INDONESIAN ECONOMY (EKO 3163)

Aloysius Gunadi Brata / aloy.gb@gmail.com / 0821-3440-9573

Description
This course provides a review on the recent developments of Indonesian economy.
Topics are chosen to show several aspects of (recent) Indonesian economy.
Organization The course will be delivered in presentation and discussion with active
participation of each student. Class attendance is required. The class will be formed
into (….) groups (depending on the number of students). Each group will
responsible to present selected topic(s) or article(s) to the class. Prepare your power
point for 30-45 minutes presentation that will be followed by comments/questions
from other groups or other students (60 minutes). Mid-term test and final test will be
determined later.

Grading
Presentation (group) : 20%
Classroom participation (individual) : 20%
Midterm test (individual) : 30%
Final test (individual) : 30%
TOTAL : 100%

Topics
W1: Introduction

W2: Eve Warburton (2016) Jokowi and the New Developmentalism, Bulletin of
Indonesian Economic Studies, 52:3, 297-320, DOI:
10.1080/00074918.2016.1249262

W3: Paul K. Gellert (2019): Neoliberalism and altered state developmentalism in

the twenty-first century extractive regime of Indonesia, Globalizations,
16(6): 894-918, DOI: 10.1080/14747731.2018.1560189

W4: Hal Hil (2018) Asia’s Third Giant: A Survey of the Indonesian Economy,
ECONOMIC RECORD, 94(307): 469–499, https://doi.org/10.1111/1475-
4932.12439

W5: Ross H. McLeod & Sitta Rosdaniah (2018) An Evaluation of Some Key
Economic Policies, Bulletin of Indonesian Economic Studies, 54:3, 279-306,
DOI: 10.1080/00074918.2018.1548245

W6: Timmer, Charles Peter (2018) : Pro-poor growth in Indonesia: Challenging

the pessimism of Myrdal's Asian Drama, WIDER Working Paper, No.
2018/103, ISBN 978-92-9256-545-9, The United Nations University World
Institute for Development Economics Research (UNU-WIDER), Helsinki,
http://dx.doi.org/10.35188/UNU-WIDER/2018/545-9
W7: Benjamin A. Olken (2019) Designing Anti-Poverty Programs in Emerging
Economies in the 21st Century: Lessons from Indonesia for the World,
Bulletin of Indonesian Economic Studies, 55:3, 319-339, DOI:
10.1080/00074918.2019.1690411

MID TEST/UTS

W8: Jordan Kyle (2018) Local Corruption and Popular Support for Fuel Subsidy
Reform in Indonesia, Comparative Political Studies, 51(11), 1472–1503, DOI:
10.1177/0010414018758755

W9: Arifin Rosid, Chris Evans & Binh Tran-Nam (2018) Tax Non-Compliance and
Perceptions of Corruption: Policy Implications for Developing Countries,
Bulletin of Indonesian Economic Studies, 54:1, 25-60, DOI:
10.1080/00074918.2017.1364349

W10: Stephen V. Marks (2017) Non-Tariff Trade Regulations in Indonesia:

Nominal and Effective Rates of Protection, Bulletin of Indonesian Economic
Studies, 53:3, 333-357, DOI: 10.1080/00074918.2017.1298721

W11: Fukunari Kimura & Lurong Chen (2018) Value Chain Connectivity in
Indonesia: The Evolution of Unbundlings, Bulletin of Indonesian Economic
Studies, 54:2, 165-192, DOI: 10.1080/00074918.2018.1505412

W12: Natasha Hamilton-Hart (2019) Indonesia’s Quest for Food Self sufficiency:
A New Agricultural Political Economy?, Journal of Contemporary Asia,
49(5): 1-26, DOI: 10.1080/00472336.2019.1617890

W13: Martha Maulidia, Paul Dargusch, Peta Ashworth, Fitrian Ardiansyah (2019)
Rethinking renewable energy targets and electricity sector reform in
Indonesia: A private sector perspective, Renewable and Sustainable Energy
Reviews 101: 231-247, https://doi.org/10.1016/j.rser.2018.11.005

W14: REVIEW/FINAL TEST/UAS

Note: You may skip technical stuff in the articles (econometric models, mathematical
formulas, or estimation results).
 Bulletin of Indonesian Economic Studies

ISSN: 0007-4918 (Print) 1472-7234 (Online) Journal homepage: http://www.tandfonline.com/loi/cbie20

Jokowi and the New Developmentalism

Eve Warburton

To cite this article: Eve Warburton (2016) Jokowi and the New Developmentalism, Bulletin of
Indonesian Economic Studies, 52:3, 297-320, DOI: 10.1080/00074918.2016.1249262

To link to this article: http://dx.doi.org/10.1080/00074918.2016.1249262

Published online: 15 Feb 2017.

Submit your article to this journal

View related articles

View Crossmark data

Citing articles: 1 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=cbie20

Download by: [FU Berlin] Date: 16 February 2017, At: 03:43

 Bulletin of Indonesian Economic Studies, Vol. 52, No. 3, 2016: 297–320

Indonesian Politics in 2016

JOKOWI AND THE NEW DEVELOPMENTALISM

Eve Warburton*
The Australian National University

Indonesia’s president, Joko Widodo (Jokowi), made a remarkable political recovery

in 2016. During his first year in office, Jokowi had been overwhelmed by a combat-
ive and divided parliament, disunity in cabinet, tensions with his own party, and
declining approval ratings. In 2016, however, Jokowi expanded his ruling coalition
and consolidated his power, and his approval rating rose to almost 70%. By mid-2016,
the president had achieved stable government for the first time since winning office.
Political stability gave us greater clarity on Jokowi’s agenda and the kind of Indonesia
he wants to shape. I suggest that in 2016 a Jokowi-styled new developmentalism be-
gan to emerge. Jokowi’s administration focused narrowly on infrastructure and de-
regulation; other problems of government were subordinated to these developmen-
talist goals. There are uncanny echoes of the past in the new developmentalism, and
its conservative and nationalist features reflect political trends that pre-date Jokowi’s
presidency. Indeed, Jokowi’s developmental strategy is neither unique nor coherent;
his decision-making is defined by ad hocery. Instead, I argue, deeper structural fea-
tures of Indonesia’s socio-political landscape are making their mark on the president
and returning Indonesia to its developmentalist moorings.

Keywords: politics, developmentalism, political parties, cabinet, presidency

JEL classification: D72, D73, O20

INTRODUCTION
In 2016, President Joko Widodo (Jokowi) made a remarkable political recovery.
During his first year in office, from October 2014, he faced multiple crises: a series
of political missteps, conflict within his cabinet, a disruptive opposition coalition
in parliament, and troubled relations with his own party. These misfortunes left
the president looking weak and out of his depth. His administration also appeared
ill-equipped to deal with Indonesia’s economic slowdown and rising inflation. By
June 2015, Jokowi’s public approval rating had sunk to just 41% (Indikator 2016),

* I wish to thank Edward Aspinall for his invaluable guidance and feedback in preparing
this article. Thanks also to reviewers at BIES, to Marcus Mietzner for comments on an earli-
er draft, and to Greg Fealy, Thomas Power, Liam Gammon, Burhanuddin Muhtadi, Colum
Graham, Ahmad Muhajir, Danang Widoyoko, and Thushara Dibley for their feedback on
the presentation of this article at the Indonesia Update in September 2016.

ISSN 0007-4918 print/ISSN 1472-7234 online/16/000297-24 © 2016 ANU Indonesia Project

http://dx.doi.org/10.1080/00074918.2016.1249262
298 Eve Warburton

his lowest since coming to office. As a result, some analysts characterised Jokowi
as the weakest of Indonesia’s presidents (Muhtadi 2015; Wall Street Journal, 23 July
2015).
Over the course of 2016, however, Jokowi managed to consolidate his political
power. By year’s end, he had transformed his governing coalition from a minor-
ity, with just 37% of seats in parliament, to a strong 69% majority (table 1). As a
result, the opposition Red-and-White Coalition was no longer a serious threat.
After a second cabinet reshuffle, in July 2016, analysts concluded that Jokowi
was now the undisputed ‘boss’ of this administration (Jakarta Post, 28 July 2016).
Inflation also cooled, the rupiah gained ground, and Indonesia’s annual rate of
GDP growth was predicted to increase from 4.8% in 2015 to 5.1% in 2016 (World
Bank 2016). In response, the president’s approval rating improved steadily from
October 2015, and surveys in August 2016 reported public satisfaction with
Jokowi to be as high as 68% (Indikator 2016).
Jokowi’s political recovery is the focus of this article. The analysis is struc-
tured around three driving questions: How did Jokowi consolidate his power
and reshape the political map over the course of 2016? In the process, did Jokowi
break with established patterns of post-Soeharto politics or simply perpetuate
them? And now that he has achieved stability, what is Jokowi’s agenda and what
kind of Indonesia is he shaping?
In response to the first question, I offer a narrative account of three strategies
the president used to consolidate his political position. First, Jokowi intervened
in the internal affairs of opposition parties and took an assertive approach to
coalition-building. Second, he asserted his authority over friends and foes alike in
a ruthless cabinet reshuffle. And, finally, he cultivated strategic alliances outside
of political parties to shore up his political position. Jokowi’s forceful tactics and
shrewd political manoeuvrings revealed a hard-headed dimension to his leader-
ship style that was absent during his first 12 months in office.
In answer to question two, I suggest that Jokowi’s hard-nosed style consti-
tutes a departure from the more conciliatory methods of his predecessor, Susilo
Bambang Yudhoyono. Ultimately, however, Jokowi applied the tried-and-true
strategy of accommodation that is the hallmark of post–New Order politics. To
consolidate power, the president had to expand his governing coalition, accom-
modate vested interests, strike deals with oligarchs, and engage in patronage dis-
tribution. The imperatives of achieving stable government thus compelled Jokowi
to embrace, rather than challenge, the usual patterns of politics in Indonesia’s
patronage-driven democracy.
Question three addresses the implications of Jokowi’s political consolidation.
Once he achieved stability, Jokowi provided greater clarity on his agenda and on
the kind of Indonesia he wants to shape. I argue that in 2016 a Jokowi-styled ‘new
developmentalism’ began to emerge. Jokowi is almost exclusively focused on a
narrow set of pragmatic economic programs—specifically, infrastructure, dereg-
ulation, and de-bureaucratisation. In its design and implementation, Jokowi’s
developmentalism exhibits a renewed commitment to a state-centred national-
ism that has deep roots in Indonesian history. For Jokowi, the state functions to
provide economic services and nurture local industry, and the state sector is the
locomotive for his infrastructure boom.
The new developmentalism is not accompanied by a substantive program for
building transparent and clean government. Jokowi calculated that anti-corruption
Jokowi and the New Developmentalism 299

TABLE 1  Proportion of Seats Held in Parliament (%)

Government Opposition

July 2016 69 31
September 2015 46 54
July 2014 37 63

Source: Data from Katadata.co.id.

reforms risk disturbing his hard-won political equilibrium and might hold back
the government’s ambitious infrastructure programs. Nor is Jokowi’s administra-
tion interested in pursuing a progressive approach to human rights and justice.
During 2016, Jokowi paid attention to what counted politically, and it became
apparent that neither human rights nor historical justice was a source of political
capital for him. As he lays the groundwork for a re-election bid in 2019, his strat-
egy is to maintain authority over parliament and cabinet and to deliver a handful
of tangible economic outcomes to the electorate. Everything else falls outside of
his tunnelled vision.
There are strong echoes of the past in this new developmentalism. It recalls
the New Order’s inwardly focused, technocratic agenda, and its conservative
and nationalist features reflect trends that pre-date Jokowi’s presidency. Yet
we should not see this new developmentalism as the product of some novel or
even coherent economic strategy. Indeed, the president himself is defined by ad
hocery; Jokowi has no deeply held ideological convictions, and he is fickle in his
approach to political and economic problems. Instead, I suggest that deeper struc-
tural features of Indonesia’s socio-political landscape are making their mark on
the president, and returning Indonesia to its developmentalist moorings.

JOKOWI CONSOLIDATES HIS POWER

In 2016, the president proved his credentials as a sharp political operator, enticing
and punishing parties and politicians in order to establish control. He took what
Mietzner (2016) has described as a ‘coercive’ approach to expanding his coalition:
intervening in the internal affairs of opposition parties and ensuring pro-Jokowi
factions took control. The president also asserted his authority by way of a cabinet
reshuffle in July. He removed ministers he deemed obstructive or overly ambi-
tious and protected his personal partisans, keeping them in strategic positions
where they can command state resources and direct policy (Power 2016). Jokowi
emerged from the year looking more authoritative, but he also made several blun-
ders and revealed some shortcomings in his leadership style.

An Assertive Approach to Coalition-Building

After the 2014 elections, Jokowi managed to form a government with minority sup-
port in the national parliament, with just 37% of the seats. According to Mietzner
(2016, 215), the president was initially unperturbed and felt he could leverage his
popular mandate against a hostile opposition. Prabowo Subianto’s Red-and-White
Coalition took over key positions in parliament and obstructed the president at
every turn (Muhtadi 2015). Tensions with Megawati Sukarnoputri, the chair of the
300 Eve Warburton

Indonesian Democratic Party of Struggle (PDI–P), also meant Jokowi could not even
rely on the support of the political party of which he was a member. By the end of
2014, it was clear that the president needed to expand his governing coalition.
Jokowi had hoped to avoid using the approach of his predecessor, who pur-
sued broad-based coalitions. Yudhoyono welcomed parties with open arms,
offering them patronage in return for loyalty. According to Mietzner (2016, 214),
Yudhoyono ‘secured political stability . . . by allowing parties to extract and
exploit state resources under their control’. Critics argued that these ‘rainbow
coalitions’ hamstrung the president and left him beholden to the interests of a
party cartel (Slater and Simmons 2012). Rather than focusing on hard reforms,
Yudhoyono was compelled to satisfy and accommodate the range of interests
in his coalition (Fealy 2011). Jokowi, in contrast, took a hard-nosed approach to
coalition-building, even as he began to realise that he needed broader support.
Instead of promising patronage in the same vein as Yudhoyono, Jokowi opted for
a more interventionist approach that ensured he was less indebted to (and hence
less constrained by) his coalition partners.
Mietzner offers unique insight into the president’s coalition-building tactics,
and his recent study demonstrates that Jokowi intervened in the affairs of oppo-
sition parties in order to support pro-government factions (Mietzner 2016). In
the wake of their loss in the 2014 presidential election, when their candidate,
Prabowo, failed in his bid to become president, PAN (National Mandate Party),
Golkar, and PPP (United Development Party) held leadership ballots in which
pro-government and pro-opposition factions were pitted against one another.
PAN’s leadership ballot was relatively straightforward. In the party congress
held in March 2015, a pro-Jokowi faction led by Zulkifli Hasan narrowly defeated
Hatta Rajasa, the incumbent chair and Prabowo’s running mate. By September
2015, PAN had joined the governing coalition.
The leadership ballots of Golkar and PPP, held in late 2014, were more com-
plicated. In both cases, pro-government and pro-opposition parties claimed
victory, triggering months of legal battles. In Golkar, Agung Laksono headed
a pro-Jokowi faction that challenged long-time party chair Aburizal Bakrie, a
prominent political oligarch who had led Golkar into opposition. In PPP, the
incumbent, Suryadharma Ali, faced off against a pro-government faction led by
Muhammad Romahurmuziy. In both parties, the ballots caused deep divisions,
which the president was able to manipulate for his own purposes. As Mietzner
(2016, 217) explains, internal ruptures allowed Jokowi to activate a rarely used
presidential tool: ‘the authority to recognize a specific faction as the legal repre-
sentative of the party’.
This authority comes in the form of a letter of decision from the Ministry of Law
and Human Rights. The minister, Yasonna Laoly, issued such letters recognising
the pro-government factions as the legal representatives of both Golkar (in March
2015) and PPP (in October 2014). Formal government recognition gives a party’s
leadership the legal status to make strategic decisions, including nominating can-
didates for district elections. Eventually, government pressure compelled both
political parties to hold a fresh ballot. In both Golkar and PPP, pro-government
factions took control of the party and entered Jokowi’s coalition.
The level of government intervention in Golkar’s internal affairs was particu-
larly unusual. Once the Ministry of Law and Human Rights had issued a letter of
Jokowi and the New Developmentalism 301

decision recognising Agung Laksono’s leadership of the party, Luhut Panjaitan,

Jokowi’s close ally and the then coordinating minister for political, legal, and
security affairs, used his personal networks within Golkar to bolster the pro-
government faction and pressure Bakrie to step aside. In what Mietzner (2016,
220) describes as a ‘remarkable intervention by the government in party manage-
ment’, the president compelled Bakrie to hold an extraordinary leadership ballot
in which he would not compete. Luhut reportedly advised Bakrie to take a break
from his political career and focus on rebuilding his corporate empire, which had
suffered from spiralling debt for years (219). Insiders suggested that as long as
Bakrie remained chair of Golkar, the government would deny him and his compa-
nies access to lucrative contracts and licences (219). Rather than offering patronage
or political posts to lure Golkar into the coalition, Jokowi and Luhut withheld legal
recognition and threatened to exclude Bakrie, a once powerful oligarch, from the
material rewards of government. The strategy worked and Bakrie stepped down.
Golkar held an extraordinary leadership ballot in June 2016 with eight competing
candidates, all of whom professed support for the Jokowi administration (220).
Luhut went even further, however, and backed a particular candidate for the
Golkar chair. Luhut continued to exploit his networks within Golkar to ensure
victory for the controversial Golkar operator Setya Novanto, who had allegedly
been involved in numerous corruption cases. The media followed the Golkar
ballot closely, and on several occasions the government was forced to deny that
Luhut’s support for Setya amounted to an official endorsement (Kompas, 11 May
2016). But many analysts believe the president agreed to the plan (Detik.com, 16
May 2016). Indeed, as Mietzner (2016) argues, Setya was an attractive option for
Jokowi on two counts: first, while parliamentary Speaker (2014–15), Setya had
assured clear passage of the government’s budget and other key legislation at a
time when Jokowi faced pressure from the opposition Red-and-White Coalition
and his own party, PDI–P. Second, Setya is a highly compromised figure and is
unpopular with the public, owing to his alleged involvement in a string of high-
profile corruption scandals (Budiartie and Warburton 2015; Jakarta Post, 23 May
2016; Tempo, 17 Nov. 2015). Setya understands his political limitations and has
no presidential ambitions. The other prominent candidate for the Golkar lead-
ership was Ade Komaruddin, a charismatic and ambitious figure who, had he
risen to the head of the party, would have been more difficult to control and
could potentially even run for president in 2019 (Mietzner 2016, 220). In a shrewd
move, Jokowi and Luhut ensured the victory of a controversial but useful ally. So
not only does the president now have a safe majority in parliament, with PAN,
Golkar, and PPP inside government; he also has a partisan supporter atop Golkar,
the second-largest party in parliament, with 91 seats.

The Cabinet Reshuffle

Having tamed parliament, the president then turned to his cabinet. In July 2016,
Jokowi conducted a ruthless reshuffle in which he not only rewarded his coalition
partners—as is the convention in Indonesian politics—but also removed obstruc-
tive or overly ambitious ministers. Jokowi dismissed five ministers, transferred
five others into new positions, and made nine new appointments. A consensus
formed quickly among commentators that Jokowi had finally flexed his politi-
cal muscles. Yet while commentators welcomed Jokowi’s assertive approach,
302 Eve Warburton

the process itself revealed high levels of dysfunction and the cabinet lost several
reform-minded ministers. Some of the president’s decisions were also made with
a striking lack of forethought or consultation, revealing an autocratic impulse in
Jokowi’s management style.
Jokowi’s reshuffle was driven by political expediency, and his principal goal
was to secure support for the 2019 elections. To this end, Jokowi rewarded his
new coalition members with cabinet posts. He put PAN’s Asman Abnur in charge
of the Ministry of Administrative and Bureaucratic Reform, a position formerly
held by Yuddy Chrisnandi of Hanura, and gave Golkar’s Airlangga Hartarto the
Ministry of Industry, replacing Saleh Husin (also of Hanura). These appointments
gave the new cabinet the feel of a Yudhoyono-style ‘rainbow’ cabinet. There are
important differences, however. Jokowi was far less willing than Yudhoyono to
bend to the wishes of his new coalition partners. For example, even though Golkar
is the second-largest party in parliament, it was offered only one ministry. PPP,
meanwhile, received no extra cabinet posts after joining the coalition. This stands
in contrast to Yudhoyono’s approach, in which he distributed posts in loose pro-
portion to each party’s seats in parliament.
Jokowi also pushed back against the demands of his original coalition part-
ners. For example, journalists and political analysts reported that Surya Paloh,
chair of the National Democrats (NasDem), had asked that his close ally, real-
estate businessman and fellow NasDem politician Enggartiasto Lukita (Enggar),
be given the Ministry of Land and Spatial Planning (pers. comm. 13 and 16 Aug.
2016); instead, Jokowi gave Enggar the Ministry of Trade. Meanwhile, Megawati
Sukarnoputri, the chair of PDI–P, had pressured Jokowi consistently for two years
to appoint her former adjutant General Budi Gunawan as chief of Indonesia’s
National Police. Jokowi initially agreed, but controversy erupted in 2014 when
the Corruption Eradication Commission (KPK) named Gunawan as a corrup-
tion suspect (Muhtadi 2015). The president subsequently resisted Megawati’s
demands and, in August 2016, appointed a younger, reform-minded officer,
Tito Karnavian, to this powerful position. Jokowi made Gunawan the head of
Indonesia’s National Intelligence Agency (BIN)—a senior but in fact less prestig-
ious and less powerful role within Indonesia’s security architecture. The president
was thus managing to both accommodate and control the parties in his coalition.
What emerged was a very different picture from the one analysts painted a year
earlier, of a president overwhelmed and trapped by Indonesia’s oligarchic politi-
cal system (Muhtadi 2015).
Yet the reshuffle also revealed the president’s shortcomings. Several contro-
versies indicated poor judgement on Jokowi’s part or an unwillingness to consult
with his advisors. For example, Jokowi appointed Wiranto, a New Order gen-
eral and the chair of Hanura, to the prestigious position of coordinating minister
for political, legal, and security affairs. Clearly, Jokowi did not seek wide coun-
sel before making this appointment. In 2003, Wiranto was indicted for crimes
against humanity by the United Nations for his role in the military-backed
massacres of East Timorese in 1999. Jokowi was either not aware of Wiranto’s
criminal past or did not believe the appointment would tarnish Indonesia’s inter-
national image. In any case, the decision illustrates a striking lack of attention to
political optics and little sensitivity to problems of security reform and historical
justice in Indonesia.
Jokowi and the New Developmentalism 303

The ruckus surrounding the new minister for energy and mineral resources
was particularly embarrassing for the president and his advisors. Jokowi removed
Sudirman Said, who was widely regarded to have been the country’s first reform-
ist energy minister in more than a decade, and replaced him with Arcandra Tahar,
a little-known engineer. (Sudirman had fallen out of favour with Jokowi after hav-
ing several public spats with other cabinet ministers.) Arcandra, however, was a
naturalised American citizen, which proved a serious problem because Indonesia
does not allow dual citizenship. Once the press got wind of Arcandra’s American
citizenship, a media firestorm raged and Jokowi was forced to remove him from
his post after just 20 days (Kompas, 15 Aug. 2016). Despite the embarrassment of
this bungled appointment, the president was reportedly adamant that Arcandra
be reinstated. The pair had built up a rapport, and so Jokowi ordered his govern-
ment to do everything it could to resolve the legal issues surrounding Arcandra’s
citizenship.
Two months later, in a surprise move that few observers could explain, the
president appointed Ignasius Jonan as the new energy minister, with Arcandra as
his deputy. Jokowi had fired Jonan from the transport ministry in the July reshuf-
fle, ostensibly for his lack of commitment to Jokowi’s policy agenda. Jonan had
no experience in the extractives sector and was told of his new role only two
hours before his inauguration. Neither the vice-president, Jusuf Kalla, nor the
coordinating minister for maritime affairs and natural resources, Luhut Panjaitan,
attended the ceremony, because they too were informed of the decision only at
the last minute (Jakarta Post, 15 Oct. 2016). This kind of erratic decision-making
has, for Jokowi, become something of a pattern.
Several reformist ministers were removed as well. These changes illustrate the
president’s pivot away from reform and towards political expediency. For exam-
ple, Anies Baswedan, a former rector of Paramadina University, was removed
from the Ministry of Education and Culture to make room for a Muhummadiyah
representative, Muhadjir Effendy.1 The palace reported that Jokowi was disap-
pointed with Baswedan’s handling of the Smart Indonesia Card, which is one of
the president’s signature programs and provides poor Indonesians with subsi-
dised education (Merdeka.com, 27 July 2016). Yet the consensus among insiders
is that Jokowi also harboured concerns about the charismatic minister’s political
aspirations and feared Anies might make a bid for the presidency in 2019. Since
leaving cabinet, Anies has indeed pursued a political career and he will compete
in Jakarta’s gubernatorial elections in 2017.
In appointing NasDem’s Enggar as the new minister for trade, Jokowi made
the outgoing Thomas Lembong the head of the Investment Coordinating Board
(BKPM)—a demotion for a reform-minded minister who many felt had the poten-
tial to invigorate Indonesia’s trade relations and roll back protectionist policies.
Lembong’s liberal approach to trade policy had irked the more nationalist politi-
cians in Jokowi’s cabinet, and the president himself became uncomfortable with
Lembong’s market-oriented position on food imports. The president decided
instead to hand the reins of trade policy to one of his coalition partners. Enggar,

1. Muhummadiyah is one of Indonesia’s largest Islamic organisations. It is not a political

party, but it received representation in Yudhoyono’s cabinet in return for political backing.
304 Eve Warburton

like Jonan, brought no specific experience to this new role, and in fact this was
the first time since 1999 that an active party politician had held the trade portfolio
(Jakarta Post, 4 Aug. 2016).
The reshuffle demonstrated that Jokowi’s priorities in 2016 were politics over
good governance, and stability over reform. Throughout the year, we gained
fresh insight into the president’s leadership style as well. Jokowi is calculating
and often hard-nosed in his approach to ministers, advisors, and even his closest
confidantes. His bringing back Sri Mulyani Indrawati as finance minister, a role
in which she gained wide respect during Yudhoyono’s presidency, showed fore-
sight and a commitment to sound financial management during a time when the
budget was under significant pressure. Yet, in other instances, Jokowi’s decision-
making was inept, and the president displayed strikingly poor judgement when
it came to several strategic appointments. The overall picture is one of inconsist-
ency. Jokowi manages government in an unsystematic way, giving the impres-
sion of a president defined largely by ad hocery.

The President’s Enablers

Jokowi could not have consolidated power without a small group of loyal ‘ena-
blers’—wealthy politico-business elites and former generals with the finan-
cial resources and political networks that the president needs in lieu of a party
machine. Power (2016), in his post-reshuffle analysis, made the compelling argu-
ment that Jokowi protects these personal partisans and keeps them in positions
known to be rich in rent and patronage resources, in large part because they are
crucial for his re-election in 2019.
Among Jokowi’s inner circle, Rini Soemarno is arguably his closest ally. Rini is
the minister for state-owned enterprises (SOEs), and she shapes the president’s
thinking on policy areas from infrastructure to energy. Rini played a key role in
Jokowi’s rise to the presidency, having helped to finance and organise his bid for
office (Power 2016). After the election, in mid-2014 Jokowi made Rini head of the
transition team that would help him design his first cabinet and lay out a strategic
plan for his government’s policy agenda. Rini quickly fell out of favour with PDI–P
over the distribution of cabinet positions, and, in 2016, Jokowi’s closeness to Rini
remained a sore point in the president’s relations with his party and its matriarch,
Megawati. For almost two years, PDI–P privately and publicly attacked Rini’s cre-
dentials and integrity, and Megawati consistently demanded Rini be removed from
cabinet (Tempo, 10 Jan. 2016). But Jokowi refused, instead keeping Rini in a strategic
post where she could command state resources and direct key government policies.
Another of the president’s close allies is Amran Sulaiman (Power 2016), a
prominent businessman (with interests primarily in agribusiness) who helped
finance and organise Jokowi’s presidential campaign. Jokowi gave Amran the
Ministry of Agriculture—another portfolio known to be rich in patronage oppor-
tunities—with the authority to distribute agricultural concessions and import
licences. Jokowi is reportedly fond of Amran, and has allowed him to drive the
government’s policy on food security and agricultural self-sufficiency. Despite
fluctuating food prices, which can be linked directly to Amran’s policies, his posi-
tion as minister remained safe (Jakarta Post, 4 Aug. 2016; Power 2016).
Luhut Panjaitan has been another key enabler since the early days of Jokowi’s
presidential campaign. Jokowi has relied on Luhut for both financial resources
Jokowi and the New Developmentalism 305

and political strategy. On coming to power, Jokowi initially made Luhut his chief
of staff at the palace, and then, in 2015, appointed him coordinating minister for
political, legal, and security affairs. From this post, Luhut used his personal net-
works and political acumen to help the president pacify the opposition and estab-
lish majority government. Yet the president gradually became uncomfortable with
Luhut’s high public profile. Luhut interpreted his brief broadly, and took the lead
on everything from internal and international security issues to natural-resource
contracts, trade, and even problems of poverty reduction. Jokowi reportedly dis-
liked the outward perception of the former general’s influence over the palace, so
in the July 2016 reshuffle Jokowi moved Luhut to the less prestigious position of
coordinating minister for maritime affairs and natural resources (Jakarta Post, 28
July 2016). As Power (2016) points out, though, ‘Luhut is now overseeing energy
and resources, sectors in which his company PT Toba Sejahtra has significant
holdings, and which remain particularly ripe for rent-seeking’. In other words,
the president placed one of his key enablers and political financiers at the apex of
the country’s most important and lucrative sectors.
In summary, since his campaign in 2014, Jokowi has not been able to rely on his
party for consistent support and so has cultivated personal alliances with power-
ful and wealthy individuals outside political parties. According to Power (2016),
Jokowi will keep these individuals in strategic ministries as a strategy to ‘resource
and manage a re-election bid that limits his indebtedness to parties’.

JUST A ‘NORMAL’ POLITICIAN?

This brings us to the second driving question of the analysis: in the process of
consolidating power and achieving stability, has Jokowi broken with established
patterns of post–New Order politics or simply perpetuated the conventions of
Indonesia’s patronage-driven democracy?
I argue that Jokowi’s interventionist tactics mark a departure from the more
conciliatory methods of his predecessor. Yudhoyono secured stable government
by welcoming parties inside the governing coalition, promising them patronage,
and providing them with cabinet positions. Yudhoyono saw himself as a ‘mod-
erator’, and his role was to mediate competing interests within Indonesia’s frag-
mented democratic polity (Aspinall, Mietzner, and Tomsa 2015). For this reason,
he sought large, inclusive political coalitions and avoided any actions or policy
decisions that might alienate coalition members. Yudhoyono also accommodated
competing interest groups and rarely disciplined recalcitrant ministers or coali-
tion partners (Aspinall, Mietzner, and Tomsa 2015). As a result, he earned a repu-
tation for equivocation and for avoiding tough decisions.
Jokowi, in contrast, has proved to be more hard-nosed in his approach to
coalition-building. As we have seen, during 2015 and 2016 he intervened in
opposition party affairs to support pro-government factions, and offered these
parties fewer rewards in the way of ministerial posts and patronage resources.
Jokowi’s decision to back Setya Novanto for the Golkar chair was a particularly
potent demonstration of the president’s embrace of realpolitik and, according to
Mietzner (2016), of his willingness to engage in a level of intervention in party
affairs that was unprecedented in the democratic era. When it came to the reshuf-
fle, Jokowi asserted his authority, removing obstructive ministers—some of them
306 Eve Warburton

reformists—and those he felt displayed too much political ambition. In 2016, we

saw a Machiavellian streak in Jokowi’s leadership style that was not on display
during his first 12 months in office. To be sure, the hostility that Jokowi faced in
parliament, and the fractiousness within his own cabinet, necessitated a particu-
larly assertive response. Nonetheless, the president’s calculating and, at times,
uncompromising nature surprised commentators and insiders alike.
Some of Jokowi’s tactics and his personal style might stand out as different, but
the president ultimately stuck to a strategy of accommodation that has been the
hallmark of post-Soeharto politics. When Jokowi first took office he flirted with
the idea of doing things differently. He thought he could rule with a minority
in parliament and build a coalition tanpa syarat, or without conditions. Jokowi
wanted his government to look different from Yudhoyono’s and to avoid the
rainbow coalitions for which his predecessor was famous.
Jokowi’s tumultuous first year in power proved that such a course of action
brings too much political risk. And so, in pursuit of power and stability, the presi-
dent embraced established patterns of post-Soeharto politics. He expanded his
coalition to form a large majority in parliament. In the process, the president cre-
ated a broad-based coalition of the sort he had shunned before coming into office.
Jokowi struck deals with oligarchs and placed wealthy, powerful partisans in
strategic posts, giving them control over state resources and authority over lucra-
tive industries. The president and his allies accommodated vested interests in
order to overcome an aggressive opposition in parliament. Like his predecessor,
Jokowi placed a premium on stability and pursued the kind of political alliances
that would deliver a coherent political landscape.
It appears, therefore, that over his first two years in office the imperatives of run-
ning a stable government in Indonesia’s fractious, patronage-driven democracy
left their mark on a president who had set out to do things differently. Reflecting
on Jokowi’s second year as president, one member of cabinet concluded: ‘In 2016,
it became clear to me and others in cabinet that in fact Jokowi was just a “normal”
politician’ (pers. comm. 25 Aug. 2016).

THE NEW DEVELOPMENTALISM

Now that Jokowi has consolidated his power and achieved political stability, we
have more insight into his agenda and what he wants for Indonesia. I suggest that
in 2016 a Jokowi-styled new developmentalism began to emerge. It is pragmatic
and nominally pro-poor in its policy agenda and demonstrates a renewed com-
mitment to a statist–nationalist ideology. This new developmentalism is rooted
in the past. It bears uncanny parallels with the New Order’s emphasis on tech-
nocratic development programs, and its statist and nationalist features have long
loomed large in Indonesian economic planning. Yet the combination of Jokowi’s
personal agenda and leadership style means that a narrow and no-frills version of
an old developmental model is taking shape in contemporary Indonesia.
The term ‘developmental state’ was originally used to capture the East Asian
model of economic growth (Woo-Cumings 1999). After the Second World War,
Japan, Taiwan, and South Korea adopted interventionist economic policies
and established efficient bureaucracies, both of which drove fast-paced indus-
trialisation. I do not use this term to evoke the characteristics of an ideal-type
Jokowi and the New Developmentalism 307

developmental state; indeed, Indonesia has never displayed the kind of bureau-
cratic efficiency that is characteristic of the East Asian developmentalist model.
Nor do I use it in the same way as scholars of the New Order. Back then,
Indonesia constituted what Feith (1981) called a ‘repressive developmental state’.
Under Soeharto, all functions of government were geared towards creating ame-
nable conditions for economic growth. The state justified its co-optation of soci-
etal groups and repression of political dissent in the name of accelerated economic
development (Feith 1981, 502). In this context, policy was exclusively the task of
bureaucratic elites, technocratic professionals, and the military, rather than that
of political or legislative elites (Leftwich 1995). These institutions of repression no
longer exist in contemporary Indonesia.
Instead, I use ‘developmentalism’ to describe how ideas and practices associ-
ated with the developmental paradigm have risen to prominence under the Jokowi
administration. Developmentalism is driven by ‘the idea that the task of the state
is to achieve fast development to overcome . . . backwardness and catch up with
advanced countries’ (Feith 1981, 502). Other political goals are subordinated to
the overarching aim of achieving rapid economic growth. State intervention in
economic life is a necessary ingredient for accelerating national development and
achieving industrialisation (Leftwich 1995). I suggest that Jokowi and his admin-
istration are committed to this core developmental agenda and espouse a related
set of narrow, pragmatic economic policy goals. (Though, as we shall see, their
success is far from guaranteed.)
The new developmentalism is in many ways conservative; it is characterised
by an aversion to politically sensitive problems of law reform, corruption, and
even good governance. Such institutional challenges are, in the eyes of the presi-
dent, subordinate to the more urgent goal of fast-paced economic development.
In addition, Jokowi apparently believes that pursuit of a reformist agenda may
jeopardise his hard-won political stability. Nor is Jokowi interested in pursuing a
progressive approach to problems of civil and political rights or historical justice.
In fact, although the Jokowi administration is not resurrecting the repressive tools
of the New Order model, we can observe in this new developmentalism a grow-
ing impatience with liberal reform and an indifference towards human rights.
The rest of this section is dedicated to explaining the new developmentalism
and its pragmatic policy focus, its statist–nationalist orientation, and its conserva-
tive approach both to problems of law reform and good governance and to issues
of rights and justice.

Jokowi’s Narrow Agenda

Jokowi views the state’s principal task as service delivery. His administration
focuses almost exclusively on a restricted set of practical services and tangible
economic outcomes. Famous initially for his free health and education programs
for the poor, the president now pours much of his energy into promoting the
government’s agenda for an infrastructure boom. While the president still frames
his infrastructure plans as a key component in addressing inequality and poverty,
in 2016 there was a notable shift away from Jokowi’s previous focus on pro-poor
and populist policies and towards a growth-focused developmentalist agenda.
In his state-of-the-nation address in August 2016, Jokowi proclaimed 2016 the
year of accelerated national development. His speech was domestically oriented,
308 Eve Warburton

with little time spent on foreign policy and with no Yudhoyono-style platitudes
about transparency, justice, or democracy. Instead, his speech concentrated on
the government’s program for reducing inequality and poverty, summed up in a
three-word mantra—infrastructure, deregulation, and de-bureaucratisation. For
infrastructure, the government has set ambitious (and most would argue impos-
sible) goals: to deliver 35,000 megawatts of electricity to the grid; to upgrade and
develop five port hubs and 19 feeder ports; to build 3,650 kilometres of new roads;
and to achieve 100% access to clean water nationwide. The program is set to cost
over $411 billion, of which only half will come from the state budget (Bloomberg.
com, 1 July 2016).
Jokowi’s deregulation policies are designed to cut red tape and attract infra-
structure investment. De-bureaucratisation similarly aims to streamline and
reduce government procedures in order to accelerate infrastructure projects. At
public events around Indonesia, in front of business people, regional parliamen-
tarians, and foreign investors, Jokowi speaks with a rare energy about his plans
to cut red tape. On this topic, Jokowi is unusually engaging; he tends to read
verbatim from prepared texts, and most of his speeches are wooden and colour-
less. But on his three favourite issues, the president can go off-script, speak ener-
getically, and elicit cheers of approval. Indeed, since the early days of his political
career, Jokowi has consistently expressed frustration with Indonesia’s burden-
some bureaucratic procedures (Mietzner 2015, 25). This is where the president’s
passions lie.
Jokowi and his government market all of these policies as pro-poor. Jokowi
maintains that an infrastructure boom will alleviate regional inequality, cre-
ate economic opportunity in the outer islands, and jumpstart stagnating growth
(Presidential Office, press release, 19 Nov. 2015). The government argues that food
and services costs are higher in Indonesia’s outer islands because of poor infra-
structure, and that improvements in physical infrastructure will bring down costs
and therefore prices. The government’s 13 recent economic deregulation packages
are designed to attract investment in these infrastructure projects. The goal is to
revitalise Indonesia’s investment regime, with the ultimate goal of modernising its
physical infrastructure.2 While nominally pro-poor, none of these policies is explic-
itly directed at Indonesia’s poorest citizens—unlike the subsidised health and edu-
cation programs Jokowi introduced during his time as Jakarta governor. Arguably,
Jokowi’s pro-poor image has lost some of its shine over the past two years.3
Jokowi’s infrastructure drive is not particularly new. In fact, most of the infra-
structure projects identified for investment under the Jokowi administration were
part of Yudhoyono’s Master Plan for the Acceleration and Expansion of Indonesian
Economic Development 2011–2025 (Davidson 2016). Jokowi, however, has sig-
nalled a more aggressive commitment than Yudhoyono to the developmental-
ist agenda. First, he has provided far more funding. After cutting fuel subsidies,
Jokowi was able to allocate $23.9 billion (Rp 290 trillion) to the government’s

2. Many economists are sceptical about the overall impact these packages will have on
growth (see Hamilton-Hart and Schulze 2016, in this issue)
3. Of course, the president is likely to reinvigorate his pro-poor and populist credentials
with new policies as he approaches the 2019 elections.
Jokowi and the New Developmentalism 309

infrastructure projects, an increase of 86% on the previous government’s budget

allocation (Davidson 2016). Second, the modernisation of Indonesia’s infrastruc-
ture constitutes a more central component of Jokowi’s political agenda. In 2016,
the administration spent comparatively little energy on marketing education,
health, or other social programs. Infrastructure development has become a hall-
mark of Jokowi’s government in a way it never did for Yudhoyono’s.

Statist–Nationalist Ideological Orientation

The new developmentalism also exhibits a renewed commitment to a statist–
nationalist ideology that is typical of developmentalism and has deep roots in
Indonesia. To be sure, the Jokowi government has selectively embraced liberal
and market-oriented policies over the last two years. Its 13 economic deregula-
tion packages are designed to encourage private-sector investment, and the 10th
package also lifted caps on foreign investment for 35 industries. The president is
acutely aware that his $411 billion infrastructure plans will require a significant
contribution from the private sector (Bloomberg.com, 1 July 2016).
But programs of liberalisation are highly circumscribed. The president sees
liberalisation as the last resort for attracting capital under conditions of seri-
ous budget constraint. Indeed, the cabinet secretary, Pramono Anung, speaking
on behalf of the government, was at pains to frame the deregulation packages
as an effort to ‘modernise, not liberalise’ Indonesia’s investment regime (BBC
Indonesia, 12 Feb. 2016). Throughout the Soeharto era, economic planning oscil-
lated between market-oriented policies, on the one hand, and a statist–nationalist
ideology that favoured the expansion of SOEs, import substitution, subsidies, and
state-funded industrial projects, on the other hand (Hill 1994). A similar pattern
pertains to contemporary economic policy-making.
Overall, however, the new developmentalism bears the hallmarks of a statist–
nationalist ideology. It is statist because the government views a strong and stable
state as a necessary component in accelerating national development, with the
state sector as a locomotive for economic growth; and it is nationalist because the
government justifies state intervention in the name of building state strength and
sovereignty and reducing dependence on foreign capital and international mar-
kets (Chalmers 1997, 27).
The state sector has received a significant boost under the Jokowi government
(Davidson 2016). The president and his minister for SOEs, Rini Soemarno, have
distributed special privileges to state-owned companies, provided them with
access to capital, and offered them strategic contracts (Ray and Ing 2016). The
government allocated SOEs $3 billion in additional state funds in 2015 (Wall Street
Journal, 20 Apr. 2015). And there are plans afoot to merge and streamline the SOE
sector by establishing large holding companies in key industries (Jakarta Post, 1
Mar. 2016). The stated goal is to improve the efficiency of the sector, and thereby
increase the capacity of SOEs to attract capital, so it can become a locomotive for
Indonesia’s development (AntaraNews.com, 29 Feb. 2016).
Jokowi appears convinced that the state should control imports and exports
and nurture local industry. His administration argues that reliance on foreign
goods undermines domestic productivity and the prospects for long-term indus-
trialisation. This mode of thinking has directed the government to limit imports
for agricultural produce, food, livestock, and the components for manufacturing
310 Eve Warburton

smartphones (Economist, 9 May 2015). The trade and agriculture ministries

recently proposed that beef import licences be granted on the condition that com-
panies demonstrate a commitment to cattle breeding, in order to boost the local
cattle industry and reduce reliance on imports (Jakarta Globe, 26 Sept. 2016). In
the mineral mining sector, meanwhile, Jokowi’s government continues to force
downstream industrialisation by dictating that companies may not export unpro-
cessed mineral ores (Busch 2016b; Warburton 2015). Despite the pressures of the
global commodity downturn, the government continues to maintain onerous con-
ditions for foreign investment in mining (Straits Times, 10 May 2014; Warburton,
forthcoming). The logic is that foreign companies have exploited Indonesia’s
resources for too long and it is now time to create more space for Indonesians to
extract their country’s riches. The aspirations underpinning these policies are self-
sufficiency and economic sovereignty.
The president himself has no deeply held commitment to any particular eco-
nomic ideology (Mietzner 2015, 25). He takes a rather ad hoc approach to economic
policy, trialling new ideas and listening to different sorts of economic thinkers.
Many within Indonesia’s private sector, and particularly the foreign business
community, expected that Jokowi’s experience as an entrepreneur would lead
him to embrace a more market-oriented and open approach to investment and
economic development (Wall Street Journal, 27 Aug. 2015). But Jokowi assumed
the reins of power during a time when nationalist ideas had become a prominent
feature of economic policy-making; indeed, Yudhoyono’s second term was char-
acterised by an increase in protectionism and by greater limitations on foreign
companies (Warburton 2015; Patunru and Rahardja 2015). Jokowi is riding this
wave of economic nationalism. He now leans towards the nationalist position,
reflecting the ideological persuasion and interests of many Indonesian politicians
and policymakers. The statist–nationalist sensibilities of Indonesia’s political class
are leaving their mark on the president, and he is responding to the structural
imperatives of economic policy-making in Indonesia.

No Anti-corruption Agenda
Jokowi’s new developmentalism comes with no substantive plan for dealing with
corruption or promoting clean government. Transparency and good governance
were once at the forefront of Jokowi’s political identity. He invested in this image
because he believed it to be an important source of political capital. For example,
as governor of Jakarta, Jokowi aligned himself with the anti-corruption move-
ment and publicly asked the KPK to assist him in his fight against corruption
(Kompas, 27 Nov. 2012). After his presidential victory, Jokowi also invited the KPK
to help vet potential ministers for the new cabinet.
Yet the president now eschews anything that resembles an anti-corruption cru-
sade, and avoids the kinds of institutional changes that might confront or disrupt
his hard-won political equilibrium and the strategic alliances he has built over the
past two years. Jokowi now subscribes to the idea that anti-corruption reforms are
disruptive and inefficient and hold back development—a conventional position
within Indonesia’s political establishment (Gammon 2015). Jokowi’s brief flirta-
tion with an anti-corruption agenda has come to an end, and transparency is no
longer a political priority for him.
Arguably the most powerful illustration of Jokowi’s retreat from reform
involves the Ministry of Energy and Mineral Resources. In the early days of the
Jokowi and the New Developmentalism 311

Jokowi presidency, the ministry appeared to be on the brink of an institutional

renaissance. The president and his minister, Sudirman Said, introduced several
ground-breaking reforms: the government removed petroleum subsidies that
had burdened the state budget for years; Jokowi and Sudirman established an
ad hoc team to investigate the notoriously corrupt oil-importing sector—some-
thing no previous president or minister had been brave enough to do—and then
disbanded Petral, the oil-trading arm of Pertamina, the state-owned oil and gas
company;4 the president appointed Amien Suryanadi, former deputy chair of the
KPK, to the oil and gas regulator, SKK Migas, which had been plagued by brib-
ery scandals and suffered from systemic corruption; and Sudirman shook up his
ministry’s bureaucracy, changing more than 1,000 positions and introducing ad
hoc teams for innovative policy design.
After six months, however, Jokowi’s zeal for reform began to fade. Sudirman’s
emphasis on technocratic governance earned him many enemies, including
Jokowi’s loyal strategist Luhut Panjaitan. Sudirman and Luhut disagreed on the
direction of Indonesia’s most strategic energy and mining contracts, and Sudirman
had regular public spats over these issues with Luhut’s ally, Rizal Ramli, at the time
the coordinating minister for maritime affairs and natural resources (Merdeka.
com, 24 Mar. 2016). Sudirman also challenged another of Jokowi’s trusted advi-
sors, Rini, and criticised the state-owned electricity provider, PLN, for its han-
dling of the government’s electrification program. These conflicts disrupted the
cabinet, irritating the president. Jokowi also became concerned about Sudirman’s
connections to the vice-president, Jusuf Kalla, whose networks in cabinet Jokowi
had been working to undercut (Jakarta Post, 27 July 2016).
The Freeport scandal that erupted in late December 2015 offers the most com-
pelling illustration of Jokowi’s abandonment both of Sudirman and of his own
reformist agenda. Sudirman went public with a recording of a conversation in
which Setya Novanto, then parliamentary Speaker, and disreputable oil trader
Reza Chalid are heard trying to extort business opportunities and shares out of
Freeport Indonesia, the largest foreign mining company in the country (Lubis
2015). They even name Luhut as the person able to secure presidential support for
an early extension of Freeport’s contract. The tape caused a media sensation and
dominated headlines for weeks (Burdiartie and Warburton 2015). One observer
even likened it to Indonesia’s Watergate moment (New York Times, 17 Dec. 2015).5
Sudirman led the charge against Setya, expecting the president’s support. But
it was not forthcoming. In fact, the only casualty of the Freeport scandal was
Sudirman himself. Luhut and Setya had proven themselves politically valuable
to Jokowi by helping to manage a hostile opposition in parliament. Sudirman’s
reformist crusade, on the other hand, was disruptive. As a result, while Setya was

4. For decades, an invincible set of disreputable rent-seekers monopolised Indonesia’s oil-

importing sector and used part of their profits to pay for political protection at the highest
levels. Sudirman caused controversy when he stated publicly that all previous attempts
to uncover the scheme had ‘stopped at [Yudhoyono’s] desk’, implying that the president
himself was protecting an alleged mafia (Kompas, 19 May 2015).
5. Almost a year later, in September 2016, Setya won his case in the Constitutional Court,
where he had challenged the constitutionality of the tape recording (because it was con-
ducted by a private citizen and not at the behest of law enforcement). Soon afterwards, the
parliamentary ethics committee cleared him of any wrongdoing.
312 Eve Warburton

forced to step down from his position as parliamentary Speaker, he still enjoyed
Jokowi’s backing for the chair of Golkar. In the July reshuffle, Luhut was moved to
the Coordinating Ministry for Maritime Affairs and Natural Resources (replacing
Rizal Ramli). This post gives Luhut direct oversight of the Ministry of Energy and
Mineral Resources and, thus, of the Freeport contract. Politics trumped reform in
this instance, and the president turned his back on a reformist minister and the
agenda they once seemed to share. Reflecting on the shift in Jokowi’s approach to
Indonesia’s energy sector, one senior member of the ministry said that the presi-
dent, in his second year on the job, ‘stopped listening to professionals, and instead
only listened to the political strategists in his inner circle. Their goal is not reform,
it is political success’ (pers. comm. 25 Aug. 2016).
The case of the energy and mineral resources portfolio is indicative of a broader
trend. In general, Jokowi’s approach to corruption aligns with the conservative
position of many within Indonesia’s political establishment. He argues for a pre-
ventative rather than punitive approach, and the government promotes a pre-
vention program that outlines positive thinking, integrity, and humility among
the bureaucracy as the tools for eradicating corruption (Presidential Office, press
release, 9 Dec. 2015). Jokowi has also concluded that criminalising corruption
slows down the bureaucracy and stalls development projects. Accordingly, he
has instructed the courts and the police to avoid investigating and criminalis-
ing bureaucrats and regional leaders involved in overseeing infrastructure pro-
jects (HukumOnline.com, 28 Jan. 2016; Liputan6.com, 24 Aug. 2016).6 In essence,
this instruction makes the rule of law subordinate to the goal of economic
development.7
The picture that emerges, therefore, is of a president who once saw political
value in an anti-corruption agenda. He flirted with institutional reform at a time
when, off the back of a unique grass-roots presidential campaign, it seemed plau-
sible that he could. But Jokowi learned that reform is complex and full of political
risk. In 2016, he began to pay attention to what counts politically. He was simply
no longer willing to risk the political stability he had achieved, or his infrastruc-
ture agenda, by engaging in an anti-corruption crusade. As a result, the president
opted for the expedient approach to corruption—one that is unobtrusive, uncon-
troversial, and far less threatening to the political establishment.

No Interest in Human Rights or Justice

Human rights and historical justice also slipped from Jokowi’s agenda. During
his presidential campaign, Jokowi promised that his government would investi-
gate past abuses, including the massacre of communist sympathisers in 1965–66
and state violence against political activists in 1997–98. As well as mentioning
these issues, the Nawacita, Jokowi’s campaign pledge, explicitly outlined the

6. Presidential Instruction 1/2016 on the Accelerated Implementation of Strategic National

Projects.
7. One important caveat: Jokowi has so far delayed parliament’s proposed revisions to Law
30/2002 on the KPK. These revisions would place new restrictions upon the institution’s
powers of investigation. Polls suggest that the changes are unpopular with the public, so
Jokowi is not keen to rush them through (Kompas, 8 Feb. 2016).
Jokowi and the New Developmentalism 313

protection of minorities as a political priority. At the end of his first year, how-
ever, scholars and activists expressed bitter disappointment at Jokowi’s inaction
(Hearman 2016; Hidayat 2015; Muhtadi 2015). Analysts generally concluded at
the time that, because Jokowi was ruling with a minority government, and with-
out a party machine, he had little choice but to appease conservative elites and
defer reform (Setiawan 2016; Hearman 2016). There was thus an implicit sense of
hope that if Jokowi could govern from a position of strength, then human rights
and justice might return to the political agenda.
Yet Jokowi’s political consolidation was not accompanied by a fresh commit-
ment to human rights or justice. Over the course of 2016, despite establishing his
authority over parliament and the cabinet, Jokowi was passive, if not silent, on a
number of pressing human-rights issues. For example, the president continued
to execute prisoners at alarming rates. Jokowi ordered 28 executions in just two
years; Yudhoyono executed 21 prisoners over ten years and oversaw an infor-
mal moratorium on executions between 2008 and 2013 (Amnesty International,
press release, 28 July 2016). International human-rights organisations Amnesty
International (2015) and Human Rights Watch (press release, 27 July 2016) con-
demned the Jokowi administration for carrying out an increasing number of exe-
cutions and failing to address endemic corruption in the judicial system.
Jokowi has been limp in his defence of Indonesia’s minorities. Like Yudhoyono,
Jokowi has chosen not to defend the rights of the Islamic sect Ahmadiyah, which
is regarded as heretical by most mainstream Muslims (Human Rights Watch
2016). Meanwhile, local authorities in East and West Kalimantan forcibly evicted
and detained thousands of members of the Gafatar (Gerakan Fajar Nusantara)
community—a highly heterodox religious movement that had been accused
of blasphemy. The home-affairs minister, Tjahjo Kumolo, officially banned the
Gafatar in early 2016 and introduced punishments of up to five years in prison for
those professing membership of the group (Human Rights Watch, press release,
29 Mar. 2016). When anti-Chinese violence broke out in Medan in July 2016, the
president did make a public statement calling for greater community tolerance
(Today, 2 Aug. 2016). But, in general, Jokowi has allowed a heavy-handed and
discriminatory approach towards religious minorities at both the national and
the regional level.
Perhaps most disturbing has been the president’s indifference towards an
outpouring of intolerance and vitriol against Indonesia’s lesbian, gay, bisexual,
and transgender (LGBT) community. In January 2016, prominent politicians,
mainstream Islamic organisations, and radical Islamist groups joined together
in a chorus of anti-gay rhetoric (Human Rights Watch 2016). The outburst was
triggered by a news article in which the minister for research and higher edu-
cation, Muhammad Natsir, criticised a campus group for offering support ser-
vices to LGBT students (Republika, 23 Jan. 2016). What followed was an exercise in
xenophobic-cum-homophobic outbidding, with leading public figures claiming
that the LGBT community was endangering the nation and constituted a threat as
serious as terrorism and drugs, and that gay activism was part of a foreign ‘proxy
war’ targeting Indonesia (Time, 11 Aug. 2016). Human Rights Watch (2016) called
the furore an ‘unprecedented attack’ on the rights of LGBT Indonesians. Jokowi
gave no public statement either condemning or condoning these attacks on the
LGBT community.
314 Eve Warburton

The Jokowi government has made limited progress towards realising its cam-
paign promise of justice for the victims of state violence. In April 2016, a con-
sortium of civil-society and government organisations held a symposium on the
anti-communist massacres of 1965–66: ‘Dissecting the 1965 Tragedy: An Historical
Approach’. The symposium offered what McGregor and Purdey (2016) call ‘a rare
and remarkable expression of political will to deal with the human rights abuses
of 1965’. Survivors were given an unprecedented public forum to describe the
anti-communist pogrom of 1965–66 and the decades of discrimination they and
their families suffered under Soeharto’s authoritarian regime.
Activists and scholars, however, were restrained in their praise of the sympo-
sium (McGregor and Purdey 2016). While the government had provided victims
with this public forum, it simultaneously shut down any possibility for achieving
justice. Luhut, then coordinating minister for political, legal, and security affairs,
spoke on behalf of the government when he refused to acknowledge the scale of
the massacre, and was adamant there would never be an apology to victims. In
fact, the government’s goal appeared to be a speedy and tidy end to the contro-
versy surrounding 1965–66. Luhut even stated publicly that the president wanted
the issue to be ‘finished’ by May 2016, with no mention of justice or reparations
for victims of the massacres (McGregor and Purdey 2016).
Jokowi’s alliances with military generals shape his increasingly conservative
position on this issue. Military figures feature prominently in Jokowi’s cabinet
and inner circle, and many of them have ultra-conservative credentials and poor
human-rights records. For example, Jokowi appointed Ryamizard Ryacudu, on
Megawati’s request, as minister for defence. Ryamizard is known as an ‘arch-
conservative’ (IPAC 2016, 3) and in early 2016 stated publicly that communists
‘deserved to die’ in 1965–66 (Asia Times, 2 June 2016). In July, Jokowi appointed
Wiranto as coordinating minister for political, legal, and security affairs; Wiranto
is a former general accused of crimes against humanity who has no inter-
est in pursuing historical justice for victims of state violence. Hendropriyono,
another New Order general, played a prominent role in Jokowi’s campaign
and has remained an informal advisor to the president over the last two years.
Hendropriyono was director of the National Intelligence Agency (BIN) during
the Megawati presidency, and was allegedly behind the murder of prominent
human-rights activist Munir in 2004 (Sydney Morning Herald, 18 Dec. 2010).
Jokowi’s new commander of the armed forces, General Gatot Nurmantyo, is
another ultra-conservative figure; he not only opposes any initiatives for his-
torical justice but also supports expanding the role of the armed forces in non-
security-related affairs (IPAC 2016).
Jokowi does not appear to be acting out a premeditated plan to resurrect the
military’s role in politics, nor is he deliberately injecting conservative militaris-
tic ideas into policy and politics. Rather, these appointments are made in isola-
tion and on the basis of discrete political calculations. But the result is that some
of the most senior and conservative military generals from the late New Order
period have now been returned to positions in which they wield significant politi-
cal influence. The effect is to block progressive initiatives that address histori-
cal injustice and to undermine reform of Indonesia’s security apparatus, where,
according to Aspinall, Mietzner, and Tomsa (2015, 17), ‘reservoirs of authoritarian
thinking’ have remained since the end of the New Order.
Jokowi and the New Developmentalism 315

At the same time, there is little political incentive for Jokowi to resist the demands
of conservative elites when it comes to sensitive problems of human rights and
historical justice. The electorate itself is increasingly conservative. During the late
Yudhoyono years, observers were widely discussing a rise in illiberal attitudes
and practices among both political and societal groups (Berger 2015; Bush 2015).
Yudhoyono failed to take a strong stance against Islamic radicalism, for example,
or prejudice towards religious and cultural minorities, because he feared alienat-
ing mainstream and conservative Islamic groups. Polls have shown for years that
Indonesians are increasingly intolerant towards people of other religions and of
minorities (Jakarta Post, 24 Feb. 2016); they demonstrate a sustained fear of com-
munism, and there is widespread admiration for the military (CSIS 2016).
Nor does the president himself demonstrate a strong personal interest in, or
a principled commitment to, universal human rights. As we have seen, in 2016
Jokowi paid attention to what counted politically, and in the process he was
responding to the conservative political environment in which he operated.

ECHOES OF THE PAST

In summary, Jokowi’s new developmentalism is pragmatic and growth-oriented
in its policies, statist–nationalist in its ideological positioning, and conservative
in its approach to problems of transparency and governance and of human rights
and justice. There are uncanny echoes of the past in this new developmentalism.
It recalls the New Order’s emphasis on pragmatic developmental programming
and its rhetorical aspiration to ‘modernise’ Indonesia. Jokowi’s personal style has
even prompted comparisons with President Soeharto—some admiring and oth-
ers critical.
Under the New Order, the state’s mission was to achieve ‘modernity’, a task
that, as Cribb (1999, 34) explains, ‘was defined . . . in a practical sense: stabilis-
ing prices, repairing physical infrastructure, making agriculture more productive,
encouraging industry’. In the late 1980s and early 1990s, Soeharto promoted a pol-
icy focus remarkably similar to Jokowi’s contemporary agenda. In his state-of-the-
nation address in 1990, for example, Soeharto explained that in order for Indonesia
to attain a more equitable distribution of wealth, the government must focus on
‘de-regulation and de-bureaucratisation’—language that anticipates that used by
Jokowi. Soeharto argued that to accelerate development the state must create more
opportunities for private investment and awaken ‘the initiative, creativity and par-
ticipation of the public in development’—though he was at pains, like Jokowi, to
emphasise that to deregulate was to modernise, not to liberalise. In much the same
way, Jokowi, in his 2016 address, implored Indonesians to be creative and com-
petitive and to contribute to the modernisation of Indonesia’s economy.
The parallels are not lost on Indonesia’s political commentators. Some in the
media even refer directly to Jokowi as a ‘little Soeharto’ (Jakarta Post, 28 July
2016). Known as the ‘Father of Development’, the New Order autocrat governed
Indonesia for 32 years and oversaw a long period of economic transformation
(Hill 1994). Soeharto also prioritised infrastructure—and roads, in particular—
such that Indonesia’s physical infrastructure in the 1990s was considered supe-
rior to that of countries at similar stages of development (Hill and Narjoko 2010;
Davidson 2015).
316 Eve Warburton

Comparisons go beyond Soeharto’s and Jokowi’s shared developmentalist

agenda. For some observers, Jokowi’s personality and leadership style are remi-
niscent of Soeharto (Jakarta Globe, 9 Dec. 2014). One senior editor wrote that Jokowi
maintains a passive outward appearance, ‘but when he thinks the time is right,
like Soeharto, Jokowi has the ruthlessness to exercise his power’ (Jakarta Post, 8
May 2016). Analysts and government insiders frequently refer to the two presi-
dents’ common Javanese heritage, drawing attention to how both presidents are
quietly spoken, low-profile developmentalists, with a deceptively cunning knack
for political manipulation.
There are, of course, fundamental differences between the two presidents. Most
obviously, Soeharto came to power as an unelected military general who relied
on repression to gain and remain in office; Jokowi is a democrat who rose to office
through Indonesia’s decentralised democratic system, and he boasts a strong
popular mandate. However, the reaction from commentators is telling. By acci-
dent or design, Jokowi has tapped into what Aspinall (2015, 3) has described as a
‘widespread—though diffuse—mood of nostalgia for the certainties of the New
Order’, and for its golden years of economic achievement.
Indeed, the polls demonstrated that Indonesians embraced Jokowi’s pivot
towards a simple and pragmatic developmental agenda. The electorate did not
punish the president for his retreat from reform or for his conservative approach
to rights and justice. After October 2015, public approval ratings improved stead-
ily for Jokowi, with an Indikator survey conducted after the July reshuffle show-
ing public approval at 68%—his highest rating since coming to office (Indikator
2016). When respondents were asked if they had faith in Jokowi’s capacity to lead,
the numbers increased to over 74%. So if the president can continue to maintain
political stability and realise some of his infrastructure goals, he may be rewarded
at the ballot box in 2019.
Although the president received a boost over the course of 2016, public approval
should be viewed as tenuous. The president’s popularity is highly contingent on
inflation (figure 1) and, to a lesser extent, oil prices; the budget is under serious
pressure; and the president’s core policy projects—the deregulation packages, the
35,000 megawatt project—all face great challenges (Busch 2016a; Straits Times, 25
Aug. 2016). The president’s personality and his leadership deficiencies could also
cause setbacks on the road to 2019. As we have seen, Jokowi is prone to making
rushed, often rash decisions, and without wide consultation. This combination
makes it difficult to predict how the president will manage an increasingly complex
set of budget challenges, and how he will realise his ambitious infrastructure goals
when they are wracked by logistical problems, corruption, and political infighting.

CONCLUSION
In 2016, President Jokowi made an extraordinary comeback. He achieved political
stability, reshaped the political map, and laid out his developmentalist agenda for
Indonesia. In the process, we gained fresh insight into Jokowi’s personality and
leadership style. Prior to becoming president, Jokowi had been depicted by most
analysts as a shining example of Indonesia’s democratic consolidation and post-
Soeharto reforms (McRae 2013). After he came to office, commentary on Jokowi’s
political persona began to shift, as analysts attempted to explain the president’s
Jokowi and the New Developmentalism 317

FIGURE 1  Jokowi’s Public Approval Rating (%) and Inflation (% year on year)
80 8

70 Approval (lhs) 7

60 6

50 5

40 4

30 Inflation (rhs) 3

20 2

10 1

0 0
Dec-2014 Apr-2015 Aug-2015 Dec-2015 Apr-2016 Aug-2016
Sources: Data from Indikator (2016) and Bank Indonesia.

retreat from reform and his embrace of realpolitik. By the end of his second year
in office, many observers were looking back to the New Order to understand the
president and the decisions he had made.
One emerging line of argument is that Jokowi should be seen as the product
of a generation ‘born, bred and educated’ under the New Order (Jakarta Globe,
31 Oct. 2015). From this perspective, Jokowi’s developmentalism and his will-
ingness to compromise on liberal reform are both functions of his New Order
rearing. Baker (2016) suggests that Jokowi personifies the New Order’s illiberal
middle class, and that his developmental agenda ‘goes beyond a simplistic per-
sonal attribute or set of beliefs: it is inherent to his class status’. As a middle-class
businessman rising to success in the late New Order years, Jokowi, Baker argues,
built alliances with the New Order establishment and learned an appreciation for
‘services and standards’—but developed little interest in liberal democratic ideas
or political rights.
These structuralist accounts of Jokowi’s political identity contrast somewhat
with earlier characterisations of Jokowi as a ‘technocratic populist’ informed by
his experiences as a businessman and driven into politics by a desire to improve
the lives of ordinary Indonesians (Mietzner 2015). Jokowi’s developmental-
ist sensibilities have undoubtedly been shaped by the socio-political realities
of Soeharto’s Indonesia and his interactions with the New Order bureaucratic-
military regime. Yet the notion that Jokowi is simply the product of a repressive
developmental state glosses over his idiosyncrasies and the way in which he has
adapted and evolved to suit Indonesia’s contemporary political landscape.
The story that emerges from the past two years is of a president who once
saw political value in a progressive agenda for clean government and cultivated
the identity of a political maverick. Jokowi flirted with institutional reform at a
time when, off the back of a unique campaign, it seemed plausible that he could
achieve success as a reformer. But as the political circumstances changed, so did
318 Eve Warburton

the president. Jokowi has learned that reform is complex and full of political risk,
and he is not willing to jeopardise the stability he has achieved or the alliances he
has built over his first two years in office. Jokowi is not consistently an economic
nationalist, a conservative, or an anti-reformist; nor does he represent a coherently
illiberal middle class. Instead, he is simply charting the course of least resistance.
Since coming to office, Jokowi has been responding to and moulding himself to
the political economy that he inherited. The president has a very simple agenda:
to improve Indonesia’s infrastructure and provide public services. Beyond this,
Jokowi is changeable—even unpredictable—and something of a blank slate.8
Jokowi is also defined by ad hocery; he often takes a reactive approach, respond-
ing to problems of government as they arise and through whatever method seems
the most practical and convenient in the circumstances.
For these reasons, we must see the new developmentalism as reflecting the
deeper ideological and structural imperatives of post-Soeharto Indonesia. In 2016,
Jokowi began mastering Indonesia’s patronage-driven democracy, embracing the
tailwinds of an economic nationalism that pre-dates his presidency, and acting
out the conservative sensibilities of Indonesia’s political class. Jokowi offers a fast,
simple, and ruthless implementation of the statist–nationalist development strat-
egy that has long loomed large in Indonesia.

REFERENCES
Amnesty International. 2015. Flawed Justice: Unfair Trials and the Death Penalty in Indonesia.
Report ASA 21/2434/2015. London: Amnesty International.
Aspinall, Edward. 2015. ‘Oligarchic Populism: Prabowo Subianto’s Challenge to Indonesian
Democracy’. Indonesia 99 (April): 1–28.
Aspinall, Edward, Marcus Mietzner, and Dirk Tomsa. 2015. The Yudhoyono Presidency:
Indonesia’s Decade of Stability and Stagnation. Singapore: ISEAS Publishing.
Baker, Jacqui. 2016. ‘The Middle Class President’. New Mandala, 5 August. http://www.
newmandala.org/comfortable-uncomfortable-accommodations/.
Berger, Dominic. 2015. ‘Human Rights and Yudhoyono’s Test of History’. In The Yudhoyono
Presidency: Indonesia’s Decade of Stability and Stagnation, edited by Edward Aspinall,
Marcus Mietzner, and Dirk Tomsa, 217–38. Singapore: ISEAS Publishing.
Budiartie, Gustidha, and Eve Warburton. 2015. ‘Indonesia’s Freeport Saga’. New Mandala,
22 December. http://www.newmandala.org/indonesias-freeport-saga/.
Busch, Matthew. 2016a. ‘Jokowi’s Cabinet Reshuffle: Implications for the Economy’.
Indonesia at Melbourne, 8 August. http://indonesiaatmelbourne.unimelb.edu.au/
jokowis-cabinet-reshuffle-implications-for-the-economy/.
———. 2016b. ‘Jokowi’s Natural Resource Gamble’. East Asia Forum, 15 April. http://
www.eastasiaforum.org/2016/04/15/jokowis-natural-resource-gamble/.
Bush, Robin. 2015. ‘Religious Politics and Minority Rights during the Yudhoyono
Presidency’. In The Yudhoyono Presidency: Indonesia’s Decade of Stability and Stagnation,
edited by Edward Aspinall, Marcus Mietzner, and Dirk Tomsa, 239–57. Singapore:
ISEAS Publishing.
Chalmers, Ian. 1997. ‘Introduction’. In The Politics of Economic Development in Indonesia:
Contending Perspectives, edited by Ian Chalmers and Vedi R. Hadiz, 1–35. London:
Routledge.

8. I wish to thank Edward Aspinall for this particular characterisation of Jokowi as a blank
slate.
Jokowi and the New Developmentalism 319

Cribb, Robert. 1999. ‘Nation: Making Indonesia’. In Indonesia Beyond Suharto: Polity,
Economy, Society, Transition, edited by Donald K. Emmerson, 3–38. Armonk, NY: M. E.
Sharpe.
CSIS (Centre for Strategic and International Studies). 2016. ‘2 tahun Jokowi: Optimisme
publik, konsolidasi kekuasaan dan dinamika elektoral; Rilis hasil survei nasional CSIS’
[Two years of Jokowi: Public optimism, consolidation of power and electoral dynamics;
Results of the CSIS national survey]. Survey findings, CSIS, Jakarta.
Davidson, Jamie S. 2015. Indonesia’s Changing Political Economy: Governing the Roads.
Cambridge, UK: Cambridge University Press.
———. 2016. ‘Eminent Domain and Infrastructure under the Yudhoyono and Widodo
Administrations’. In Land and Development in Indonesia: Searching for the People’s
Sovereignty, edited by John F. McCarthy and Kathryn Robinson, 167–85. Singapore:
ISEAS Publishing.
Fealy, Greg. 2011. ‘Indonesian Politics in 2011: Democratic Regression and Yudhoyono’s
Regal Incumbency’. Bulletin of Indonesian Economic Studies 47 (3): 333–53.
Feith, Herb. 1981. ‘Repressive-Developmentalist Regimes in Asia’. Alternatives: Global,
Local, Political 7 (4): 491–506.
Gammon, Liam. 2015. ‘Jokowi’s Year of Living Cautiously’. East Asia Forum, 12 December.
http://www.eastasiaforum.org/2015/12/12/jokowis-year-of-living-cautiously/.
Hamilton-Hart, Natasha, and Günther G. Schulze. 2016. ‘Taxing Times in Indonesia: The
Challenge of Restoring Competitiveness and the Search for Fiscal Space’. Bulletin of
Indonesian Economic Studies 52 (3): 265–95.
Hearman, Vannessa. 2016. ‘No “Magic Bullet”’. Inside Indonesia 123 (January–March).
http://www.insideindonesia.org/no-magic-bullet.
Hidayat, Nurkholis. 2015. ‘Jokowi’s First Year: Human Rights’. Indonesia at Melbourne,
19 October. http://indonesiaatmelbourne.unimelb.edu.au/jokowis-first-year-human-
rights/.
Hill, Hal. 1994. ‘The Economy’. In Indonesia’s New Order: The Dynamics of Socio-Economic
Transformation, edited by Hal Hill, 54–122. Sydney: Allen & Unwin.
Hill, Hal, and Dionisius A. Narjoko. 2010. ‘Managing Industrialisation in a Globalising
Economy: Lessons from the Soeharto Era’. In Soeharto’s New Order and Its Legacy: Essays
in Honour of Harold Crouch, edited by Edward Aspinall and Greg Fealy, 49–66. Canberra:
ANU E Press.
Human Rights Watch. 2016. ‘“These Political Games Ruin Our Lives”: Indonesia’s
LGBT Community Under Threat’. https://www.hrw.org/report/2016/08/10/
these-political-games-ruin-our-lives/indonesias-lgbt-community-under-threat.
Indikator. 2016. ‘Public Assessment of Jokowi’s Government Post-Second Reshuffle: Results
from a National Survey; 1–9 August 2016’. http://www.indikator.co.id/agenda/
details/31/National-Survey-Release-Public-Assessment-of-Jokowis-Government-Post-
Second-Reshuffle.
IPAC (Institute for Policy Analysis of Conflict). 2016. Update on the Indonesian Military’s
Influence. Jakarta: IPAC.
Leftwich, Adrian. 1995. ‘Bringing Politics Back In: Towards a Model of the Developmental
State’. Journal of Development Studies 31 (3): 400–427.
Lubis, Uni. 2015. ‘Luhut, Golkar dan Jokowi di pusaran #papamintasaham Freeport’ [Luhut,
Golkar and Jokowi in Freeport #papamintasaham maelstrom]. Rappler, 16 December.
http://www.rappler.com/indonesia/116079-luhut,-golkar-dan-jokowi-di-pusaran-
papamintasaham-freeport.
McGregor, Katharine, and Jemma Purdey. 2016. ‘Indonesia Takes a Small but Critical Step
toward Reconciliation’. Indonesia at Melbourne, 26 April. http://indonesiaatmelbourne.
unimelb.edu.au/indonesia-takes-a-small-but-critical-step-toward-reconciliation/.
McRae, Dave. 2013. ‘Indonesian Politics in 2013: The Emergence of New Leadership?’.
Bulletin of Indonesian Economic Studies 49 (3): 289–304.
320 Eve Warburton

Mietzner, Marcus. 2015. Reinventing Asian Populism: Jokowi’s Rise, Democracy, and Political
Contestation in Indonesia. Policy Studies 72. Honolulu, HI: East-West Center.
———. 2016. ‘Coercing Loyalty: Coalitional Presidentialism and Party Politics in Jokowi’s
Indonesia’. Contemporary Southeast Asia 38 (2): 209.
Muhtadi, Burhanuddin. 2015. ‘Jokowi’s First Year: A Weak President Caught between
Reform and Oligarchic Politics’. Bulletin of Indonesian Economic Studies 51 (3): 349–68.
Patunru, Arianto, and Sjamsu Rahardja. 2015. ‘Trade Protectionism in Indonesia: Bad
Times and Bad Policy’. Lowy Institute Analysis, Lowy Institute, Sydney.
Power, Thomas. 2016. ‘Cashing in’. New Mandala, 8 August. http://www.newmandala.
org/cashing-in/.
Ray, David, and Lili Yan Ing. 2016. ‘Addressing Indonesia’s Infrastructure Deficit’. Bulletin
of Indonesian Economic Studies 52 (1): 1–25.
Setiawan, Ken. 2016. ‘“Old Guard” Blocking Human Rights Reform in Indonesia’.
East Asia Forum, 4 February. http://www.eastasiaforum.org/2016/02/04/
old-guard-blocking-human-rights-reform-in-indonesia/.
Slater, Dan, and Erica Simmons. 2012. ‘Coping by Colluding: Political Uncertainty and
Promiscuous Powersharing in Indonesia and Bolivia’. Comparative Political Studies 46
(11): 1366–93.
Warburton, Eve. 2015. ‘Indonesia: Why Economic Nationalism Is So Popular’. Lowy
Interpreter, 25 August. http://www.lowyinterpreter.org/post/2015/08/25/Indonesia-
Why-economic-nationalism-is-so-popular.aspx.
———. Forthcoming. ‘Resource Nationalism in Indonesia: Ownership Structures and
Sectoral Variation in Mining and Palm Oil’. Journal of East Asia Studies.
Woo-Cumings, Meredith. 1999. The Developmental State. Ithaca, NY: Cornell University
Press.
World Bank. 2016. Indonesia Economic Quarterly: Resilience through Reforms. June.
Washington, DC: World Bank.
 Globalizations

ISSN: 1474-7731 (Print) 1474-774X (Online) Journal homepage: https://www.tandfonline.com/loi/rglo20

Neoliberalism and altered state developmentalism

in the twenty-first century extractive regime of
Indonesia

Paul K. Gellert

To cite this article: Paul K. Gellert (2019): Neoliberalism and altered state developmentalism
in the twenty-first century extractive regime of Indonesia, Globalizations, DOI:
10.1080/14747731.2018.1560189

To link to this article: https://doi.org/10.1080/14747731.2018.1560189

Published online: 24 Jan 2019.

Submit your article to this journal

Article views: 36

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=rglo20
GLOBALIZATIONS
https://doi.org/10.1080/14747731.2018.1560189

Neoliberalism and altered state developmentalism in the twenty-first

century extractive regime of Indonesia
a,b
Paul K. Gellert
a
KITLV, Leiden, the Netherlands; bUniversity of Tennessee, Knoxville, TN, USA

ABSTRACT KEYWORDS
This article examines neoliberal leverage, states and ‘deep marketization’ in Neoliberalism; deep
relation to resource extraction in Indonesia. The concept of ‘altered’ state marketization; resource
developmentalism within a world-historical analysis of the semi-peripheral extraction; resource curse;
developmental state;
zone of the world-economy is introduced to address how much and in what
Indonesia
ways resource extraction persists and even continues to expands. While most
theories of development assume either that natural resource extraction will
gradually fade or that the ‘curse’ of natural resources requires institutional
reform, Indonesia’s ‘extractive regime’ persistently and under different political
systems relies on extraction of land, forest, and mineral resources. Altered state
developmentalism occurs, from this perspective, within a world-historical cycle
of resource nationalism (CRN) as US hegemony declines. In contrast to the A
first CRN’s put more emphasis on nationalizing ownership, e for progressive
redistribution aims and industrial transformation. In contrast to the first CRN’s
emphasis on nationalizing ownership, empirical evidence from the oil, mining,
and oil palm sectors illustrate how the advance of neoliberalism is tempered in
the second CRN by altered developmentalism that aspires for growth and
poverty reduction in a more market-oriented direction pushed by neoliberal
leverage. While accommodating some pressures for land rights against the
expansion of extractive frontiers, the overall thrust expands the conditions for
yet more deep marketization.

Neoliberal leverage, states and marketization in relation to resource extraction

Much of the literature addressing neoliberalization in the political economy does so writ large with-
out attention to place, politics, or empirical histories. In this article, it is grounded in the persistence
of Indonesia’s extractive regime (Gellert, 2010). Broadly, neoliberalism needs to be understood as a
political project of (re)distribution of resources brought about through the leverage of powerful
actors and institutions on weaker ones, supported by a legitimizing ideology. Neoliberalism con-
gealed around deregulation and privatization, and has spread unevenly since the 1970s due to the
politics of implementation. Distinct from the raw power of IMF structural adjustment programmes,
leverage refers to the ability to get weaker actors and institutions to accept and internalize neoliberal
precepts.1 In the latest phase of neoliberalism, then, ‘deep marketization’ works not only ‘on’
downsizing or dismantling the developmentalist state (the Washington Consensus) and ‘through’
reshaping it by implementation of ‘good governance’ techniques (the Post-Washington Consensus)
but also ‘around’ the state by facilitating the mobile flexibility of financial capital (Carroll, 2017b;
Neusiedl, 2017).

CONTACT Paul K. Gellert pgellert@utk.edu

© 2019 Informa UK Limited, trading as Taylor & Francis Group
2 P. K. GELLERT

The main themes of the articles in this special issue focus on the transition through various phases
of neoliberalism from the Washington consensus (Williamson, 1990) through a period of more
socially oriented policies under a post-Washington consensus (PWC) (Birdsall & Fukuyama,
2011) and of late to an emphasis on ‘deep marketization’ (Carroll, 2012; Carroll & Jarvis, 2014a).
These versions have been dubbed neoliberalism 1.0, 2.0 and 3.0, respectively (Gonzalez-Vicente, Car-
roll, & Jarvis, 2017). The idea of deep marketization of version 3.0 is accompanied by the empirical
observation that neoliberalism is, in effect, returning to and being implemented in the geographical
locales of its intellectual origins in the global North, albeit with considerable contestation including
right-wing populism in the United States, Great Britain and elsewhere. Although neoliberalism is
variegated (Brenner, Peck, & Theodore, 2010), it is important to recognize that the uneven spread
of neoliberalism is not simply due to the degree of traction of ideas of neoliberalism among elites
and policymakers (Mukand & Rodrick, 2016; Rodrick, 2014). Rather, it is the outcome of political
struggles and material and ideological conflicts over the global spread of capitalism. These conflicts
are salient in and shaped by the commodities exported from natural resource semi-peripheries like
Indonesia, the focus of this article.
Materially, the advance of neoliberalism is supported by the expansion of natural resource extrac-
tion. This article, therefore, attempts to analyze the deep marketization processes and practices in
relation to resource extraction in the global South (as well as the global North). It asks whether
and how the diffusion of neoliberalism is exemplified in states that rely on natural resource extrac-
tion and export for growth and ‘development’. Specifically, the article focuses on the case of Indo-
nesia to illustrate how deeply marketized neoliberalism is working to legitimize a mode of semi-
peripheral development that continues to rely on resource extraction, as well as intertwined land
acquisition necessary for the expansion of oil palm plantations and coal mining.2 The marketization
thesis, even in its more politically contested and historical rendering, needs to be nuanced to account
for the predominance of natural resource exporters like Indonesia.
Although deep marketization has been spreading in the last decade, natural resources are in some
ways the least likely productive resources to be fully marketized. Private capital usually relies on ter-
ritorial states for access to and control over resources. Chile, for example, was the ‘poster child’ of
neoliberal development based on natural resource exports, but expansion relied on considerable
state action (Kurtz, 2001). There is a long history of capitalist expansion in which the (capitalist)
state has played a significant role in constructing and controlling property relations from the Enclo-
sures to the creation of ‘political forests’ in the colonial states of Southeast Asia (Peluso & Vander-
geest, 2001) which in turn served to legitimize the allocation of rights to extract, to cultivate, and to
export and sell. Political forests have been firmly established through legal processes and an ideologi-
cal construction of ‘rule of law’ as universal rather than biased (Gellert & Andiko, 2015). These pol-
itical developments are important to the expansion of the world-system as a set of ‘world-ecological’
processes, combining exploitation and appropriation (Mies, 1986; Moore, 2014, 2015). Expansion
relies, in other words, on both the exploitation of labour to produce value in a system of (ecologi-
cally) unequal exchange and the appropriation of human and nonhuman value, e.g. via household
labour (Dunaway, 2014), states (Parenti, 2015), and nature (Bunker, 1985; see also Frey, Gellert, &
Dahms, 2019).
Notably, in some cases, resource extraction has led to the emergence and solidification of ‘extrac-
tive regimes’ (Gellert, 2010; Nem Singh & Bourgouin, 2013). An extractive regime is ‘a sociopolitical
formation that relies for its power and longevity on extraction of natural resource wealth as com-
modities and on the importance of these commodities to the world-system’s core’ (Gellert, 2010,
p. 32). The extraction of resources that provides the raw materials for core consumption and, in
 GLOBALIZATIONS 3

rare cases, hegemonic ascent in the world-system comes from peripheral and semi-peripheral areas
(Bunker & Ciccantell, 2005). Extractive regimes are a sub-set that tend to be found in semi-peripheral
countries (or situations). In these cases, two conditions hold: first, extraction of multiple natural
resources forms a significant basis for the production and accumulation of value, and second, there
is a resilient economic and political regime of growth and accumulation that relies on political dom-
ination and legitimation. The latter condition of some growth and accumulation distinguishes extrac-
tive regimes of the semi-periphery from extraction in the periphery of the world-system that is more
often of an enclave character. These conditions and other characteristics also distinguish extractive
regimes from the better known (but contested) idea of a ‘resource curse’, as explained below.
In twenty-first century Indonesia, this paper argues, a form of ‘altered’ state developmentalism is
evident. Altered state developmentalism is a variant of earlier cases of state developmentalism (Car-
roll & Jarvis, 2017; Woo-Cumings, 1999). It retains the professed goals of industrial transformation
and GDP growth but relies less on state or dirigiste strategies than market-based strategies to do so.
As such, the developmental state is not ‘dead’ (Pirie, 2017). As a legitimating ideology it continues to
be powerful, but the state that is legitimated by ‘development’ is altered in the means to achieve it.
The result is a mix of neoliberal market strategies and policies and top-down statist, even nationalist,
ones.
The article proceeds in several parts. In the first, the expectation that resource exports would be a
temporary stage in a range of theoretical traditions, including the relationship between the evolution
of neoliberalism and the persistence of resource exports, is briefly reviewed. Second, I define and the-
orize the extractive regime by way of explaining how it differs from two other perspectives, the
resource curse and new institutionalism. Neither of these sets of literature adequately addresses
the historical and geographical political economy of resources in the world-system. Resource extrac-
tion’s material and ideological characteristics have changed as early to mid-twentieth century
resource nationalism gave way, first to aspects of developmental statism under the New Order
and in the last 15 years to increased but far from complete marketization that I call altered state
developmentalism. In the third section, I situate ‘altered’ state developmentalism within a world-his-
torical analysis of the semi-peripheral zone of the world-economy to address how much and in what
ways resource extraction persists and even continues to expand in Indonesia. I use this world-sys-
tems framing to define the extractive regime and to explain the oscillation of the extractive regime’s
politics in Indonesia across two cycles of resource nationalism (CRNs) (Kaup & Gellert, 2017) that
include the longevity of the Suharto authoritarian period and the deepening marketization of the
reform and democratization years since. In the fourth section, I explore more carefully the recent
advances of neoliberalization and deep marketization with examples from the oil, mining, and oil
palm sectors. To be clear, marketization is advancing but is still shaped by a state that claims prop-
erty rights or control of access to two-thirds of the land in the archipelago designated as ‘forest area’
(kawasan hutan) (Gellert & Andiko, 2015) and that both opens to foreign investment and privatiza-
tion of state-owned enterprises (SOEs) while also pushing for ownership stakes by various levels of
government from national to regional. In other words, the leverage and power of deep marketization
is tempered by the altered developmentalism of political control over (nominally) forest lands, com-
modification of resources and the investment and trade of these commodities.

Resource persistence: counter to expectations of ‘development’

The persistence of resource extraction and exports evident in extractive regimes such as Indonesia
runs counter to the assumptions of most theories of development, which posit that natural resource
4 P. K. GELLERT

exports are a temporary stage in transition to better and more Western development with higher
incomes per capita and growing levels of consumption. In modernization theory, for example,
which contained this Eurocentric and normative approach to development, as well as in some
forms of classical Marxism, there was an assumption of stages (Peet & Hartwick, 1999). The ‘back-
wards’ (i.e. non-capitalist) agricultural and natural resources sector were to be superseded in a sub-
sequent stage by industrialization and the growth of manufacturing jobs in the urban centres.
Development economists only debated whether growth should be balanced in some kind of big
‘push’ for high growth rates and broad-based industrialization or unbalanced. The latter group
pinned optimistic hopes on the creation of development via growth poles and linkages, including
fiscal linkages from natural resource extraction (Hirschman, 1977).
Partly in response, state and institutional theorists (discussed more in the following section) pos-
ited that pathways from the periphery could be built on the foundation of effective, developmental
states with embedded autonomy (Evans, 1995; Haggard, 1990). In particular, industrial transform-
ation was to be built by constructing comparative advantage such as Japan and South Korea, as well
as China. However, these cases of developmental state success in rapid economic transformation and
advance in the world hierarchy of nation states have been rare, historically produced – and rejected
by the advocates of neoliberalism. More commonly, efforts to use linkages as the basis for regional
and national economic development and to build have been partial or complete failures in cases from
Brazil to Indonesia (Bunker, 1985, 1989).
The relationship between neoliberalism and the persistence of resource extraction is less obvious.
Neoliberal advice to follow ‘comparative advantage’ supports continued reliance on natural
resources. However, neoliberal economists also harbour modernizationist assumptions, including
the idea that with sufficient private capital investment and growth, comparative advantage will
not be static, and economies will likely move beyond resource exports into other sectors. To do
so, neoliberalism posits that private investors will build industries and this transition will (and
should) occur without a state dirigiste role. Not surprisingly, therefore, accusations of corruption
and cronyism have been used as a cudgel against states in order to bring about ‘good governance’,
as happened in Indonesia in the wake of the Asian Financial Crisis (Carroll, 2010). In the global
North, we have seen financialization of land, even forest lands on which vertically integrated indus-
tries used to rely (Gunnoe, 2014; Gunnoe & Gellert, 2011). Because the marketization of ‘nature’
requires the state to appropriate and exploit human and nonhuman nature, semi-peripheral states
are locations of accentuated state involvement in the economy, as the next section explores.

Theorizing extractive regimes of the semi-periphery: challenging resource curse

and new institutionalist approaches
Unlike the dominant assumption of resource exports as a passing phase, two streams of literature do
address the persistence of resource exports from the periphery: resource curse and instiutionalism In
this section I make the case for a more historical and geographically sensitive concept of ‘extractive
regime’ against existing but limited literatures on the resource curse and (new) institutionalism. The
limits of these literatures lead to an explication of the semi-periphery where extractive regimes most
often reside. As introduced above, extractive regimes (Gellert, 2010) rely on natural resources for
capital accumulation and government revenue; they have proven to be resilient sociopolitical for-
mations in which resources are deemed necessary for growth – ideologically justifying negative
impacts on the people and (nonhuman) ecologies affected by their extraction; and they are most
 GLOBALIZATIONS 5

often situated in semi-peripheral locations in the world-system with exploitation of resources in

extractive peripheries either inside or outside the state’s territory.

Resource curse revisited

The resource curse argument is well-known. As its foundational authors explained, counter to liberal
and neoliberal intuition, national endowments of natural resources are associated with negative out-
comes (Auty, 1993; Sachs & Warner, 2001). High levels of resource exports are associated with harm
to national development through lower rates of economic growth, heightened corruption (of the
easily available resource rents), and authoritarianism (even if authoritarianism is also, ironically,
part of the developmental state institutional apparatus). In addition, currency appreciation from
resource revenue leads to sectoral distortion known as the ‘Dutch disease’. All of this, resource
curse theorists argue, requires attention from state policymakers to be effectively managed. Some lit-
erature on Indonesian development takes up the voluntarist approach, suggesting that the resource
curse was managed well by technocrats who made wise investments in agriculture and avoided the
‘Dutch disease’ misallocation productive resources (Ascher, 1998; Booth, 1995; Bresnan, 1993), with-
out accounting for the domestic politics and multiple resource exports that made this path desirable
and possible. One important exception is Ross (2001) who introduced ‘rent seizing’ or efforts by
states to allocate rents in relation to timber booms in Southeast Asia.
The most straightforward distinction between resource curse arguments and the extractive regime
is that the former focuses on single commodity dependence (especially petroleum) while extractive
regimes exploit more than one commodity. In his careful empirical analysis of the resource curse,
Ross moved away from the forest resource curse and concluded that the resource curse should be
renamed an oil or mineral curse. He explains, ‘It is more accurately a mineral curse, since these mala-
dies are not caused by other kinds of natural resources, like forests, fresh water, or fertile cropland.
… The resource curse is overwhelmingly an oil curse’ (Ross, 2013, p. 1). Another difference is that
curse scholars find wholly negative outcomes (for growth, democracy, civil war) while extractive
regimes are more mixed, with some sharing of wealth accumulation especially by state and dominant
class participants in extraction, and persistence under democratic rule. A third is that curse studies
associate resources with nondemocratic political systems while extractive regimes have flexible poli-
tics. Resource extraction persists and co-exists with multiple political regimes, from authoritarian to
more liberal democratic.
The resource curse literature coalesces around a need for states with the capacity to intervene in
the economy. Ross’s earlier work on rent seizing noted a downward spiral in which rent seeking
(further) weakens state institutions. Even critics of the resource curse highlight the importance of
intervening factors such as state institutions and their willingness to invest in human capital as deci-
sive (Kurtz & Brooks, 2011). The emphasis on building institutions provides a bridge to the second
literature to be distinguished from extractive regimes.

States and institutions

Interestingly, the resource curse literature and new institutionalism, including statist strands of pol-
itical economy, come together in identifying state capacity as crucial. The former criticizes states for
being authoritarian, violent, and ultimately so weak or lacking in capacity in their focus on the
resource – usually, oil – that economic growth suffers, let alone diversification or ‘development’.
The latter, institutionalist critics who see divergent outcomes take a state-centered view in which
6 P. K. GELLERT

“good” political institutions, that is, those that promote accountability and state competence, are shown
to create incentives for politicians to utilize resource wealth in growth-enhancing ways (Dunning, 2005;
Robinson, Torvik, & Verdier, 2006) whereas “bad” institutions serve to reinforce political and economic
underdevelopment. (Kurtz & Brooks, 2011, p. 752)

Both positions reveal a bias against the perceived venality of elites and posit the need for ‘clean’ refor-
mers to fix corruption and build strong state institutions. Also, they share the hope – albeit weaker
among adherents to the resource curse thesis – that states and their institutions can fix things. Both
take an individualist and ideational approach to politics, exemplified in the varying evaluations of the
strength of Indonesian President Joko Widodo in taking a serious stance against corruption (War-
burton, 2016).
This position is precisely what early statists like Peter Evans (1995, 1989) argued. Developing
economies are said to need a state with effective bureaucratic capacity, internal solidarity to reinforce
the formal bureaucratic structures and an organized network of external ties that simultaneously
embeds the state in social relations. Given these institutional capacities, states would achieve indus-
trial transformation, growth and development. Thus, institutionalists spend much of their effort try-
ing to dissect whether and why particular states do or do not cultivate these institutions and the non-
corrupt esprit de corps for their effective pursuit of policymaking. While Korea, Japan and China have
been pointed to as successful developmental states, a number of critics have noted both the rarity of
developmental states emerging – per Cumings’ (1984) memorable phrase that ‘many are called, few
are chosen’ – and the literature’s emphasis on states and capitalists to the omission of labour,
let alone environment (Deyo, 1989). It is debatable whether Indonesia ever fit the classic definition
(Davidson, 2017) and its achievements of high growth rates and declining poverty must be placed in
the context of the Cold War, but as explained below, the Indonesian state did attempt to guide
investment and the economy for a time.3
Astute observers have highlighted how the developmental state has been dying, even in its erst-
while home in Korea and Taiwan (Pirie, 2017) and wondered what comes ‘beyond’ or ‘after’ the
developmental state (Carroll & Jarvis, 2017; Fine, Saraswati, & Tavasci, 2013). Embedded in the diag-
nosis of state failure and corruption are two facets of development ideology stressed in all versions of
neoliberalism: maintaining growth and building (the ‘right’) institutions (Carroll, 2018; Carroll &
Jarvis, 2015). In the post-Washington Consensus language, ‘governance failures’ are the cause of per-
sistent weaknesses in resource ‘management’ among resource-dependent states. For the institution-
alists who now emphasize the spread of ideas, the answer is that new (or rather old neoliberal) ideas
spread to new elite policymakers who then adopt and adapt them to their variegated situations.
Development outcomes are a question of policymaker ‘intent’, to quote Henley’s comparison of
Asian success with African failure (Henley, 2015).
However, in East and Southeast Asia, as Carroll and Jarvis (2017, p. xviii) point out, ‘fetishizing of
growth and institutions in the region has too often come minus commensurate attention to the
impact of growth trajectories under globalization upon social relations and classes … and, in turn,
institutional change’. In resource exporters, I would add, the growth trajectories rely on available
supplies of ‘nature’. As neoliberalism advanced to work ‘through’ states by promoting ‘good govern-
ance’ (see Carroll, 2012), one observes – with considerable irony – the same international financial
institutions that promoted neoliberal dismantling of (‘bad’) states now promoting state capacity for
good governance in managing the revenue streams from resource exports while minimizing
corruption.
Importantly, the institutionalist recommendations of scholars and development policy prac-
titioners for addressing the resource curse have often led to the use of leverage and occasionally
 GLOBALIZATIONS 7

blatent power, especially at moments of crisis like the Asian Financial Crisis (AFC) of 1997–98. In
doing so, however, they have created what I term ‘altered states’. Altered states are developmental
states using different tools. Rather than dirigiste, they accept market mechanisms and the role of pri-
vate capital, thereby altering the methods of achieving development in the direction of markets. Like
earlier developmental states, they aspire to bring development by conventional measures (growth,
improved health, poverty reduction) to their populations and to industrialize and transform their
economies. However, they have altered goals in that they are less focused on state domination of
ownership or regulation. Resource nationalism in these states, as we shall see, is selective.
Altered developmentalism is evident across Asia in the two decades since the AFC as state invol-
vement in the economy has been rolled back. One might think that it is captured in the advance of
variegated forms of neoliberalism. However, it is vital to recall the overall thrust of neoliberal
advance, whereas Brenner et al. (2010) stress the evolutions and variations. Moreover, such varie-
gated capitalism perspectives, especially in their most recent emphases on ideas over interests
(Mukand & Rodrick, 2016; Rodrick, 2014), miss the material, class and world-systemic bases of
extractive regimes and changes wrought through the modalities of power identified by Gonzalez-
Vicente et al. (2017) as public private partnerships, financialization, and promotion of a conducive
business climate. In other words, altered states in general are neoliberalizing and maintain only lim-
ited roles due to the leverage of neoliberalisms’s dominant actors and institutions and the politics of
international-national relations.
Thinking through extractive regimes offers ideas to enhance the contribution of the critical stream
of research on neoliberalization viewed as the ‘deep marketization’ of the state under late capitalism.
Extractive regimes are locations of unevenness in the spread of neoliberal ideas and the limits of
‘deep marketization’(Carroll & Jarvis, 2017), and that may have something to do with the biogeophy-
sical bases of resources. The role of resource extraction has been significant, though not a curse, for
Malaysia’s ‘political business state’ (Jarvis, 2017) and Indonesia’s dominant politico-bureaucrats
(Robison & Hadiz, 2004).

Semi-peripheral extractive regimes and cycles of resource nationalism

World-systems analysis provides an alternative explanation for the political economy of persistent
resource extraction to the determinism of the resource curse and the voluntarism of the institution-
alist perspective. It is a more historical structural explanation for the emergence of developmental
states in the Cold War era and within regional and global political economies. The world-system,
as Dunaway and Clelland (2017b, p. 512, original emphasis) insist, ‘consists of the ever-changing tri-
modal structure of exploitative relationships that are loosely “governed” by an interstate system com-
prised of nation-states – all of which operate in contradictory ways to both support and resist the
survival of the system’. It is worth emphasizing that in their dialectical thinking ‘semiperipheral elites
inconsistently abet and resist core agendas’. Developmental states occur in the semi-periphery zone
of the world-system that are both exploited (by the core) and exploiter (of internal peripheries).
These are not simply states in the middle on a continuum of income levels as in the World Bank
tables. They have frequently been authoritarian, highly exploitative of labour and natural resources,
anti-union and demobilizing of civil society, while also providing significant public resources to edu-
cation, health, and welfare programmes. Wallerstein (1979) was an early proponent of the view that
semi-peripheral states tend to be ‘strong’ states. Bond (2012) characterizes the behaviour of such
states as Brazil and South Africa of late as ‘subimperialist’ due to the need of their capitalists to
expand territorially, frequently exploiting natural resources (see also Frame, 2018; Harvey, 2003).
8 P. K. GELLERT

From a world-systems perspective, with the diffusion of neoliberal policies, such states remain semi-
peripheral but their inconsistency is expressed in an ‘altered’ character. There are changes in the
modalities of rule, including key relations among state, domestic capital, and international capital,
as well as labour and other societal groups and classes often omitted from developmental state ana-
lyses. These states still appear ‘simultaneously developmental and predatory’ but ideologically they
are more in line with neoliberalism (Gomez & Lafaye De Micheaux, 2017; Witt & Redding, 2014).
By turning to extractive regimes and examining how their state institutions are altered by the poli-
tics of leverage over time, we avoid the ahistorical and apolitical weaknesses of the resource curse and
institutional perspectives. Extractive regimes are shaped by the position of states within the global
hierarchy of nation-states in the world-system.4 Of special importance for thinking about the role
of extraction in shaping polities and economies in the semi-periphery is that one of two preferred
development strategies of semi-peripheries is ‘state marketing of ecological resources (most lying
within ethnic communities)’ (Dunaway & Clelland, 2017a, p. 425). In relation to extraction, it is
important to also note how ‘semiperipheries engage in subimperialism, i.e. they behave like the
core in their economic and political relationships with peripheries and other semiperipheries
(Chase-Dunn, 1998, p. 210)’ (Bond, 2012; Dunaway & Clelland, 2017a, p. 438, emphasis original).5
The patterns of investment have changed in recent years, and this is particularly so in the high (but
not highest) capital-intensive sectors. By Dunaway and Clelland’s (2017a, p. 438) calculation, for
example, two-thirds of the largest land grabs between 2000 and 2016 were undertaken by corpor-
ations based in 6 semi-peripheral countries: Malaysia, Singapore, Saudi Arabia, India, China and
Argentina. The oil palm plantation expansion is a case in point.
These investments are how the world-system is tied together in global commodity chains (GCCs),
but GCC analysis has tended to ignore the politics of extraction by focusing instead on policy and
regulation. A few studies have demonstrated the significance of investigating the political strategies
by which extractive GCCs and the exploitation inherent in them are constructed (Ciccantell & Smith,
2009; Gellert, 2003), although not yet producing an ‘explicit theorization of the state’ (Smith, 2015, p.
294). Such theorization of the state within a world-systems perspective, additionally, does well to
understand the longue durée of ‘systemic cycles of accumulation’ (Arrighi, 1994/2010) and how
cycles shape the possibilities – and ideologies of – resource exporting (semi)peripheries.
Arrighi sees the material intensification and spatial expansion of the capitalist system unfolding in
alternating phases of material expansion and financial expansion. In the former, capitalists profit pri-
marily from investment in trade and production while in the latter, capitalists profit primarily from
accumulating property titles and other types of claims on future income (Arrighi, 2004, p. 532). The
shift from a phase of material expansion to a phase of financial expansion occurs when hegemons
experience crises of overaccumulation. When such crises occur, capitalists begin to experience
declining rates of profit from their investments in the territorial confines of an existing hegemon
and seek to resolve such crises by investing in different regions of the world and different economic
sectors that potentially offer higher and/or more stable rates of return. In response, places compete
for the investments of capitalists seeking out new sites of profit. In sum, for Arrighi the spatial expan-
sion of the capitalist economy occurs largely during crises of overaccumulation, as capitalists from an
existing hegemon seek out new sites to invest their mobile capital.
Expanding on Arrighi’s analysis of hegemonic ascent and Bunker and Ciccantell’s (2005) atten-
tion to the race for resources, Kaup and Gellert (2017) propose that cycles of resource nationalism
(CRNs) have occurred during two historical eras of hegemonic decline, the late 1930s to the early
1970s decline of Great Britain and the early 2000s to the present decline of the U.S. (Kaup & Gellert,
2017). In thinking about the emergence of new state modalities and the leverage exerted towards
 GLOBALIZATIONS 9

‘deep marketization’, it may be useful to also consider how the strategies of global powers aimed at
benefiting from the world’s raw materials simultaneously shaped resource nationalism, as Kaup and
Gellert describe for Bolivia and Indonesia. The forms of resource nationalism that emerged
responded to the extensive or intensive form of accumulation.
Great Britain sought to expand global capitalism through an extensive regime of colonial land
acquisition to secure access to raw material supplies. In response, resource (semi)peripheries pursued
anti-colonial development trajectories through which they made calls for greater national sover-
eignty. In the first CRN, extractive (semi) peripheries sought to own their natural resources and con-
trol the wealth that came from them. Ironically, it was the openings created by the United States and
its competitors that provided extractive peripheries with limited opportunities to pursue resource
nationalism and capture some of gains. By contrast, the United States deployed a more intensive
regime of accumulation through the nation-state system and Bretton Woods institutions, followed
later by processes of neoliberalization to ‘consolidate’ its power via mechanisms such as the establish-
ment of rules to benefit Western-based transnational corporations (Arrighi, 2004, p. 535). In
response, resource (semi)peripheries have sought to temper the power of transnational corporations
through calls for greater equity and fairness (Kaup & Gellert, 2017, p. 280).
The second CRN is frequently mischaracterized as a resurgence of resource nationalism
simply due to market opportunities from the commodity price boom. Warburton, too, rejects
the dominant view of resource nationalism as a temporary opportunistic response to price
booms. Similar to altered developmentalism, she describes it as ‘part of a re-emergent develop-
mentalist model’. However, in contrast to my world-historical stance she explains resource
nationalism largely in terms of its ideational basis among state policy makers (Warburton,
2018, p. 91; see also Haslam & Heidrich, 2016). In this period, we are witnessing the decline
of US hegemony, and potential rivals are offering new windows of opportunity in the peripheries
and semi-peripheries. Most prominent is China which has deployed an extensive regime of
accumulation and is expanding capitalism to both lesser incorporated frontiers and previously
more insulated market sectors of the world, working with extractive peripheries entertaining
new and upholding old policies of resource nationalism. Other rising economies such as India
and Brazil also seek to improve their position in the world economy by investing in and creating
the opportunities for extractive peripheries to gain greater control or revenue from their pursuit
of resource nationalism. In the following, I apply this deeply historical concept of an extractive
regime to Indonesia’s history of resource extraction and the cycles of hegemony and of resource
nationalism that have responded to it.
The extractive regime spanned the rule of Indonesia’s first two Presidents, Sukarno and Suharto.6
This period overlaps with early neoliberalism (version 1.0), but in Cold War Asia, dirigiste and
resource nationalist policies were accommodated. The extractive regime even outlived the violent
birth of the pro-Western capitalist New Order in 1965–66 when between half a million and a million
people were killed for being members of the Indonesian Communist Party (PKI), the largest in the
world at the time, or accused of being members of one of its affiliated organizations (such as the Bar-
isan Tani Indonesia or BTI). During three decades of authoritarian rule (1965–1998), the relation-
ship between extraction and rule became clear. Admittedly, some analysts and policymakers held
ambitions for Indonesia’s manufacturing ‘miracle’ to raise the nation-state up in the global hierarchy
(Hill, 1996). But despite manufacturing gains in the late 1980s and 1990s, it was natural-resource
based exports that remained vital to growth. In what follows, I briefly consider two periods, the
end of the first CRN in the 1950s to 70s and the beginning of the second CRN in the 1980s and
1990s. These periods correspond as well to neoliberalism 1.0 and 2.0.
10 P. K. GELLERT

Indonesia’s extractive regime, 1950s to 70s

The diffusion of neoliberalism 1.0 gained significantly as the economy was opened up to foreign
investment by Suharto in 1968, and the resource nationalist aims of state ownership and control
of oil, tin and rubber (Robison, 1986). Yet the New Order was not a fully liberalized government.
Post-colonial Indonesia, including both Sukarno’s amalgamation of socialism and nationalism
and Suharto’s New Order authoritarian capitalism, was ‘deeply engaged in non-liberal efforts
in relation to the economy’ (Carroll, 2017a, p. 138). The petroleum sector fell under a unified
national company, Pertamina, whose military leadership did not pursue national refineries but
continued Sukarno era demands for production sharing agreements from foreign companies
(Simpson, 2008) and benefited from the oil boom (see Figure 1). The forests were opened
to commercial exploitation, including a range of smaller domestic logging operations and
large foreign investors, but the latter were to come under increasing pressure by the 1980s
(Gellert, 2003).
The longevity of the extractive regime has an affinity with the resource nationalist possibilities
attendant to its cycles. Thus, the birth of the extractive regime goes back to the first cycle of resource
nationalism (Kaup & Gellert, 2017), during the waning years of British hegemony. Shaped by Great
Britain’s hegemonic strategy of free trade imperialism, the Dutch followed Britain’s laissez faire

Figure 1. Indonesian exports as percent of total constant value, 1962–2015.

Note: Colours, from bottom to top, represent the following SITC4 categories at medium resolution: food and live animals for food (yellow);
drinks & tobacco (pink); crude materials, inedible (light red); fuels, lubricants & related materials (brown); animal and vegetable oils
(yellow-brown); chemicals (magenta); manufactured goods (red); machinery & transport (light blue); other manufactured (turquoise);
other (dark blue). ‘Articles of’ refers to Articles of apparel and clothing accessories.
Source: The Atlas of Economic Complexity, Center for International Development at Harvard University, Accessed 22 December 2018 at
http://www.atlas.cid.harvard.edu.
 GLOBALIZATIONS 11

policies and opened up the NEI to private plantations, mining enterprises, and shipping vessels start-
ing in the 1860s. By the early twentieth century and before independence was declared in 1945 (and
longer before sovereignty was acknowledged by the Dutch only in 1949), resource nationalism
emerged amid anti-colonial struggles.7 It was only after independence in the 1950s that Dutch-
owned oil fields and plantations were nationalized. US capitalist interests were severely threatened,
and thus it was not too surprising that the US government supported clandestine separatist attacks
on Sukarno’s Indonesia in the late 1950s (Kahin, 1995) nor that it supported the army takeover and
genocide-scale killings in 1965–66 (Robinson, 2018; Roosa, 2006; Simpson, 2008).

Indonesia’s extractive regime 1980s to 90s

By the 1980s, Suharto’s New Order was firmly established, aided by large influxes of foreign aid and
the economy was open to foreign investment and global currency trade (Robison, 1986; Winters,
1996). Yet, despite that, the state continued to play a significant role in natural resource extraction,
and resource exports continued to be vital to the Indonesian economy. It was in the late New Order
period of the 1980s to 1990s that Indonesia’s state, like others in East Asia was considered a devel-
opmental state, even if it was ‘less apparent’ in Southeast Asia (Carroll, 2017a, p. 138). Neomercanti-
list policies were made possible by the dynamics of the Cold War period (Anderson, 1998; Felker,
2003), although Suharto’s picking of winners was less calibrated to world market success than clien-
telism and cronyism lubricated resource extraction. Manufacturing expanded (see Figure 1), but it
was oil money that greased the wheels of development spending (Winters, 1996), and ‘Bob’
Hasan, one of his key allies, parlayed presidential trust into a powerful tropical plywood industry
(Gellert, 2003).
The extractive regime was built on a variety of natural resources including oil, gas, mining, timber,
and fisheries. Even excluding the most lucrative oil and gas sector, other resource-based commodities
accounted for over 42% of exports in 1990–1994 and that only declined mildly to 36% in 2000–2004
(Gellert, 2010, p. 37).8 Moreover, in order to support capital accumulation based on extraction, the
state facilitated the concentration of control of forest resources on the one hand and the monopol-
ization of trade channels on the other. Thus, the regime was built on extraction, limited industrial
processing such as from logs to plywood, and export of these resource-based commodities. Indone-
sia’s tropical plywood industry came to dominate Japanese and other world markets through this
strategy. What the strategy did not do, however, was instil long-term investment in sustainable forest
management to keep up the resource. Indonesia’s developmental state, while pressured to liberalize
exchange rates and banking in the 1980s (Robison, 1986), did withstand the first wave of neoliber-
alism (neoliberalism 1.0). It was, however, unlike the quintessential developmental state of South
Korea, which built up domestically owned corporate powerhouses into globally recognized brands
of automobiles and electronics. Instead, after the Asian Financial Crisis and massive street protests
led to Suharto’s resignation, during the post-Suharto democratic reform period, more resources were
opened to extraction.
The multi-commodity extractive regime has persisted and, if anything, deepened in post-author-
itarian Indonesia. The leverage and power of capitalist liberalizers continue to extend the spatial
frontiers of commodification, while resource nationalists aim to retain the benefits for the putative
national community or public good. But whereas resource nationalism from the 1950s to early 1970s
emphasized ownership of extraction and processing and developmental statism in the 1980s and
early 1990s advocated broader industrial transformation, in the last decade marketization increas-
ingly relies on ‘comparative advantage’ and expanded extractivism. This altered developmentalism,
12 P. K. GELLERT

or ‘new developmentalism’ (Warburton, 2016, 2018), proposes a greater role for the state without
rejecting private capital.

Marketizing the extractive regime, altering the state, 2000 to present

Overall, in line with the increased global leverage and decreased space for autonomous national pol-
icy of neoliberalism 3.0, the Indonesian economy has continued to open up to foreign investors
through the last fifteen years. Most recently, with the election of private sector furniture businessman
President Joko (‘Jokowi’) Widodo in 2014, foreign investors were hopeful for a significant opening
(Emont, 2016). Jokowi announced a ‘big bang’ of deregulation and opening to foreign investment in
2016 (Reuters, 2016a). The World Bank’s Ease of Doing Business measure has become the de rigueur
instrument to discipline developing governments (Carroll & Jarvis, 2014b, p. 18, 2017, p. xix). In
2017, Indonesia moved up 15 places to 91st rank, as noted by the English-language Jakarta Globe
(Jakarta Globe, 2016). Nonetheless, this ranking still places Indonesia far behind Malaysia (23), Thai-
land (46), Mongolia (64), China (78), and Vietnam (82) (World Bank, 2017). Facing this embarrass-
ment, according to the American Chamber of Commerce in Indonesia (2017), Jokowi’s target is to
reach the top 40.
The commodity mix of exports has changed significantly and their weight in government reven-
ues has declined. The current cycle of resource nationalism in Indonesia’s extractive regime inter-
sects with an ‘altered’ state. However, state projects (proyek) continue to drive development
spending and contribute to informal sharing (bagi-bagi) of revenue down hierarchical ranks of gov-
ernment (Winters, 2011).9 As various others have noted about neoliberalism in general, the state has
not disappeared. The altered developmental state is still a gatekeeper to the ‘free gifts’ of nature but
does so through politicizing investment opportunities in ways that potentially make up, at least for
individuals and political party machines, for declines in (national) revenue from resources.
In this final section of the article, I consider how Indonesia is being repositioned as a resource
exporter. First, I explain why I view this as a ‘reperipheralization’ of Indonesia’s economy. Next, I
offer brief examinations of how the extractive regime has opened to foreign investment in oil
palm plantations and mining in part by addressing international (and domestic) capitalism’s need
for secure access to land for investment (Le Billon & Sommerville, 2017). However, this openness
has been counterbalanced by the second CRN, with policymakers pushing for downstream invest-
ments in mineral processing, although not as much in the palm oil sector (Warburton, 2017). In
the face of global pressures to continue to liberalize the economy, their tactics seem more related
to a distortion of resource nationalist sentiment as a kind of easy strategy than to a developmentalist
push to increase investment and build domestic capital(ists).

Reperipheralizing Indonesia
While development economists continue to comment on and advocate for Indonesia developing a
manufacturing industry (in the 1990s) and moving away from resource dependency (in the
2000s), the resource commodity boom of the first decade of the century exacerbated the emphasis
on resource exports in Indonesia. In fact, manufacturing declined as a share of exports and in absol-
ute levels for computers and parts (Coxhead & Shrestha, 2016; Pangestu, Rahardja, & Ing, 2015).
Thus, rather than moving from resource exports to manufactured goods to a service economy in
some linear path, Indonesia’s economy was reperipheralized but with new resource exports replacing
the old.10
 GLOBALIZATIONS 13

Oil and plywood, the key resource commodities of the earlier period, declined significantly. Oil
had accounted for 80% both of exports and of government revenue in the 1980s, but that declined
precipitously in the 1990s. Oil is no longer the revenue source it was. Indonesia became a net oil
importer in 2009.11 Additionally, timber products, which had been the second most important export
commodity shaped by the capital-state alliance of the prior period also declined significantly. With
the Indonesian Wood Panel Association’s leader Bob Hasan and his ally President Suharto gone
from the political scene (and Hasan briefly jailed for corruption on a forest aerial photography con-
tract), as well as its successor Wood Industry Revitalization Board weakened, the industry declined
precipitously.12 Table 1 shows the decline in plywood export volume from a peak of 8.6 million m3 in
1993 to less than 2 million just after the global financial crisis in 2009. Table 2 shows an equally dras-
tic decline in export revenue from over US$4 million to just over 1 million.
The dominance of world plywood exports in this sector that had been built by controlling the for-
ests and external trade relations by this semi-peripheral extractive regime (Gellert, 2003) similarly
declined in volume from 85% to 30% and in value to less than 40%. It is important to note the decline
in volume began before the 1997–98 crisis and continued despite recovering prices, which helped
revenue (see Japanese plywood import price in Figure 2). Industry capacity utilization fell to 42%
in 2005 and production fell from its peak of 9 million m3 to 3.3 million in 2012 (Makkarennu &
Nakayasu, 2013, p. 190). The main causes were lack of (domestic) investment in industrial efficiency,
thereby losing market share to Malaysia and China, the 2008 Great Recession decline in demand
from Japan and other markets, and especially the ‘imbalance’ between supply and demand. The latter
antiseptic term refers to the decades of forest degradation and the increased logging for local gains

Table 1. Volume of Indonesian nonconiferous plywood (in 1000 m3) as share of world exports, 1990–2016.
Year Indonesian plywood export volume World plywood export volume Percent of world exports
1990 7500 8779 85.43%
1991 7807 9331 83.67%
1992 8331 10,439 79.81%
1993 8571 11,754 72.93%
1994 7492 11,415 65.64%
1995 7280 11,585 62.84%
1996 6907 11,894 58.08%
1997 7143 11,919 59.93%
1998 6785 10,964 61.88%
1999 6188 10,416 59.41%
2000 6016 10,626 56.62%
2001 5738 10,616 54.05%
2002 5675 10,666 53.21%
2003 5262 10,690 49.23%
2004 4991 11,054 45.15%
2005 3652 9600 38.04%
2006 3028 9072 33.38%
2007 3001 8617 34.83%
2008 2146 6407 33.49%
2009 1919 6281 30.56%
2010 2040 6323 32.26%
2011 2080 5881 35.37%
2012 2150 5544 38.78%
2013 2322 5870 39.56%
2014 2318 6045 38.35%
2015 2338 5469 42.75%
2016 2556 5623 45.45%
Note: This data has no specific status comments in the ITTO source but see Note on Table 2 for more general problems with data accuracy.
Source: ITTO Biennial Review of Statistics updated 2017/06/01 (Accessed 12 October 2017 at http://www.itto.int/annual_review_output/?
mode = searchdata).
14 P. K. GELLERT

Table 2. Value of Indonesian nonconiferous plywood (in US$) as share of world exports, 1990–2016.
Year Indonesia plywood exports value World plywood exports value Percent of world exports
1990 2,626,337 3,036,681 86.5%
1991 2,727,616 3,229,962 84.4%
1992 3,098,560 3,805,466 81.4%
1993 4,047,876 5,409,349 74.8%
1994 3,551,779 5,192,665 68.4%
1995 3,330,219 5,101,029 65.3%
1996 3,285,724 5,419,320 60.6%
1997 3,399,771 5,335,504 63.7%
1998 2,157,840 3,298,931 65.4%
1999 2,426,845 3,844,201 63.1%
2000 2,384,669 3,797,951 62.8%
2001 1,989,475 3,281,050 60.6%
2002 2,001,633 3,398,788 58.9%
2003 1,956,360 3,446,016 56.8%
2004 2,193,840 4,214,249 52.1%
2005 1,914,623 3,845,190 49.8%
2006 1,689,624 3,991,013 42.3%
2007 1,533,511 3,996,644 38.4%
2008 1,280,033 3,426,574 37.4%
2009 1,047,163 2,749,984 38.1%
2010 1,414,967 3,341,033 42.4%
2011 1,634,731 3,736,875 43.7%
2012 1,599,600 3,446,470 46.4%
2013 1,741,500 3,660,050 47.6%
2014 1,854,580 3,891,133 47.7%
2015 1,858,710 3,441,616 54.0%
2016 2,032,020 3,594,048 56.5%
Note: ITTO reports perennially reveal the difficulties in collecting this data and the severe problems with accuracy. ITTO notes reveal that
1990–96 and 2008–10 data are from Comtrade database; 1997–2007 and 2011–12 from COMTRADE mirror statistics; and 2013–2016
are ITTO estimates.
Source: ITTO Biennial Review of Statistics updated 2017/06/01 (Accessed 12 October 2017 at http://www.itto.int/annual_review_output/?
mode = searchdata).

under decentralization, as well as conversion for plantations, initially fast-growing tree plantations
but increasingly oil palm. Further export increases have been recorded in the last couple years despite
Chinese domestic demand tapering (Forth, 2016). Tables 1 and 2 show the increases of late.
Despite these old resource sectors of oil and timber declining and broader news of Indonesia
becoming a consumer-driven middle-income country, however, the economy remained heavily
dependent on resource exports. What changed was the mix of exports. As Table 3 shows, gas exports
doubled in value from 2006 to 2011, palm oil exports tripled, and coal exports quadrupled. More-
over, throughout the period, resource-based exports as a percent of total exports has hovered around
forty percent.

Limits to ‘deep marketization’ in the oil, mining and palm oil sectors
Amid pressure to make resources available for foreign investors, Carroll has argued there is ‘the
death of development as any sort of national project’ in Indonesia (Carroll, 2017b, p. 107). While
this may be so, there are limits to deep marketization as the natural resource extractive sectors
are the focus of a new cycle of resource nationalism (CRN), structurally enabled by competition
for access by multiple players – China, India and other potential global powers (Kaup & Gellert,
2017). These sectors, including not only minerals and natural gas but also oil palm plantations
(which extract territory and value), are the object of continued and even renewed state ownership.
While Indonesia’s SOE to GDP ratio is very low at 2%, compared to Thailand, China, Brazil and
 GLOBALIZATIONS 15

Figure 2. Japanese index price for Indonesian plywood.

Source: ITTO Biennial Review of Statistics updated 2017/06/01 (Accessed 12 October 2017 at http://www.itto.int/annual_review_output/?
mode = searchdata), Table 4-3b.

Vietnam, which are all over 25% (Nem Singh & Chen, 2017:Figure 1b), the state continues to play
other roles through its control over land and ownership. In effect, state and capital have facilitated
palm oil land accumulation by dispossession and a coal mining boom with pressures to divest own-
ership. Economically, the slack of the major decline in oil and timber sectors has been taken up by
the booms in palm oil and coal. From 2% to 3% of exports, respectively, they jumped during the
global commodity boom to 9% and 14% by 2012 (Coxhead & Shrestha, 2016). Coal demand declined
in the last few years due to Chinese economic slowing but recovered in 2017–18.

Table 3. Resource-based exports from Indonesia, in US millions.

Total
Paper & resource- Resource-based
Crude Palm paper Wood Tuna & based as percent of
Year Total oil Gas Coal oil prod’s prod’s Metals Shrimp exports total exports
1999 48,665 3200
2000 62,124 6090 6625 1087 3414
2001 56,321 5715 5732 1081 3112
2002 57,159 5228 5578 1762 2092 2098 2781 1806.3 967 22,311 39.03%
2003 61,058 5621 6477 1980 2455 2007 2676 1914.2 965 24,094 39.46%
2004 71,585 6241 7750 2749 3442 2229 2905 1910.8 950 28,176 39.36%
2005 85,660 8146 9154 4354 3756 2325 2612 3451.0 965 34,762 40.58%
2006 100,799 8169 10,197 6086 4818 2859 2257 4863.5 1101 40,350 40.03%
2007 114,101 9226 9984 6681 7869 3375 1987 4821.1 1073 45,016 39.45%
2008 137,020 12,419 13,161 10,485 12,376 3797 1665 3868.9 1153 58,923 43.00%
2009 116,510 7820 8936 13,817 10,368 3405 1315 5378.9 953 51,993 44.63%
2010 157,779 10,403 13,670 18,499 13,469 4242 1748 7414.6 1059 70,504 44.69%
2011 203,497 13,829 22,872 27,222 17,261 4215 2124 6128.6 1285 94,936 46.65%
2012 190,020 12,293 20,521 26,166 17,602 3972 2048 4084.5 1365 88,052 46.34%
2013 182,552 10,205 18,129 24,501 15,839 3802 2098 4692.3 1623 80,890 44.31%
2014 175,980 9215 17,180 20,819 17,465 3780 2294 1769.8 1917 74,440 42.30%
2015 150,366 6479 10,341 15,999 15,385 3606 2243 3277.2 1459 58,790 39.10%
Note: Metals is composed of nickel and copper but only copper in 2015 due to missing nickel data; wood products is total of sawnwood,
veneer and plywood.
Source: BPS data, except ITTO Biennial Review of Statistics for wood products.
16 P. K. GELLERT

In fact, Indonesia’s state has faced a dual problem with resource-based exports of lack of diver-
sification and limited revenue. On the one hand, resource exports continue to be a vital part of the
economy, providing valuable investment capital and export earnings that are relatively slow in other
sectors. When calculated as minerals (coal, oil and gas); animal and vegetable byproducts (predomi-
nantly palm oil); metals; paper; wood products; vegetable products (coffee, etc.); and animal products
(shrimp, etc.), resource-based exports (with limited agricultural crops) accounted for 51.8% of US
$141Bn in exports in 2016 (Simoes & Hidalgo, 2011).13
That figure represented a decline from 58.2% of US$224Bn in 2011 at the peak of the commodity
price boom and 57.1% of US 167 Bn in 2008 (see also Hamilton-Hart & Schulze, 2016). However, the
ratio was not unfamiliar as in 1999, after a decade of manufacturing growth and just after the econ-
omic (and political) crisis, this set of products accounted for 48.6% of US$58.5 Bn in exports. Figure
1 illustrates this trend in its summary of Indonesian exports over the longer time period from 1965 to
2015. In Table 3, I present similar data calculated from Central Bureau of Statistics on export data (to
2015) for 8 commodities or categories of commodities. The sum of these resource-based exports rose
from 34% to 45% and returned to 36%. The blue and turquoise bands of manufactured exports (in
Figure 1) grow from 1985 until the late 1990s. It is precisely in 1998 when the Asian financial crisis
hit, that the yellow band of oils (i.e. palm oil) starts to expand. The growth of palm oil was facilitated
by the collapse of the Indonesian rupiah and IMF leverage to open the sector to foreign investors.
The persistence of the extractive regime continues to follow the pattern of semi-peripheral states
taking a large role in providing access to and supporting the exploitation of nature and people
(labour). On the other hand, given the altered state developmentalism without dirigiste direct own-
ership or control of sectors, these non-oil commodity booms have not provided the same kind of
revenue to the central government as oil and gas did in earlier periods. Indonesia’s state-owned
oil company, Pertamina, had pioneered production-sharing agreements. In forestry, based on
pressure to enforce good forest stewardship, the Ministry of Forestry had introduced a reforestation
fee (officially a bond or deposit) that provided millions. When the Ministry was deemed corrupt
during the Asian financial crisis and under IMF pressure, the management of the funds were
moved to the Treasury (Barr, 2001).
By contrast, for palm oil, there is an export tax in effect that pivots on market competition and is
only collected during periods of high market prices. In September 2017 for example, there was no tax
because the price fell below the benchmark $750 per ton (Reuters, 2017). For coal there is a royalty
ranging from 3% to 7% for open pit mines (PwC, 2016). Altogether these taxes do not amount to as
much revenue for the central government. While markets thereby shape revenues, the state main-
tains a role in the allocation of oil palm plantation permits. That is why the regional governments,
under decentralization, have been so interested in allocating permits via informal sales, and cliente-
listic control prevails over elections as various NGOs and political scientists have found (Berenschot,
2018; Gecko Project and Mongabay, 2018). It is also why there has been such a struggle between cen-
tral and regional governments over control of this process, with both levels of government couching
their arguments in ‘rule of law’ (Gellert & Andiko, 2015).
The social and environmental dislocation of palm oil expansion have been tremendous, although I
do not have space here to elaborate the environmental. While defenders of expansion claim that it is
a poverty alleviation tool, the results have been quite the opposite. Whereas during the authoritarian
period the so-called nucleus estates shared land with smallholders who together controlled 80% of
land, these ratios have been inverted so that companies control 75–80% (Potter, 2016).14 Moreover,
even though the oil palm plantations have provided jobs, the boom has increased informality and
depressed wages (Coxhead & Shrestha, 2016). Adding insult to injury, those who are precariously
 GLOBALIZATIONS 17

employed in the sector now are often those who did not have secure and recognized rights before-
hand and have been dispossessed from the accumulation and land grab associated with it (Gellert,
2015; Li, 2010).
President Jokowi’s land reform efforts importantly respond to long-standing demands from indi-
genous land rights advocates for the government to recognize and return rights to them. However,
land reform simultaneously works to ensure legal certainty and ‘rule of law’ for expanded extraction
(Gellert & Andiko, 2015). As explained by Minister Siti Nurbaya (MOEF, 2018, p. ix, emphasis
added), the government is working to resolve land tenure conflicts via the ‘Land Reform Program
(TORA) and the expansion of social forestry and encourage corporate [sic] to involve or partner
with the communities’. Jokowi’s land reform plan targets 9 million hectares for reform, precisely
one half for redistribution to local and indigenous (adat) communities and the other 4.5 million hec-
tares for ‘asset legalization’ of land use permit areas and release from the forest estate controlled by
the ministry. The redistribution is significant, but critics note that providing certificates only to ‘clean
and clear’ lands without resolving the hundreds of land conflicts is insufficient (Chandran, 2018).
Interestingly, of the asset legalization portion, 2.17 million ha are described as ‘no longer productive’
production forest areas (MOEF, 2018, p. 35) presumably due to the history of logging and forest
degradation (and without mention of people living in and around these areas). At the same time,
as of November 2017, despite efforts to maintain its institutional hold on forest areas (kawasan
hutan) (Gellert & Andiko, 2015), the ministry had also ‘released’ 6.8 million ha ‘for the establishment
of (mostly) oil palm plantations’ (MOEF, 2018, p. 38). Doing so is a de facto legalization of the plan-
tation sector’s acquisition of land. Meanwhile, of the 12.7 million hectares to be allocated for social
forestry, only 1.7 million had been transferred as of mid 2018. In line with the objective of legal cer-
tainty – a longstanding goal of international financial institutions like the World Bank and part of the
leverage of neoliberalism 3.0 – Minister of Agrarian Affairs Sofyan Djalil explained, ‘Our target is
that by 2025, all land in the country is registered’ (Chandran, 2018).
In contrast to an altered developmentalist balancing stance in the palm oil sector between local
and indigenous demands for land and the powerful palm oil sector, the state has taken a more
resource nationalist stance in the mining sector. This sectoral difference may be caused by the ‘inte-
grated ownership structures’ of Indonesian, Malaysian, and Singaporean capital in the palm oil sec-
tor in contrast to the ‘polarized’ divisions between foreign and largely Western capital opposed by
Indonesian capital in mining (Warburton, 2017, p. 304). As part of the mining sector, coal is subject
to the revision of previous ‘Contracts of Work’ agreements that began decades ago. The new Law on
Mineral and Coal Mining No. 4 of 2009 has faced significant political contestation. Officially it came
into effect in 2014 and required divestment – in a sequence including regional government partici-
pation – up to 49%, as well as downstream processing. In tandem with that, coal mining experienced
a shift from 75% foreign-owned companies in 2003 to 95% domestically owned by 2012 (Warburton,
2017, p. 298).
Global capitalists, notably copper and gold mining company Freeport MacMoran, voiced stren-
uous objections to rising resource nationalism in Indonesia, particularly the downstream processing
and divestment requirements. In early 2017, after the election of Jokowi, the government reversed
course on the goal of banning raw minerals exports, possibly to assist state-owned nickel company
Antam’s goal of acquiring Freeport. Antam had suffered Rp. 743 billion losses in 2014 and received a
state injection of Rp. 7 trillion in 2015 to foster investment in smelting (Warburton, 2017). In 2018,
the deal to divest 51% was sealed but with the state aluminum company, Inalum, which had become
a holding company in possession of 65% of Antam. Jakarta’s neoliberal-friendly English-language
newspaper issued an editorial aimed at the worries and leverage of foreign investors which assured
18 P. K. GELLERT

readers that the deal ‘guaranteed fiscal and legal certainty’. It further stated, based in part on Freeport
retaining leadership of the underground mining operations, ‘All parties involved are comfortable
with the deal, thereby refuting all the allegations of rising resource nationalism in Indonesia’ (Jakarta
Post, 2018).

Conclusion: the persistence of the semi-peripheral extractive regime

This article has argued that in the interstices of a global convergence towards neoliberal ‘deep mar-
ketization’ bolstered by various forms of leverage, Indonesia’s semi-peripheral extractive regime has
persisted but with an altered state developmentalism. Counter to the expectations of various theor-
etical perspectives on development, resource commodification and export retains a grip on Indone-
sia’s economy. For decades, Indonesia’s export profile has been dominated by a mix of resource
commodities extracted from its internal peripheries, but the mix has changed from oil and timber
to palm oil and coal. In the contemporary extractive regime, resource extraction continues to
shape the political economy even as the mix of commodities being extracted has changed.
The effect of ongoing resource commodification and exports, however, is not a simple story of the
resource ‘curse’ of low growth nor is the path away from extraction possible merely based on strong,
ethical leaders who have the will and ideas to construct ‘good’ (neo-liberal) institutions. Instead, like
most states in the world-system, Indonesia persists in its status in the semi-periphery, but it is also
shaped by long historical cycles of resource nationalism in periods of hegemonic decline and uncer-
tainty about new hegemonic power. In the first CRN, Indonesia’s extractive regime was built on a
state developmentalism that was skeptical of global markets and aimed toward nationalizing own-
ership of resources for progressive redistribution aims, industrial transformation, and collective
action to improve market prices. In large part, this unravelled during the Suharto’s New Order,
but some elements continued in the resource sectors. In the current CRN, altered developmentalism
is taking a more market-friendly stance shaped by the continued leverage of neoliberalism. In both
cycles there is a kind of sub-imperialism involved in the exploitation of the numerous internal per-
ipheries by a semi-peripheral state that ‘inconsistently abets and resists core agendas’. Although still
aspiring to bring growth and poverty reduction, the altered developmental state is now proceeding
via modalities of joint ownership and operation of mines and integrated ownership of plantations, as
well as local and regional dominance of the politics of obtaining land permits. The altered develop-
mentalism is accommodating some pressures for recognition of land rights from local and indigen-
ous groups, but the overall thrust of creating and expanding legal certainty expands the conditions
for ‘deep marketization’ through further and more stable commodification of the land and resources
of Indonesia. How long this extractive regime can continue, especially as it relies directly on spatial
expansion of resource frontiers that are eventually exhausted, and to the benefit of which foreign and
domestic dominant classes and actors are open questions.

Notes
1. Babb and Chorev (2016) identify ‘loose coupling’ of the development regime composed not only of
direct power over money and resources but also of the capacity to make and enforce international
rules and moral authority.
2. Land in general and agriculture in particular are not generally included by economists in their division of
economies into industry, agriculture and services. However, the spread of plantations often has charac-
teristics of extraction in its ongoing reliance on spatial expansion and the lack of reinvestment in pro-
ductive resources. Oil palm expansion in Indonesia illustrates this extractive trend as well as any. More
 GLOBALIZATIONS 19

broadly, as Jason Moore (2015) has argued, capitalism expands through not only (human) labour exploi-
tation in different parts of the world-system but through the appropriation of (nonhuman) nature. These
processes are intertwined in his concept of world-ecology that expands and degrades the oikios in its
path. See Gellert (2019) for an evaluation of Moore’s position in comparison to John Bellamy Foster
and Stephen Bunker.
3. Other recent efforts to address resource extraction and states have focused on Latin America (Massi &
Nem Singh, 2018; Saunders & Caramento, 2018) where the developmental state was also weak.
4. The notion of an extractive regime may call to mind other uses of the term regime. IR scholars since
Krasner refer to regimes in terms of implicit and explicit principles, norms, rules and decision-making
procedures around which actors expectations converge. This definition and approach, like the recent
contributions on ideas in IPE, omit the uneven power dynamics to reach these areas of convergence.
A second use of regime is merely to refer to the political system, e.g. authoritarian regime. As noted
above, extractive regimes work within multiple political systems. A third usage of regime is in Fou-
cault-inspired works that examine regimes of truth in discursive constructions.
5. Others are less fixed or rigid in their categorization of the semi-periphery. Terlouw (2003, p. 89) com-
plains, for example, that the semi-periphery concept ‘has wide applications, but it is difficult to dis-
tinguish clear examples. The semi-periphery is an analytical category for the analysis of changing
spatial patterns of inequality. As such it is a useful tool, but no substitute for that analysis’ Terlouw
(2003).
6. Arguably, it began even earlier during centuries of domination by the Dutch colonial state and before
that the VOC (Vereenigde Oostindische Compagnie). Yet, even limiting our attention to the last century,
the persistence of Indonesia’s extractive regime across multiple historical periods and the evolution of
the modalities of the regime are evidence of (neo)liberalism’s gradual diffusion and evolution.
7. In the important but short historical period of World War II and the Japanese occupation of Netherlands
East Indies, the idea of national development was also popularized. It is important as well to note its class
dimension as
the social strata most captivated by this shimmering vision of a transcendent, alternative Asian
modernity, and in turn its most vocal driving force, were representatives of the rising middle
classes, its bourgeoisies, or what we can also call modernizing, nation-building subelites. (Mark,
2006, p. 464)
8. Gellert (2010) codifies resource based exports as including SITC 2-digit categories 63 (wood products);
03 (fish); 23 (rubber); 28 (metallic ore); 42 (vegetable oils); 82 (furniture); 32 (coal); 68 (nonferrous
metals) and 64 (paper). He may have erred in including nonferrous metals, which includes aluminum,
although that is a very small sub-category.
9. The bagi-bagi phenomenon has also been examined in urban community development from an anthro-
pological perspective in Medan, Indonesia. See Jakimow (2018).
10. Note the caveat that one major limitation with trade data is the lack of accurate reporting, ‘especially true
for exports’ (Babones, 2013, p. 35). Foreign investment in Indonesia is also notoriously difficult to
measure due to poor reporting (see Carney & Hamilton-Hart, 2015). Recent data for 2015 show a dis-
crepancy between Bank of Indonesia data showing US$603.02 million of investments from the US while
the Bureau of Economic Analysis in the US shows $13,546 million (converted from rupiah using an
exchange rate of Rp. 11,900 = US$1) – over twenty times as much (U.S. Department of Commerce,
2017). See also Hamilton-Hart and Schulze (2016).
11. A few years later it somewhat confusingly re-joined OPEC only to have its membership frozen by OPEC
and then suspended by Indonesia itself in 2016 (Reuters, 2016b). Now it is reportedly headed towards
becoming a net gas importer by 2020 as well (The Straits Times, 2017).
12. While many elites were accused of corruption during Suharto’s rule, Hasan was the most prominent
figure actually convicted. He served three years in jail (2001–2004) and was fined $243 million for his
misuse of government funds intended for aerial photographs and mapping of the condition of the forest.
13. All figures in this paragraph were calculated from running various years of the UC Atlas of Global
Inequality and examining data at the 2-digit HS level. For example, 2016 data was accessed 17 October
2017 at http://atlas.media.mit.edu/en/visualize/tree_map/hs92/export/idn/all/show/2016/. Earlier years
use SITC codes and are not directly comparable.
20 P. K. GELLERT

14. In some instances (Potter, 2016; pers comm. W. Berenschot), local people have been told to find their
own partnership land to supply the company mills, even though land is not available for non-property-
holding rural folks.

Acknowledgements
I would like to express thanks to all the participants in the Florence Workshop that led to this special issue,
especially Toby Carroll and Darryl Jarvis, for their feedback and camaraderie. This paper was produced during
an academic leave when I was an Affiliated Fellow at KITLV. I appreciate Ward Berenschot’s helpful com-
ments, as well as those of one of the anonymous reviewers.

Disclosure statement
No potential conflict of interest was reported by the author.

Notes on contributor
Dr. Paul K. Gellert is Associate Professor in the Department of Sociology and Director of Global Studies at the
University of Tennessee. His research interests centre on the political economy of natural resources and the
politics of development and environment, especially in Indonesia. He has published in International Journal
of Comparative Sociology, Journal of Asian Studies, and Journal of Contemporary Asia. He co-edited Ecolo-
gically Unequal Exchange: Environmental Injustice in Comparative and Historical Perspective (Palgrave,
2019) and a 2017 special issue of the open-access Journal of World-Systems Research (http://jwsr.pitt.edu/).
New collaborative work (with Paul Ciccantell) on the global political economy of coal appears in The Oxford
Handbook of Energy and Society.

ORCID
Paul K. Gellert http://orcid.org/0000-0002-1832-0120

References
American Chamber of Commerce. (2017). Policy package 16 postponed. Retrieved from https://www.amcham.
or.id/politics/5619-policy-package-16-postponed
Anderson, B. R. O. (1998). From miracle to crash. London Review of Books, 20(8), 3–7.
Arrighi, G. (1994/2010). The long twentieth century: Money, power, and the origins of our times. London: Verso.
Arrighi, G. (2004). Spatial and other fixes of historical capitalism. Journal of World-Systems Research, 10(2),
527–539.
Ascher, W. (1998). From oil to timber: The political economy of off-budget development financing in
Indonesia. Indonesia, 65, 37–61.
Auty, R. (1993). Sustaining development in mineral economies: The resource curse thesis. London: Routledge.
Babb, S., & Chorev, N. (2016). International organizations: Loose and tight coupling in the development
regime. Studies in Comparative International Development, 51(1), 81–102.
Babones, S. J. (2013). Methods for quantitative macro-comparative research. New York: Sage.
Barr, C. (2001). Banking on sustainability: Structural adjustment and forestry reform in post-Suharto Indonesia.
Washington, DC: WWF Macroeconomics Program Office and Center for International Forestry Research
(CIFOR).
Berenschot, W. (2018). The political economy of Clientelism: A comparative study of Indonesia’s patronage
democracy. Comparative Political Studies. doi:10.1177/0010414018758756
Birdsall, N., & Fukuyama, F. (2011). The post-Washington consensus: Development after the crisis. Foreign
Affairs. Retrieved from https://www.foreignaffairs.com/articles/2011-02-16/post-washington-consensus
 GLOBALIZATIONS 21

Bond, P. (2012). Financialization, corporate power and South African subimperialism. In R. Cox (Ed.),
Corporate power and globalization in U.S. Foreign policy (pp. 114–132). New York: Routledge.
Booth, A. (Ed.). (1995). The oil boom and after: Indonesian economic policy and performance in the Soeharto
era. New York: Oxford University Press.
Brenner, N., Peck, J., & Theodore, N. (2010). Variegated neoliberalization: Geographies, modalities, pathways.
Global Networks, 10(2), 182–222. doi:10.1111/j.1471-0374.2009.00277.x
Bresnan, J. (1993). Managing Indonesia: The modern political economy. New York: Columbia University Press.
Bunker, S. G. (1985). Underdeveloping the Amazon: Extraction, unequal exchange, and the failure of the modern
state. Chicago: The University of Chicago Press.
Bunker, S. G. (1989). Staples, links, and poles in the construction of regional development theories. Sociological
Forum, 4(4), 589–610.
Bunker, S. G., & Ciccantell, P. S. (2005). Globalization and the race for resources. Baltimore, MD: The Johns
Hopkins University Press.
Carroll, T. (2010). Delusions of development: The World Bank and the post-Washington consensus in Southeast
Asia. London: Palgrave-MacMillan.
Carroll, T. (2012). Working on, through, and around the state: The deep marketisation of development in the
Asia-Pacific. Journal of Contemporary Asia, 42(3), 378–404.
Carroll, T. (2017a). Capitalism, contradiction and the onward march of variegated neoliberalism in Southeast
Asia. In R. Westra (Ed.), The political economy of emerging markets: Varieties of Brics in the age of global
crises and austerity (pp. 135–158). London: Routledge.
Carroll, T. (2017b). Late capitalism and the shift from the ‘developmental state’ to the variegated market state.
In T. Carroll & D. S. L. Jarvis (Eds.), Asia after the developmental state: Disembedding autonomy (pp. 93–
123). Cambridge: Cambridge University Press.
Carroll, T. (2018). Neoliberalism and multilateral development organisations in Southeast Asia. In A.
McGregor, L. Law, & F. Miller (Eds.), Routledge handbook of Southeast Asian development (pp. 48–60).
New York, NY: Routledge.
Carroll, T., & Jarvis, D. S. L. (2014a). Introduction: Financialisation and development in Asia under late capit-
alism. Asian Studies Review, 38(4), 533–543. doi:10.1080/10357823.2014.956284
Carroll, T., & Jarvis, D. S. L. (2014b). Theorising Asia’s marketisation under late capitalism: Risk, capital and
the new politics of development. In T. Carroll & D. S. L. Jarvis (Eds.), The politics of marketising Asia (pp. 1–
23). Abingdon: Palgrave MacMillan.
Carroll, T., & Jarvis, D. S. L. (2015). The new politics of development: Citizens, civil society, and the evolution
of neoliberal development policy. Globalizations, 12(3), 281–304. doi:10.1080/14747731.2015.1016301
Carroll, T., & Jarvis, D. S. L. (Eds.). (2017). Asia after the developmental state: Disembedding autonomy.
Cambridge: Cambridge University Press.
Carney, R. W., & Hamilton-Hart, N. (2015). What do changes in corporate ownership in Indonesia tell us?
Bulletin of Indonesian Economic Studies, 51(1), 123–145.
Chandran, R. (2018, September 24). Millions left out as Indonesia pushes land reform, activists say. Reuters.
Retrieved from https://www.reuters.com/article/us-indonesia-landrights-lawmaking/millions-left-out-as-
indonesia-pushes-land-reform-activists-say-idUSKCN1M41NZ
Chase-Dunn, C. (1998). Global formation: Structures of the world-economy. Lanham, MD: Rowman &
Littlefield Publishers.
Ciccantell, P., & Smith, D. A. (2009). Rethinking global commodity chains. International Journal of
Comparative Sociology, 50(3–4), 361–384. doi:10.1177/0020715209105146
Coxhead, I., & Shrestha, R. (2016). Could a resource export boom reduce workers’ earnings? The labour-mar-
ket channel in Indonesia. Bulletin of Indonesian Economic Studies, 52(2), 185–208. doi:10.1080/00074918.
2016.1184745
Cumings, B. (1984). The origins and development of the Northeast Asian political economy: Industrial sectors,
product cycles, and political consequences. International Organization, 38(1), 1–40.
Davidson, J. S. (2017). Survival of the weakest? The politics of independent regulatory agencies in Indonesia. In
T. Carroll & D. S. L. Jarvis (Eds.), Asia after the developmental state: Disembedding autonomy (pp. 237–260).
Cambridge: Cambridge University Press.
Deyo, F. C. (1989). Beneath the miracle: Labor subordination in the new Asian industrialism. Berkeley, CA:
University of California Press.
22 P. K. GELLERT

Dunaway, W. A. (2014). Bringing commodity chain analysis back to its world-systems roots: Rediscovering
women’s work and households. Journal of World-Systems Research, 20(1), 64–81.
Dunaway, W. A., & Clelland, D. A. (2017a). Moving toward theory for the 21 century: The centrality of non-
western semiperipheries to world ethnic/racial inequality. Journal of World-Systems Research, 23(2), 399–
464.
Dunaway, W. A., & Clelland, D. A. (2017b). Response to commentators. Journal of World-Systems Research, 23
(2), 511–514.
Dunning, T. (2005). Resource dependence, economic performance, and political stability. Journal of Conflict
Resolution, 49, 451–482.
Emont, J. (2016, October 20). Visionary or cautious reformer? Indonesian president Joko Widodo’s two years in
office. Time. Retrieved from http://time.com/4416354/indonesia-joko-jokowi-widodo-terrorism-lgbt-economy/
Evans, P. (1995). Embedded autonomy: States and industrial transformation. Princeton, NJ: Princeton
University Press.
Evans, P. B. (1989). Predatory, developmental, and other apparatuses: A comparative political economy per-
spective on the third world state. Sociological Forum, 4(4), 561–587.
Felker, G. B. (2003). Southeast Asian industrialisation and the changing global production system. Third World
Quarterly, 24(2), 255–282. doi:10.1080/0143659032000074583
Fine, B., Saraswati, J., & Tavasci, D. (Eds.). (2013). Beyond the developmental state: Industrial policy into the
twenty-first century. New York: Palgrave Macmillan.
Forth, K. D. (2016, October 4). Global plywood market: China losing edge; Indonesia gains. Woodworking
Network. Retrieved from https://www.woodworkingnetwork.com/news/woodworking-industry-news/
global-plywood-market-china-losing-edge-indonesia-gains
Frame, M. (2018). The role of the semi-periphery in ecologically unequal exchange: A case study of
land investments in Cambodia. In R. S. Frey, P. K. Gellert, & H. F. Dahms (Eds.), Ecologically
unequal exchange: Environmental injustice in comparative and historical perspective (pp. 75–106).
London: Palgrave Macmillan.
Frey, R. S., Gellert, P. K., & Dahms, H. F. (Eds.). (2019). Ecologically unequal exchange: Environmental injustice
in comparative and historical perspective. London: Palgrave Macmillan.
Gecko Project and Mongabay. (2018). Ghosts in the machine: The land deals behind the downfall of
Indonesia’s top judge. In Mongabay series: Indonesia for sale, Indonesian forests, Indonesian palm oil.
Mongabay. Retrieved from https://news.mongabay.com/2018/04/ghosts-in-the-machine-the-land-deals-
behind-the-downfall-of-indonesias-top-judge/
Gellert, P. K. (2003). Renegotiating a timber commodity chain: The politics of the Indonesia-Japan plywood
link. Sociological Forum, 18(1), 53–84.
Gellert, P. K. (2010). Extractive regimes: Toward a better understanding of Indonesian development. Rural
Sociology, 75(1), 28–57.
Gellert, P. K. (2015). Palm oil expansion in Indonesia: Land grabbing as accumulation by dispossession. In J.
Shefner (Ed.), Current perspectives in social theory (pp. 65–99, vol. 34). Bingley: Emerald Press.
Gellert, P. K. (2019). Bunker’s ecologically unequal exchange, foster’s metabolic rift, and Moore’s world-ecol-
ogy: Distinctions with or without a difference? In R. S. Frey, P. K. Gellert, & H. F. Dahms (Eds.), Ecologically
unequal exchange: Environmental injustice in comparative and historical perspective (pp. 107–140). London:
Palgrave Macmillan.
Gellert, P. K., & Andiko. (2015). The quest for legal certainty and the reorganization of power: Struggles over
forest law, permits, and rights in Indonesia. The Journal of Asian Studies, 74(3), 639–666. doi:10.1017/
S0021911815000613
Gomez, E. T., & Lafaye De Micheaux, E. (2017). Diversity of Southeast Asian capitalisms: Evolving state-
business relations in Malaysia. Journal of Contemporary Asia, 47(5), 792–814. doi:10.1080/00472336.
2017.1322629
Gonzalez-Vicente, R., Carroll, T., & Jarvis, D. S. L. (2017, October 26–27). The political economy of neoliber-
alism’s diffusion as operating system for the world market. Paper presented at the Workshop on ‘Leverage and
Marketisation: The Diffusion of the Neoliberal Development Agenda from North to South and Back Again’,
Florence.
Gunnoe, A. (2014). The political economy of institutional landownership: Neorentier society and the financia-
lization of land. Rural Sociology, 79(4), 478–504. doi:10.1111/ruso.12045
 GLOBALIZATIONS 23

Gunnoe, A., & Gellert, P. K. (2011). Financialization, shareholder value, and the transformation of timberland
ownership in the USA. Critical Sociology, 37(3), 265–284.
Haggard, S. (1990). Pathways from the periphery: The politics of growth in the newly industrializing countries
(P. J. Katzenstein, Eds.). Ithaca, NY: Cornell University Press.
Hamilton-Hart, N., & Schulze, G. G. (2016). Taxing times in Indonesia: The challenge of restoring competi-
tiveness and the search for fiscal space. Bulletin of Indonesian Economic Studies, 52(3), 265–295. doi:10.1080/
00074918.2016.1249263
Harvey, D. (2003). The new imperialism. Oxford: Oxford University Press.
Haslam, P. A., & Heidrich, P. (Eds.). (2016). The political economy of natural resources and development: From
neoliberalism to resource nationalism. New York: Routledge.
Henley, D. (2015). Asia-Africa development divergence: A question of intent. New York: Zed Books.
Hill, H. (1996). The Indonesian economy since 1966: Southeast Asia’s emerging giant. New York: Cambridge
University Press.
Hirschman, A. O. (1977). A generalized linkage approach to economic development with special reference to
staples. In N. Manning (Ed.), Essays on economic development and cultural change in honor of Bert
F. Hoselitz (pp. 67–98). Chicago, IL: University of Chicago Press.
Jakarta Globe. (2016, October 26). Indonesia climbs 15 places to rank 91st in World Bank’s ease of doing
business. Retrieved from http://jakartaglobe.id/business/indonesia-climbs-15-places-rank-91st-world-
banks-ease-business/
Jakarta Post. (2018, October 2). Freeport acquisition sealed. The Jakarta Post. Retrieved from http://www.
thejakartapost.com/academia/2018/10/02/freeport-acquisition-sealed.html
Jakimow, T. (2018). A moral atmosphere of development as a share: Consequences for urban development in
Indonesia. World Development, 108, 47–56. doi:10.1016/j.worlddev.2018.03.023
Jarvis, D. S. L. (2017). The state and development in Malaysia: Race, class and markets. In T. Carroll & D. S. L.
Jarvis (Eds.), Asia after the developmental state: Disembedding autonomy (pp. 201–236). Cambridge:
Cambridge University Press.
Kahin, A. R. (1995). Subversion as foreign policy: The secret Eisenhower and Dulles Debacle in Indonesia.
New York: The New Press.
Kaup, B. Z., & Gellert, P. K. (2017). Cycles of resource nationalism: Hegemonic struggle and the incorporation
of Bolivia and Indonesia. International Journal of Comparative Sociology, 58(4), 275–303. doi:10.1177/
0020715217714298
Kurtz, M. (2001). State developmentalism without a developmental state: The public foundations of the ‘free
market miracle’ in Chile. Latin American Politics and Society, 43(2), 1–25.
Kurtz, M. J., & Brooks, S. M. (2011). Conditioning the “resource curse”: Globalization, human capital, and
growth in oil-rich nations. Comparative Political Studies, 44(6), 747–770. doi:10.1177/0010414011401215
Le Billon, P., & Sommerville, M. (2017). Landing capital and assembling ‘investable land’ in the extractive and
agricultural sectors. Geoforum, 82, 212–224. doi:10.1016/j.geoforum.2016.08.011
Li, T. M. (2010). To make live or let die? Rural dispossession and the protection of surplus populations (pp. 66–
93, vol. 41). Wiley-Blackwell.
Makkarennu, & Nakayasu, A. (2013). Prospective Indonesian plywood in the global market. Journal of Life
Sciences and Technologies, 1(3), 190–195. doi:10.12720/jolst.1.3.190-195
Mark, E. (2006). “Asia’s” transwar lineage: Nationalism, marxism, and “greater Asia” in an Indonesian inflec-
tion. The Journal of Asian Studies, 65(3), 461–493. doi:10.1017/S0021911806001100
Massi, E., & Nem Singh, J. (2018). Industrial policy and state-making: Brazil’s attempt at oil-based industrial
development. Third World Quarterly, 39(6), 1133–1150. doi:10.1080/01436597.2018.1455144
Mies, M. (1986). Patriarchy and accumulation on a world scale: Women in the international division of labor.
Atlantic Highlands, NJ: Zed Books.
MOEF. (2018). The state of Indonesia’s forests 2018. Jakarta: Ministry of Environment and Forestry, Republic of
Indonesia.
Moore, J. W. (2014). The value of everything? Work, capital, and historical nature in the capitalist world-ecol-
ogy. Review (Fernand Braudel Center), 37(3–4 (World-Ecological Imaginations)), 245–292.
Moore, J. W. (2015). Capitalism in the web of life: Ecology and the accumulation of capital. New York: Verso.
Mukand, S. W., & Rodrick, D. (2016). Ideas versus interests: a unified political economy framework. Retrieved
from https://drodrik.scholar.harvard.edu/files/dani-rodrik/ … /ideasinterestsapr10sm_dr.pdf
24 P. K. GELLERT

Nem Singh, J., & Bourgouin, F. (Eds.). (2013). Resource governance and developmental states in the global south:
Critical international political economy perspectives. New York: Palgrave MacMillan.
Nem Singh, J., & Chen, G. C. (2017). State-owned enterprises and the political economy of state–state relations
in the developing world. Third World Quarterly, 1–21. doi:10.1080/01436597.2017.1333888
Neusiedl, C. (2017). The deep marketisation of development in Bangladesh. Third World Quarterly, 38(7),
1639–1654. doi:10.1080/01436597.2016.1229567
Pangestu, M., Rahardja, S., & Ing, L. Y. (2015). Fifty years of trade policy in Indonesia: New world trade, old
world treatments. Bulletin of Indonesian Economic Studies, 51(2), 239–261. doi:10.1080/00074918.2015.
1061915
Parenti, C. (2015). The environment making state: Territory, nature, and value (the 2013 antipode AAG lec-
ture). Antipode, 47(4), 829–848. doi:10.1111/anti.12134
Peet, R., & Hartwick, E. (1999). Theories of development. New York, NY: The Guilford Press.
Peluso, N. L., & Vandergeest, P. (2001). Genealogies of the political forest and customary rights in Indonesia,
Malaysia, and Thailand. The Journal of Asian Studies, 60(3), 761–812.
Pirie, I. (2017). Korea and Taiwan: The crisis of investment-led growth and the end of the developmental state.
Journal of Contemporary Asia, 1–26. doi:10.1080/00472336.2017.1375136
Potter, L. (2016). Alternative pathways for smallholder oil palm in Indonesia: International comparisons. In R.
Cramb & J. McCarthy (Eds.), The oil palm complex: Smallholders, agribusiness and the state in Indonesia and
Malaysia (pp. 155–188). Singapore: National University of Singapore Press.
PwC. (2016). Mining in Indonesia: Investment and taxation guide. Retrieved from https://www.pwc.com/id/
en/energy-utilities-mining/assets/May2016/PwCIndonesia-mining-in-Indonesia-survey-2016.pdf
Reuters. (2016a, February 12). Indonesia plans “big bang” opening of economy to foreign investment.
Retrieved from https://www.reuters.com/article/us-indonesia-jokowi/indonesia-plans-big-bang-opening-
of-economy-to-foreign-investment-idUSKCN0VL0W9
Reuters. (2016b, December 1). Net oil importer Indonesia leaves producer club opec, again. Reuters. Retrieved
from https://www.reuters.com/article/us-opec-meeting-indonesia/net-oil-importer-indonesia-leaves-producer-
club-opec-again-idUSKBN13Q3M7
Reuters. (2017, August 28). Indonesia to keep export taxes for cocoa, palm oil at zero in September. Retrieved
from https://www.agriculture.com/markets/newswire/indonesia-to-keep-export-taxes-for-cocoa-palm-oil-
at-zero-in-september
Robinson, G. (2018). The killing season: A history of the Indonesian massacres, 1965–66. Princeton, NJ:
Princeton University Press.
Robison, R. (1986). Indonesia: The rise of capital. North Sydney: Allen & Unwin.
Robison, R., & Hadiz, V. R. (2004). Reorganising power in Indonesia: The politics of oligarchy in an age of mar-
kets (Vol. 4). New York: Routledge Curzon.
Robinson, J. A., Torvik, R., & Verdier, T. (2006). Political foundations of the resource curse. Journal of
Development Economics, 79, 447–468.
Rodrick, D. (2014). When ideas trump interests: Preferences, worldviews, and policy innovations. Journal of
Economic Perspectives, 28(1), 189–208.
Roosa, J. (2006). Pretext for mass murder: The September 30th movement and Suharto’s Coup d’etat in
Indonesia. Madison, WI: University of Wisconsin Press.
Ross, M. L. (2001). Timber booms and institutional breakdown in Southeast Asia. New York, NY: Cambridge
University Press.
Ross, M. (2013). The oil curse: How petroleum wealth shapes the development of nations. Princeton, NJ:
Princeton University Press.
Sachs, J. D., & Warner, A. M. (2001). Natural resources and economic development the curse of natural
resources. European Economic Review, 45, 827–838.
Saunders, R., & Caramento, A. (2018). An extractive developmental state in Southern Africa? The
cases of Zambia and Zimbabwe. Third World Quarterly, 39(6), 1166–1190. doi:10.1080/01436597.2017.
1409072
Simoes, A. J. G., & Hidalgo, C. A. (2011). The economic complexity observatory: An analytical tool for under-
standing the dynamics of economic development. In Workshops at the Twenty-Fifth AAAI Conference on
Artificial Intelligence: Creative Commons Attribution-Sharealike 3.0 Unported License.
 GLOBALIZATIONS 25

Simpson, B. R. (2008). Economists with guns: Authoritarian development and U.S.-Indonesian relations, 1960–
1968. Stanford, CA: Stanford University Press.
Smith, A. (2015). The state, institutional frameworks and the dynamics of capital in global production net-
works. Progress in Human Geography, 39(3), 290–315. doi:10.1177/0309132513518292
Terlouw, K. (2003). Semi-peripheral developments: From world-systems to regions. Capitalism Nature
Socialism, 14(4), 71–90. doi:10.1080/10455750308565547
The Straits Times. (2017, August 16). Indonesia’s oil and gas sector slumps, falling to 3% of nation’s GDP.
Retrieved from http://www.straitstimes.com/business/indonesias-oil-and-gas-sector-slumps-falling-to-3-
of-nations-gdp
US Department of Commerce. (2017). Section 9.4 foreign direct investment and foreign portfolio investment stat-
istics congress. Retrieved from https://www.export.gov/article?id = Indonesia-foreign-direct-investment-
statistics
Wallerstein, I. (1979). Semiperipheral countries and the contemporary world crisis. In I. Wallerstein (Ed.), The
capitalist world-economy: Essays (pp. 95–118). New York: Cambridge University Press.
Warburton, E. (2016). Jokowi and the new developmentalism. Bulletin of Indonesian Economic Studies, 52(3),
297–320. doi:10.1080/00074918.2016.1249262
Warburton, E. (2017). Resource nationalism in Indonesia: Ownership structures and sectoral variation in
mining and palm oil. Journal of East Asian Studies, 17(3), 285–312. doi:10.1017/jea.2017.13
Warburton, E. (2018). Nationalism, developmentalism and politics in Indonesia’s mining sector. In A.
Patunru, M. Pangestu, & C. Basri (Eds.), Indonesia in the new world (pp. 90–108). Singapore: ISEAS-
Yusuf Ishak Institute.
Williamson, J. (1990). What Washington means by policy reform. In J. Williamson (Ed.), Latin American
adjustment: How much has happened? (pp. 5–20). Washington, DC: Institute for International Economics.
Winters, J. (1996). Power in motion: Capital mobility and the Indonesian state. Ithaca, NY: Cornell University
Press.
Winters, J. A. (2011). Oligarchy. New York: Cambridge University Press.
Witt, M., & Redding, G. (Eds.). (2014). The Oxford handbook of Asian business systems. Oxford: Oxford
University Press.
Woo-Cumings, M. (Ed.). (1999). The developmental state. Ithaca, NY: Cornell University Press.
World Bank. (2017). Ranking of economies – ease of doing business.
 ECONOMIC RECORD, VOL. 94, NO. 307, DECEMBER, 2018, 469–499

Asia’s Third Giant: A Survey of the Indonesian

Economy*
HAL HILL
Australian National University, Canberra, ACT, Australia

This paper surveys the Indonesian economy and the drivers of

socioeconomic development over the past half-century. It highlights
the country’s rapid economic development in the face of unfavour-
able ‘initial conditions’. We examine episodes in economic devel-
opment, in particular comparing and contrasting the two main sub-
periods, of high economic growth during the authoritarian Suharto
era, 1966–98, and moderate economic growth during the demo-
cratic era since 1999. The paper emphasises the importance of
sound macroeconomic management, economic openness, inclusive
social progress and institutional development. For all the chal-
lenges that Indonesia faces, and its unfinished reform agenda, the
major conclusion is one of development success, broadly defined.

I Introduction facility) and strict adherence to orthodox macroe-

Few countries have had such varied economic
outcomes and policies as Indonesia. It has been
variously described as a ‘chronic economic
dropout’ (Higgins, 1968), a ‘miracle’ economy
(World Bank, 1993), a ‘showcase state’ (Mor-
timer, 1973), and an ‘improbable nation’ (Pisani,
2014). Its economic history has been described as
one of ‘missed opportunities’ (Booth, 1998,
2016). Its monetary policy history has featured
both a major episode of hyperinflation (including
the collapse of the central bank note-printing
*Thank are due to the following: Professor Anu
Rammohan for her editorial interest and advice; Chan-
dra Tri Putra for indispensable research assistance; two
anonymous Economic Record referees for very detailed
and helpful comments; seminar participants at the
ANU, University of Indonesia, and Gadjah Mada
University for their comments; colleagues and affiliates
past and present in the ANU’s Indonesia Project; and
several co-authors whose work I have drawn on, as
indicated. Special thanks go to the late Professor Heinz
Arndt for first sparking my interest in the subject
matter.
JEL classifications: N15, O53
Correspondence: Hal Hill, Australian National
University, Canberra, ACT, Australia. Email: hal.hil-
l@anu.edu.au
469
© 2018 Economic Society of Australia
doi: 10.1111/1475-4932.12439
conomic rules. Its political history has been
equally mixed. It was once a member of the
‘Peking–Pyongyang–Hanoi–Phnom Penh–Jakarta
axis of newly emerging forces’, as it withdrew
from virtually all major international institutions.
Then for a quarter of a century after 1966 it was a
close ally of the United States during the Cold
War era. Its foreign investment regimes have
swung from the nationalisation of all foreign
property by the early 1960s to an open-door
posture a few years later (Lindblad, 2015).
Indonesia matters to its neighbourhood, includ-
ing Australia, and increasingly to the world. It is
the world’s fourth most populous nation, with
about 265 million people. It has the largest
number of adherents to Islam. It is the dominant,
albeit low-key, power in the 630-million popula-
tion of the Association of Southeast Asian
Nations (ASEAN), the developing world’s most
successful and durable regional grouping. On
some projections, by 2050 it may also be the
world’s fourth largest economy; currently it is the
ninth largest (on a purchasing power parity (PPP)
basis). It has experienced one of the most severe
economic crises in modern global history, with a
growth collapse of some 20 percentage points in
1998. Yet out of this crisis came two positive
470 ECONOMIC RECORD
surprises: a moderately rapid economic recovery
and a sudden transition from one of the world’s
most durable authoritarian regimes (the 32-year
Suharto presidency) to a functioning and credible
democracy, on most indicators the freest both in
Southeast Asia and in the developing Muslim
world.
Indonesia also features enormous subnational
diversity. It is the world’s largest archipelagic
state, with an estimated 17,000 islands and more
than 300 ethnic groups and spoken languages.
Within its borders, there is massive diversity on
practically every socioeconomic indicator. For
example, the richest kabupaten (the districts to
which central government finances and adminis-
trative authority have been devolved since 2001)
has a per capita income about 50 times that of the
poorest. Its flora and fauna comprise many unique
tropical species.
The country’s size is of such significance that it
has on occasion had global environmental effects.
It has the world’s third most extensive tropical
rainforests. When these forests are depleted more
rapidly than usual, the country has been the fourth
largest emitter of carbon dioxide. The Krakatoa
and Tambora volcanic eruptions in 1883 and
1815, respectively, significantly affected global
climatic patterns, including several years without
summer in Europe. Indonesia also straddles one
of the world’s most important ‘choke points’ for
international commerce, the narrow Malacca
Straits through which East Asia’s shipping trade
with Europe, the Middle East and Africa passes.
Against this backdrop, this paper surveys the
Indonesian economy and the drivers of socioeco-
nomic development in the modern era. The major
themes are development success but daunting
policy challenges. Section II provides a scene-
setting overview, along with some of the param-
eters and context within which to examine
development outcomes and policy processes.
The sections that follow investigate sectoral
growth and structural change; investment and
savings; international trade; macroeconomic
management; subnational development dynam-
ics; poverty, inequality and living standards;
environmental dimensions; and institutions and
governance. Section XI sums up.
Several points need to be emphasised at the outset.
Indonesia is a new nation. It did not exist as a concept
until the nationalist struggles of the early twentieth
century. It did not exist as a formal nation state until
1945, and as a state that could be governed peacefully
until 1949. Its international boundaries have changed
DECEMBER
three times since independence. The country’s inde-
pendent history typically divides into three eras,
bookmarked by crisis years (1945, 1965, 1998) when
the country’s prospects were (wrongly) considered to
be exceptionally gloomy. These periods correspond to
the early independence era (1945–65), followed by
rapid economic growth with authoritarian governance
(1966–98), and then moderate growth with democracy
(1999 to the present).
The first two turbulent decades of indepen-
dence resulted in a sparse scholarly literature and
the near-collapse of what was in any case a
rudimentary statistical system. While emphasis-
ing the importance of conditioning historical
factors, this survey focuses mainly on the past
half-century, the country’s first era of sustained
economic growth and rising living standards,
punctuated by a deep economic and political
crisis. This division also enables us to make some
observations about economic outcomes and pol-
icy processes under two very different gover-
nance structures. 1
II Setting the Scene
(i) The Context and Early History
History and geography have shaped Indonesian
economic development. Almost 350 years of
Dutch colonial rule, until well into the twentieth
century, extractive and exploitative in nature,
defined not only the country’s geographical bound-
aries but also its dominant ideological currents.
Acemoglu and Robinson’s (2012, p. 250)
observation, drawing on Reid (1993), is apposite:
Dutch colonialism fundamentally changed
[Indonesia’s] economic and political develop-
ment. The people . . . stopped trading, turned
inward. . . . In the next two centuries, they
would be in no position to take advantage of the
innovations that would spring up in the indus-
trial revolution.
That is, the colony was increasingly integrated into
the global economy, but as an appendage to the
priorities and dictates of the imperial power, with
very little commercial and political autonomy.
1
A general note on sources is relevant at the outset.
Wherever comparative statistics are presented, the
source is the World Bank’s World Development Indi-
cators. Otherwise Indonesian statistics are used,
sourced from Statistics Indonesia (Badan Pusat Statis-
tik, www.bps.go.id).
© 2018 Economic Society of Australia
2018 A SURVEY OF THE INDONESIAN ECONOMY
Although Indonesia declared independence in
August 1945, as the occupying Japanese troops
were withdrawing, the Dutch attempted to retake
their colony and a bitter four-year war ensued. In
Asia, Indonesia’s decolonisation struggle was
second only to Vietnam’s in its severity. At
independence, Indonesia was a classic ‘dual econ-
omy’, with a small modern sector owned mainly by
foreigners and the numerically small but relatively
prosperous ethnic Chinese community, alongside
the indigenous community engaged in agriculture,
cottage industry and petty trade. 2
These events and conditions were to shape the
country’s political destiny for many years. Sukarno, the
country’s founding president, espoused nationalism and
socialism during his two-decade rule. Concerned mainly
with nation-building and preserving the country’s terri-
torial integrity, he had little interest in economics. As his
government increasingly lostcontrol of its budget, deficit
financing led to escalating inflation. Black markets
proliferated, particularly for foreign exchange. Directing
his ire abroad, in 1964 Sukarno told Western donors to
‘go to hell with your aid’ (Legge, 2003).
In 1965–6, Sukarno was deposed in controver-
sial circumstances. A gruesome episode of kill-
ings followed, and a military-backed regime took
control under President Suharto. The following
quotes by two of the most knowledgeable foreign
economists at the time summarise economic
conditions in the mid-1960s:
Indonesia must surely be accounted the number
one failure among the major underdeveloped
economies . . . a chronic economic dropout.
(Higgins, 1968, p. 678)
A decade of ever-increasing economic mis-
management had brought a degree of economic
breakdown with few parallels in modern his-
tory. The country was literally bankrupt, unable
to meet payments due to foreign debt. . . .
Export earnings had fallen to a level where they
were barely sufficient to finance half the
country’s minimum requirements, excluding
debt service. (Arndt, 1984, p. 29)
Van der Eng (2002) was the first economist to
carefully construct national accounts for the
country during the twentieth century. He
2
In fact, the international literature on dualism,
originating from the work of the Dutch economist J.H.
Boeke, was developed on the basis of theorising from
economic conditions and behaviour in colonial Indone-
sia (Szirmai, 2015).
© 2018 Economic Society of Australia
471
concluded that per capita income in 1965 was
lower than it had been half a century earlier. Thus
by the early 1960s Indonesia was one of the
poorest countries in the world. Table 1 presents
some socioeconomic indicators around 1960
compared to its neighbours. The data reveal how
far it lagged, especially on social indicators:
adults over the age of 25 had on average just
1.1 years of schooling, about half that of Malay-
sia and Thailand, and little more than one-third
that of the Philippines. Its infant mortality rates
were about double those of most of its neighbours
and among the highest in the developing world.
Geography has also been a crucial conditioning
factor in at least two respects. First, as a vast
archipelagic state adjacent to the very open
Singapore and Malaysian economies, Indonesia
is by definition trade-oriented, but only if the
policy environment is permissive. As a leading
Indonesian economist put it, ‘Indonesia was born
a free trader yet is consistently reluctant to accept
globalization’ (Basri, 2012, p. 46). Second,
Indonesia is richly endowed with energy, miner-
als, maritime, forest and land resources. It there-
fore has to manage the inherent macroeconomic
volatility and complex political economy associ-
ated with resource abundance, including the
struggle to avoid the ‘resource curse’.
These legacies continue to shape contempo-
rary economic policy: inflation aversion (a
response to the 1960s hyperinflation); ambiva-
lence about markets and globalisation; a philo-
sophical attraction to egalitarianism; concern
about inter-ethnic inequality; institutions still at
an embryonic stage of evolution; and a strong
commitment to the preservation of territorial
integrity. We return to these issues throughout
the paper.
An additional strand has been policy pragma-
tism, also shaped by the history of economic
collapse and its political ramifications. Owing to
the controversial beginnings of his presidency
and his suppression of human rights, Suharto’s
legitimacy depended on economic growth, and for
that he turned to professional economists as
technocratic ministers. The following quotations
from two of the key policy actors are illustrative
of their approach to policy:
When we started out attracting foreign investment in
1967 everything and everybody was welcome. We
did not dare to refuse; we did not even ask for
bonafidity of credentials. (Mohammed Sadli, quoted
in Palmer, 1978, p. 100)
472 ECONOMIC RECORD
T ABLE 1
Comparative Socioeconomic Conditions, c. 1960
GDP per Years of
capita Trade schooling, for
(constant (% of those aged 15
Country 2010 $) GDP) and above
Indonesia 577 11.6 1.6
Malaysia 1,408 85.7 2.8
Philippines 1,059 38.3 3.5
Thailand 571 34.9 2.6
In 1954/55 [as finance minister], I was a strong
protagonist of foreign exchange controls. . . .
Then I saw what happened under . . . Sukarno. I
know how easy it is to smuggle goods, and I
know that those who are close to the sources of
power will get their hands on the foreign
exchange. (Djojohadikusumo, 1984, p. 38)
(ii) The Long-Term Growth Record
In one of the most important turning points in
Asian economic development, economic growth
accelerated quickly under the Suharto regime and
was sustained almost constantly throughout the
next three decades (Figure 1). 3 If ever there were
a country example to refute notions of ‘path
dependence’, Indonesia would surely be it. Fig-
ure 1 also identifies the country’s major sub-
periods: first, the latter years of slow growth in
the period prior to 1966; second, the period of
high growth, 1966–96, averaging 7.3 per cent per
annum; third, the Asian financial crisis (AFC) of
1996–9, when the economy contracted severely
(including by 13.4 per cent in 1998); and fourth,
the return to growth in the democratic era
beginning in 2000, when growth averaged 5.1
per cent. Hence, without drawing any inferences,
growth was about 2 percentage points faster in the
authoritarian period than in the democratic era.
There were also various episodes within these
3
For an authoritative account of economic develop-
ment during the first 15 years of Suharto’s rule, see
Booth and McCawley (1981). The recollections of
several of the key economic policy makers in the early
Suharto era are contained in Thee (2003). The major
growth accelerations in the other developing Asian
giants, generally dating from 1978 for China (Naugh-
ton, 2006) and 1991 for India (Joshi, 2017), were also
pronounced, but with the difference that these two
economies were at least registering positive if slow
growth prior to the reforms.
DECEMBER
Years of Life
schooling, for expectancy Infant mortality
those aged 25 at birth (deaths per
and above (years) 1,000 population)
1.1 47.0 166.7
2.3 57.9 81.1
3.0 57.1 86.5
2.1 53.3 108.9
periods, especially 1966–96, which included the
1970s oil boom, and the reform and readjustment
to the era of low energy prices beginning in the
early 1980s. As a result of this growth, per capita
gross domestic product (GDP) has also risen
rapidly, as indicated on the left-hand scale: over
the period 1966–96 it increased almost fourfold,
while for the period as a whole, including the
AFC, it has increased more than sixfold.
There are two common criticisms of these
growth numbers. First, they may be a ‘myth’, as
Krugman (1994) and others have argued was the
case for some Asian ‘miracle’ economies, in the
sense that the GDP growth simply reflects rapid
factor accumulation. This results in ‘growth with-
out productivity’, which is therefore not sustain-
able. However, this has not been the case for
Indonesia. Total factor productivity (TFP) growth
has been positive throughout the periods of rapid
economic growth, typically in the range 2–4 per
cent, and has exhibited year-to-year fluctuations
that mirror those of GDP quite closely. 4 ,5
The second criticism is that growth has been
purchased at the cost of running down the stock of
natural capital, particularly non-renewable natu-
ral resources. As we will show below, environ-
mental degradation has been a serious issue for
Indonesia. Estimates of ‘green growth’ are very
4
See, for example, the TFP estimates computed
across countries and over time for major Asian
economies by the Asian Productivity Organization
(http://www.apo-tokyo.org).
5
Another way of evaluating economic performance
is to compute Indonesian per capita income relative to
the ‘frontier’, typically proxied by US per capita
income. For the record, over the period 1965–2013,
Indonesian per capita income (in PPP terms) approxi-
mately doubled relative to that of the US and is now
approaching 20 per cent (Aswicahyono & Hill, 2016,
pp. 106–7).
© 2018 Economic Society of Australia
2018 A SURVEY OF THE INDONESIAN ECONOMY 473

F IGURE 1
Annual GDP Growth and GDP per Capita, 1960–2016

GDP per capita (constant 2010 IDR) GDP growth (annual %)

GDP per capita (constant 2010 IDR)

40 15

Millions
35 10

GDP growth (%)

30 5

25 0

20 –5

15 –10

10 –15

5 –20

Year

approximate and sensitive to the assumptions country belongs to a very small group of mainly
made. For what they are worth, as we show East Asian economies that have achieved excep-
below, they suggest that this concept of growth tionally rapid growth for a sustained period. Two
has been perhaps 1 percentage point slower than major studies by the World Bank (1993, 2008)
GDP growth. Importantly, however, there is not a illustrate this proposition. The Bank’s ‘miracle’
strong correlation between Indonesia’s growth study (1993) singled out seven East Asian
rates and its terms of trade. That is, the economy economies, including Indonesia, for their very
has grown quite quickly during periods of both high growth. The Growth Commission report
high and low commodity prices. This is in part (2008) asked which economies over the preceding
because administrations have tended to be more century had grown exceptionally fast, defined as
reform-oriented when ‘forced’ to by lower com- GDP growth averaging at least 7 per cent for at
modity prices. We return to this important polit- least a decade. The authors concluded that there
ical economy proposition below. were just 13 economies among the 150 for which
Two general questions on the growth record are they could obtain reliable estimates. Indonesia
pertinent. First, how does Indonesia’s record was one of these 13, for the period 1966–96.
compare internationally? And second, what do The country therefore belongs to a group of
the occurrence of crisis episodes and the stellar economic performers. It may perhaps be
responses to them tell us about the quality of churlish to attach a qualifier to this conclusion. It
Indonesian economic management? is that, among the East Asian stars, there is a
The comparative evidence is clear: although divide between Northeast and Southeast Asia,
there is no obvious comparator for Indonesia, 6 the placing Singapore analytically in the former
group. The first group – China, and four Newly
Industrialising Economies (NIEs) (and earlier
6
In the 1970s the usual comparator was Nigeria, Japan) – achieved such rapid growth that per
given its size, tropical location and endowment of capita income since around 1960 has risen 12- to
energy resources. Around 1970 Nigeria’s per capita 16-fold, whereas the increase for Indonesia and
income and human capital were somewhat higher than other high-growth Southeast Asian economies has
Indonesia’s. However, the two countries parted com- been a still very respectable six- to eight-fold.7
pany thereafter. By 1990, Indonesia’s per capita income
was three times that of Nigeria. See Bevan et al. (1999)
7
for a detailed study of the two countries. See Perkins (2013) who draws out this distinction.

© 2018 Economic Society of Australia

474
Per Capita Income, Indonesia and Comparators, 1960–2016
China
GDP per capita (constant 2010 USD)
8,000
7,000
6,000
5,000
4,000
3,000
2,000
1,000
0
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Figure 2 traces the comparison among the three
developing Asian giants. In the early 1960s, all
were extremely poor. As the first to reform,
Indonesia pulled away from the other two from
the late 1960s. By 1990, its per capita income was
still above China’s, and well above India’s. How-
ever, by this time China was growing even faster
than Indonesia, while neither China nor India was
affected by the AFC. Thus China overtook Indone-
sia around the mid-1990s, and the gap has widened
as Indonesia returned to slower growth from 2000.
India has narrowed the gap since the mid-1990s.
The figure also includes the Philippines, a country
that was well ahead of all the giants in the 1960s,
and which shares some similarities with Indone-
sia. 8 However, Indonesia overtook it in the early
1990s and, notwithstanding different experiences
during the AFC, Indonesia remains ahead.
Second, on the record of crisis management,
since the early 1960s Indonesia has experienced
four crisis or ‘near-crisis’ episodes. The record in
all cases but one is a positive one, as Figure 1
illustrates. The early to mid-1960s episode was
entirely home-grown in origins, and, as noted, the
regime change led to rapid growth and crisis
resolution, albeit at great human cost. The early
8
That is, location, archipelagic geography, and a
deep economic crisis triggering a sudden transition to
democracy.
ECONOMIC RECORD DECEMBER
F IGURE 2
India Indonesia Philippines Thailand
Year
to mid-1980s presaged a decade of debt and
contraction for most developing country energy
exporters, and at the beginning of the period
Indonesia looked highly vulnerable owing to its
heavy dependence on the energy sector for
exports and government revenue. However, prac-
tically alone among these countries, it experi-
enced only a mild and relatively brief slowdown
in growth. At the margin, external financial
support was helpful. But the main explanation
was adroit anticipation and management of a
looming crisis (Gelb and Associates, 1987; Hill,
2000). The government had invested the windfall
gains from the oil boom quite effectively, in
agriculture, infrastructure and education. Its debt
levels were moderate owing to its self-imposed
‘balanced budget’ rule. As the terms of trade fell
sharply, the authorities quickly depreciated the
currency while maintaining prudent fiscal policy.
The result was a very large and durable real
exchange rate depreciation. Major microeco-
nomic reforms were also introduced, freeing up
the business environment and triggering a rapid
supply-side response in manufactured exports.
Of the remaining two episodes, we examine the
special case of the AFC in more detail below.
This was the one major interruption to economic
growth in the last half-century, but at least the
economic recovery commenced quite quickly,
albeit at a permanently slower growth rate. The
© 2018 Economic Society of Australia
2018
20
15
Annual growth (%)
10
5
0
–5
–10
–15
–20
1960 1965
fourth episode was largely a ‘non-event’, in the
sense that during the global financial crisis
growth slipped by only 1.5 percentage points for
the year as a whole. Thus with the exception of
the AFC, Indonesia’s record of crisis response
and mitigation has been effective. This of course
is a key reason for its overall good economic
performance.
III The Sectors and Structural Change
The process of sectoral growth and structural
change has been a largely conventional one, with
a few significant exceptions. Before the onset of
rapid economic growth, more than half of
Indonesian GDP originated in the agriculture
sector and about 70 per cent of the workforce
was employed in it. Yet malnutrition was wide-
spread, affecting over half of the population. It
was economic growth that both shifted workers
out of agriculture and greatly improved nutri-
tional intake. Figures 3 and 4 show this process of
sectoral growth and structural change.
While agriculture shrank from almost one-half
to about one-tenth of the economy, it performed
strongly for most of the period. 9 Perhaps
ironically for an authoritarian and centralised
regime, the Suharto presidency took agriculture
seriously. Indonesia was a relatively late adopter
of the green revolution, but it caught up quickly.
9
Timmer (2015) and references cited therein to his
earlier work are the key literature.
© 2018 Economic Society of Australia
A SURVEY OF THE INDONESIAN ECONOMY 475
F IGURE 3
Growth of the Main Sectors, 1960–2016
Agriculture, value added (annual % growth)
Industry, value added (annual % growth)
Services, etc., value added (annual % growth)
1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Year
The 1970s oil boom provided resources for
investment in rural development. The country
was transformed from being the world’s largest
rice importer, in some years importing one-third
of the world’s traded rice, to being largely self-
sufficient by the mid-1980s. The tropical cash
crop sector was also rehabilitated, and the country
returned to being one of the world’s leading
producers of rubber, palm oil, and other crops
(Barlow, 1997), facilitated by proximity to
Malaysia, then the world leader in these crops.
Rapid economic growth also led to a major
industrial transformation (Hill, 1997). Shut off
from global markets, finance, technology and
supply chains, Indonesia had a very small
manufacturing sector in the mid-1960s. The
opening of the economy led to double-digit
manufacturing growth in almost every year from
1967 to 1996. This was initially a catch-up
process, followed by import substitution behind
trade barriers as industrial protection increased
in the 1970s. The falling terms of trade in the
1980s led to a major outward reorientation in
industrial policy, and for the first time Indonesia
became a significant industrial exporter. The
service sector also began to expand very
quickly, with massive government investments
in infrastructure and liberalisation in the finan-
cial, transport and other sectors.
The two exceptions to this otherwise conven-
tional story of structural change have been the
mining sector and the accelerated growth of
476 ECONOMIC RECORD DECEMBER

F IGURE 4
Structural Change, 1960–2016

Agriculture, value added (% of GDP, based on const. price)

Industry, value added (% of GDP, based on const. price)
Services, etc., value added (% of GDP, based on const. price)
60
Share of GDP (%)

50

40

30

20

10

0
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Year

services after 2000. Mining sector growth has
been shaped by trends in international commodity
prices. In turn, when prices have been high it has
had the familiar Dutch disease consequences of
squeezing profitability in the other tradeable
sectors, especially manufacturing, since agricul-
ture is only partly tradeable. Since 2000, as
Figures 3 and 4 show, apart from the commodity
boom, services have emerged as the principal
growth engine. Several factors explain this trend:
a second commodity boom that resulted in an
appreciation of the real exchange rate and thus
rendered manufacturing less competitive (Gar-
naut, 2015; see also Figure 9 below); an agricul-
tural policy that began to emphasise rents over
productivity (Patunru & Rahardja, 2015); liber-
alisation and technology change in key service
sectors such as transport, telecommunications and
business services; and the rise of China and its
effect of lowering the global price of semi- and
unskilled manufactures.
IV Investment and Savings
Rapid economic growth also created a virtuous
circle of rising savings and investment that in
turn sustained the growth. The anaemic growth
prior to 1966 was both a cause and a consequence
of low savings and investment. Higgins (1968)
speculated that Indonesia may have been running
down its capital stock for most of the period
since the 1920s. By the late 1950s, capital
inflows had dried up, as foreign property was
nationalised and Indonesia shunned all but a
small flow of aid from the Soviet bloc, most of
dubious value. Government investment also col-
lapsed.
The increase in both savings and investment
was dramatic, rising from approximately 5–10
per cent of GDP to over 20 per cent in just a
decade, and higher still subsequently (Figure 5).
Indonesia joined the rest of East Asia as a high-
savings economy. Increased economic security
and positive real interest rates provided house-
holds with an incentive to save. The government
returned the budget to balance and then, as the
1970s oil boom took hold, invested much of the
windfall gains productively. Foreign aid and
investment flowed in on a large scale. Reflecting
the attractiveness of the country as an investment
destination, Indonesia ran a persistent but mod-
erate current account deficit for most of the
Suharto era.
The AFC impacted significantly on these aggre-
gates. Investment from all sources (corporate,
government and foreign) fell sharply. Savings also
fell as households dissaved as a consumption-
smoothing strategy. The budget shifted from
balance to a moderately large deficit (see below).

© 2018 Economic Society of Australia

2018
Gross domestic savings (% of GDP)
Investment minus Savings (I-S) % of GDP (RHS)
40
35
Share of GDP (%)
30
25
20
15
10
5
0 –10
1970 1975
Foreigners were no longer willing to lend to the
country. As a result, the current account swung
from deficit to surplus. The depth and severity of
the crisis, and policy changes in its wake, are
illustrated by the persistence of the current
account surplus for over a decade after the AFC,
and the fact that it took savings and investment
levels almost a decade to return to those of the pre-
crisis period.
V International Trade
Commercial policy and international trade
have been central to Indonesia’s economic trans-
formation. It has benefited from its international
engagements even while much influential opinion
has been sceptical of the benefits of openness. 10
Indonesia’s increased economic openness,
along with that of most of its neighbours, is
revealed in its rising trade shares, that is,
merchandise trade relative to GDP (Figure 6).
The increase is particularly evident from the late
1960s as commercial channels were reopened
and many formal trade barriers were abolished.
It should be noted, however, that this measure of
openness provides only a partial picture of
Indonesian trade policy. First, the increase in
the late 1960s was as much the result of the
10
The best recent examination of Indonesia’s
ambivalence towards globalisation is Patunru et al.
(2018).
© 2018 Economic Society of Australia
A SURVEY OF THE INDONESIAN ECONOMY 477
F IGURE 5
Investment and Savings, 1960–2016
Gross capital formation (% of GDP)
10
Share of GDP (%)
5
0
–5
1980 1985 1990 1995 2000 2005 2010 2015
Year
regularisation of trade, as the incentives for
extensive smuggling were abolished and the
exchange rate was unified. 11 Second, Indonesian
trade values are affected by international com-
modity prices: higher prices translate into higher
trade values but they do not generally mean
more open trade policy. In fact, it is often the
reverse, in the sense that the forces of economic
nationalism, and hence more restrictive trade
policies, are in the ascendancy when the terms of
trade are high. Third, the sharp increase in trade
to GDP in the late 1990s reflects mainly the
collapse of the denominator (GDP measured in
US dollars). Along with the rising international
trade, foreign direct investment (FDI) has also
increased rapidly over this period, except during
11
Technical and physical smuggling of course
persists, albeit on a smaller scale. A general rule of
thumb is that when tariffs or their equivalent exceed
about 25 per cent, especially for high-value-to-weight
items, the smugglers are in business. Smuggling also
increases in periods of political turbulence where
there is a breakdown in central authority. For exam-
ple, in the late 1990s several district governments
along Sumatra’s east coast, which faces nearby
Malaysia and Singapore, were known to be operating
their own (illegal) import–export businesses. Owing to
the sensitivities involved, Singapore, one of Indone-
sia’s major trading partners and Southeast Asia’s
historic entrepot centre, still does not publish its trade
statistics with Indonesia.
478 ECONOMIC RECORD DECEMBER

F IGURE 6
International Trade Shares, Indonesia and Comparators, 1960–2016

China India Indonesia Philippines Thailand

160
Trade share of GDP (%) 140
120
100
80
60
40
20
0
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Year

F IGURE 7
FDI Shares, Indonesia and Comparators, 1960–2016

China India Indonesia Philippines Thailand

50
FDI stock share of GDP (%)

40

30

20

10

0
1980 1985 1990 1995 2000 2005 2010 2015
Year

the AFC when many investors exited the country
owing to political instability (Figure 7).
Trends in Indonesia’s export composition reflect
the interplay of changing comparative advantage,
policy reform, and international commodity prices
(Figures 8 and 9). Through to around 1970 the
country’s modest exports were mainly agricultural,
with rubber the most important (Figure 8). The 1970s
oil boom saw a dramatic transformation in export
patterns such that, by the end of the decade, Indonesia
was in some respects a ‘petroleum economy’, with
about three-quarters of its exports (and about two-
thirds of government revenue) originating from the
energy sector. Manufactures were minuscule. The
collapse in oil prices in the first half of the 1980s then
resulted in a spectacular rise in manufactured exports,

© 2018 Economic Society of Australia

2018
Share of total merchandise exports (%)
to the point that they became the country’s major
export group by the end of the decade. In effect, three
channels were at work: first, statistically, the effect of
lower oil prices on export values; second, the
declining terms of trade had reverse Dutch disease
effects, mainly through the real exchange rate (Fig-
ure 9); and third, economic difficulties triggered
sweeping commercial policy liberalisation that par-
ticularly freed up the manufacturing sector. Since the
AFC, the value of mineral and agricultural exports
has fluctuated in line with international prices, while
manufactured exports have lost their dynamism.12
In the twenty-first century Indonesia missed an
important commercial opportunity through the
government’s reluctance to adjust to the require-
ments of the most dynamic sector of East Asian
trade and industrialisation, the global production
networks. This segment, which now accounts for
over half of intra-East Asian and intra-ASEAN
trade, requires very open trade and investment
regimes and high-quality logistics. Neighbouring
countries, including late-comers such as Vietnam,
12
The second major commodity boom, running for
about a decade from around 2004, differed from the first
in several respects. The major export commodities during
the latter period were coal and palm oil. Combined with
Indonesia’s now democratised and decentralised gover-
nance, this had important implications for the manage-
ment of the boom and its distributional consequences.
© 2018 Economic Society of Australia
A SURVEY OF THE INDONESIAN ECONOMY 479
F IGURE 8
Indonesian Export Composition, 1970–2016
Agriculture, raw materials and food Fuel, ores and metals
Manufactures Miscellaneous
90
80
70
60
50
40
30
20
10
0
1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Year
have reformed more quickly in these areas
(Soejachmoen, 2012; Athukorala, 2014).
Formal investigations of Indonesia’s trade
regime illustrate its general openness together
with considerable inter-industry variations in
industry assistance, and continuing resort to
non-tariff barriers. During the Suharto era, heavy
industry, much of it state-owned, was the most
highly protected. While remnants of this protec-
tion remain, some agricultural sectors now also
receive high protection, including a ban on rice
imports for extended periods. 13
VI Macroeconomic Management and Economic
Crises
Since the late 1960s, macroeconomic manage-
ment has been an Indonesian success story, except
for the special and short-lived case of the AFC.
There has been a broad political consensus never to
return to the 1960s episodes of hyperinflation and
multiple exchange rates. Early in his presidency,
the technocrats persuaded Suharto to enact a
13
The most detailed studies of effective protection are
by Fane and Condon (1996) and Marks and Rahardja
(2012). Basri (2001) provides a comprehensive political
economy analysis of trade protection during the Suharto
era. Patunru and Rahardja (2015) examine the political
economy of trade policy during the democratic era. See
also Pangestu et al. (2015) for a general survey.
480 ECONOMIC RECORD
F IGURE 9
Terms of Trade and Real Exchange Rate
‘balanced budget’ rule, whereby the government
only spent the resources it had available to it from
domestic revenue and foreign aid.14 Bank Indonesia,
the central bank, was subservient to the Ministry of
Finance, but its goal of keeping inflation under control
was facilitated by fiscal prudence. Unusually in that
era, the international capital account was opened, both
to restore the country’s reputation in global capital
markets (on the premise that capital would be more
likely to enter the country if it was free to leave) and as a
check on government excesses.15 A fixed but adjus-
table exchange rate regime was adopted.16
These settings underpinned almost three decades of
rapid growth. They unravelled only during the AFC,
when a combination of circumstances resembling a
14
There was of course considerable – if poorly
documented – off-budget activity, mainly in the military
and later among the Suharto family business empires.
This had distributional rather than macroeconomic con-
sequences, in the sense that the activities did not cause
any significant increase in fiscal deficits. It is estimated
that perhaps only one-third of military revenue came
from official sources. Much of it came from illegal
exactions on firms and wealthy households (mainly the
ethnic Chinese community) and privileged licensing
facilities in lucrative areas such as forestry and trade.
15
The Mundell–Fleming ‘impossible trinity’ was not
violated, essentially because the domestic financial
sector was in its infancy, and the only capital flows of
any significance were government borrowings and FDI.
16
See Woo et al. (1994) for an examination of
Indonesian macroeconomic policy up to the early 1990s.
DECEMBER
‘perfect storm’ occurred.17 By the early 1990s mobile
capital to emerging markets in search of higher yields
was rising. Indonesia looked very attractive. It still had
an open capital account, and by the late 1980s it had
implemented far-reaching but incomplete financial
liberalisation, in both the banking sector and the stock
market. But its policy settings were problematic in two
respects: the fixed but adjustable exchange rate was not
able to adapt to the very large short-term capital
inflows then occurring,18 and financial sector supervi-
sion was lax.19 When Thailand got into difficulty in
17
There is a very large literature on Indonesia and the
AFC. See, for example, the four-monthly ‘Surveys of
Recent Developments’ in the Bulletin of Indonesian
Economic Studies, as well as McLeod and Garnaut (1998),
Corden (2007), Ito (2007) and Arndt and Hill (1999).
18
If allowed to float, the rupiah would almost certainly
have appreciated, as it did on each occasion that the narrow
official trading band was slightly widened. A reasonable
conjecture is that an appreciating rupiah would have slowed
growth and exerted a cautionary effect on unhedged foreign
currency borrowings, both responses in all likelihood moder-
ating the subsequent economic crisis.
19
Cole and Slade (1996) provide a detailed (and
largely positive) examination of Indonesia’s extensive
but incomplete financial liberalisation prior to the AFC.
On the connections between the financial sector and the
Suharto family business interests, see the insightful
study of the major Indonesian business conglomerate of
that era by Borsuk and Chng (2014). The key point to
note during the crisis was that both ‘crony’ and
‘professional’ banks collapsed.
© 2018 Economic Society of Australia
2018 A SURVEY OF THE INDONESIAN ECONOMY
July 1997, having exhausted its reserves in defence of
the baht, the contagion quickly spread around the
neighbourhood. Indonesia initially looked secure: its
fiscal position was sound and its current account
deficits were smaller than Thailand’s. However, its
financial institutions were highly vulnerable to the
sudden exodus of capital, exacerbated by several key
policy missteps as the crisis took hold. Crucially, the
formerly very close relations between the Suharto
government and the International Monetary Fund (and
hence the bilateral donors) deteriorated to the point of
being dysfunctional.20 The economy then went into
free-fall, the banking sector collapsed, and the
Rupiah–dollar exchange rate fell from Rp2,500 to at
one point Rp17,500. In May the following year the
seemingly impregnable Suharto presidency collapsed.
The country’s economic and political prospects then
appeared exceptionally gloomy.
In the event, Indonesia (and its most affected
neighbours, Malaysia and Thailand) came
through the crisis surprising quickly. Unexpect-
edly, a functioning democratic system was con-
structed amazingly quickly. The bank and
corporate bailouts were extremely costly, and
the new, weak democratic regime had to manage
a public debt equivalent to about 100 per cent of
GDP, an almost fourfold increase over the pre-
crisis figure. New macroeconomic rules had to be
developed quickly. Bank Indonesia was given
autonomy (a positive outcome from the IMF LOI)
and the floating exchange rate regime, which had
been forced on the government during the crisis,
was maintained. A Fiscal Law was enacted in
2003, with the government essentially adopting
the Maastricht principle of a maximum 3 per cent
(of GDP) fiscal deficit and 60 per cent public debt.
A major driver of the reform was that it facilitated
early exit from the deeply unpopular Fund con-
ditionality. Moreover, unlike the European
Union, Indonesia stuck to its budget rule. Fiscal
discipline, economic growth and (modest) asset
recoveries led to a successful fiscal consolidation,
20
A major early mistake was the sudden closure of
several banks without adequate explanation or support.
This triggered a generalised bank run. The various IMF
Letters of Intent (LOIs) were also highly onerous and
complex. Both the Suharto presidency and the IMF
made major mistakes over this period. Stiglitz (2002)
offers a stridently critical assessment of the IMF, while
Boughton’s (2012) (chapter 11) semi-official history
presents the IMF case. Blustein (2001) provides an
interesting and balanced assessment. See also the
references in footnote 17, especially Corden and Ito.
© 2018 Economic Society of Australia
481
with public debt falling to around 25 per cent of
GDP by the end of that decade.
Figures 9–11 show these macroeconomic out-
comes since the 1960s. They demonstrate how
quickly inflation was brought under control in the
late 1960s, and it has remained so ever since,
apart from a brief episode of near-hyperinflation
in 1998. Fiscal prudence is also evident, as shown
by the continuously modest deficits (Note also
that a negative figure for the deficit in Figure 11
indicates a surplus.). But the data also point to
substantial remaining challenges. For most of the
Suharto and early democratic periods, inflation
has arguably been too high. One result is the
constant need for nominal exchange rate depre-
ciations to restore competitiveness (Figure 9). As
Figure 10 shows, the rupiah–dollar exchange rate
has declined from about Rp400 to below
Rp12,000 over a 50-year period. The government
has also struggled to develop a revenue base that
is sufficient to provide the goods and services that
the community expects (Figure 11). This is
especially so in periods of low commodity prices,
such as at present when the tax effort is just 12
per cent of GDP and tax buoyancy is less than
unity. Persistent and misdirected subsidies, espe-
cially for petroleum products and electricity, have
further squeezed government capacity while
having perverse distributional consequences.21
VII Subnational Development Dynamics
Regional (subnational) development dynamics
and disparities matter more to Indonesia than
most countries. Historians emphasise that arbi-
trary lines drawn on an imperfect map centuries
ago in far-off European capitals in effect created
this ‘improbable’ nation. Indonesia shares a land
or maritime border with six countries. Histori-
cally, parts of the country were more closely
connected with these neighbours than with the
rest of the country. At several junctures through-
out its history the possibility of territorial
21
The electricity and fuel subsidies have been
particularly serious, occasionally reaching 4 per cent
of GDP, more than one-quarter of the total budget,
during periods of high international oil prices when the
domestic price has remained fixed. These subsidies
have declined during the current administration, but
there is no certainty that they will not return in a future
era of high commodity prices. They are not justified on
efficiency, equity or environmental grounds (Burke &
Kurniawati, 2018).
482 ECONOMIC RECORD DECEMBER

F IGURE 10
Inflation and Exchange Rate, 1960–2016

Inflation ≤ 25%, consumer prices (annual %)

Inflation > 25%, consumer prices (annual %)
Official exchange rate (LCU per US$, period average)
25

Exchange rate (IDR/USD)

14,000
Annual inflation rate (%)

20 12,000
10,000
15
8,000
10 6,000
4,000
5
2,000
0 0
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Year

F IGURE 11
The Budget, 1960–2016

Government Expenditure (% of GDP)

Government Revenue, excluding grants (% of GDP)
Government Expenditure – Revenue (excluding grants) (% of GDP)

Fiscal deficit share of GDP (%)

30 3
Share of GDP (%)

24 0

18 –3

12 –6

6 –9

0 –12
1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Year

disintegration was very real. Current policies
therefore need to be judged in this light. There are
also analytical issues related to domestic eco-
nomic integration and to subnational comparative
advantage reflecting the country’s unusual eco-
nomic geography: the central island of Java has
just 6 per cent of the country’s land area but about
60 per cent of its population and economic
activity, and a near-monopoly of national polit-
ical power.
Figure 12 provides a picture of the subnational
mosaic, presenting summary statistics for the

© 2018 Economic Society of Australia

2018 A SURVEY OF THE INDONESIAN ECONOMY 483

The Subnational Socioeconomic Mosaic

F IGURE 12

© 2018 Economic Society of Australia

484 ECONOMIC RECORD
country’s 33 provinces.22 Several features are
immediately apparent. First, per capita incomes
vary enormously, with the capital, Jakarta, and
some resource-rich regions (such as East Kali-
mantan) much richer, by a factor of about 6, than
the poorer regions, mainly located in eastern
Indonesia. If these data were presented for the
approximately 500 kabupaten, the district below
the provinces and to which the major public
sector resources flow, the differential is far
greater, about 50:1. Regional poverty incidence
and, to a lesser extent, other social indicators are
quite highly correlated with per capita income.
Even after adjusting for regional price differ-
ences, poverty is very low (less than 10 per cent)
in the richer provinces and stubbornly high
(above 20 per cent) in the poorer ones. 23
The result of these subnational differences is
quite pronounced regional comparative advantage
and specialisation. Indonesia has a single
exchange rate, uniform national laws 24 and few
restrictions on the movement of people and
goods. But as would be expected, most of the
manufacturing and higher-value services are
located on Java, with its larger markets, stronger
human capital base and proximity to national
government. Off-Java the major activities tend to
be agriculture and mining, and the manufacturing
processing and services that support them. As a
result of these large differences, almost every
22
In the democratic era the number of subnational
political units has increased quickly, reflecting both the
assertions of local identity and implicit fiscal incentives
to carve out new regions. The proliferation has been
particularly rapid at the kabupaten (sub-provincial)
level, with the number almost doubling. The number of
provinces has also increased, from 26 pre-crisis (i.e.
excluding the former East Timor province) to 34
currently. The statistical database used in the map does
not yet incorporate the newest province, North Kali-
mantan.
23
There are two caveats to the assertions in this
paragraph. First, there are a few regions, most notably
the two Papua provinces, where resource enclaves in
traditional subsistence economies are of such scale that
they result in high provincial income surrounded by
high levels of poverty. Thus non-mining regional GDP
per capita is a better measure of economic welfare in
these cases. Second, there are several poor regions
which, for a variety of social, cultural or historical
reasons, have above average educational achievement.
See the papers in Hill (2014) for an examination of
these issues.
24
Except for one province, Aceh, where Shariah Law
has now been introduced.
DECEMBER
major economic event and policy has subnational
consequences. For example, protection for man-
ufactures is implicitly a subsidy for Java. As we
have seen, the AFC led to a sharp exchange rate
depreciation which mainly benefited off-Java
agricultural exporters. The Dutch disease effects
of commodity booms benefited the commodity
exporters off-Java, but they also squeezed the
Java-based manufacturers.
At the time of independence Indonesia inher-
ited very large regional inequalities. As best as
can be measured (regional accounts have only
been produced since the mid-1970s) interregional
inequality has remained high but reasonably
stable ever since (Hill & Vidyattama, 2016).
Given the centrifugal economic forces at work, in
some respects this is a significant achievement,
especially as regional inequality is either higher
or has been increasing in the other developing
giants, such as Brazil, China and India (Milano-
vic, 2005). The stability reflects a combination of
factors, including central government subsidies
that have mildly favoured poor regions, migration
from poorer to richer regions (which, unlike in
China, is unrestricted), and the resulting remit-
tance flows.25
Paradoxically, at various times the fear of
territorial disintegration has been the motive for
both centralised and decentralised governance.
During the 1950s several regions attempted to
secede from Jakarta’s rule. Then the military-
backed Suharto regime, determined to keep a firm
grip on these regions, and other restless outlying
provinces such as Aceh, East Timor (as it then
was) and Papua, maintained tight centralised
control. However, in 1999, as East Timor voted
to secede and civil war intensified in Aceh, the
government rushed through a ‘big bang’ decen-
tralisation reform, allocating, mainly on an
unconditional basis, about one-third of the
national government’s budget to the kabupaten,
which were then also able to elect their own
governments. Because the decentralisation was so
hasty, and implemented by a national government
grappling with the biggest economic and political
crisis in the country’s history, the reform could be
judged either a success, as it achieved the aim of
maintaining the nation-state (Mietzner, 2014), or
25
Note that if inequality is measured at the kabu-
paten level, however, the picture is less reassuring.
Nevertheless, the data are available for a shorter time
span, and frequent border changes complicate empirical
estimations.
© 2018 Economic Society of Australia
2018
50
Poverty headcount ratio (%)
40
30
20
10
0 0
1975
one that had limited tangible effects other than
enriching local elites (Sjahrir et al., 2014). Cer-
tainly there is little evidence to date of improved
local service quality (Lewis, 2014).
Almost two decades after this major decentral-
isation program, Indonesia is still struggling to
develop an effective relationship between central
and subnational governments. The latter now
spend about one-third of the consolidated state
budget. But they have made little effort to raise
their own revenue sources, which account for
only about 15 per cent of their total receipts. The
result is a major vertical fiscal imbalance and
persistent shirking of expenditure responsibili-
ties. The relationship between the major subna-
tional tiers of government, particularly the
provinces and districts, also remains ill-defined.
These problems arise from incoherent incentives
and frequent modification of the regulations
governing fiscal flows and expenditure responsi-
bilities (Lewis & Smoke, 2017).
VIII Living Standards 26
Compared to the 1960s, the Indonesian people
are now very much better fed, educated and
housed, and they generally live longer, healthier
26
Thanks go to Anne Booth and Asep Suryahadi for
helpful discussions on some of the material in this
section.
© 2018 Economic Society of Australia
A SURVEY OF THE INDONESIAN ECONOMY 485
F IGURE 13
Poverty and Inequality, 1976–2017
Poverty, series 1 Poverty, series 2 Gini ratio
0.5
0.4
Gini ratio
0.3
0.2
0.1
1980 1985 1990 1995 2000 2005 2010 2015
Year
lives. These improvements have been over-
whelmingly driven by economic growth, but
there have also been beneficial policy interven-
tions, particularly in education. There have been
costs, too: Indonesia is now a more unequal
society (to regard this as a ‘negative’ is of course
a value judgement) and, as we shall see in the
following section, environmental amenities have
mostly deteriorated.
Poverty incidence has fallen quickly. The
‘headcount poverty’ measure, that is, the per-
centage of the population with estimated con-
sumption below some sort of acceptable
minimum standard, has fallen from about 60
per cent to 11 per cent. Figure 13 presents the
official estimates for the period 1976–2017.
Before 1976 the data are incomplete, but as
Booth (1992, pp. 342–5) and others have shown,
on the basis of fragmentary but plausible data,
there was a significant decline in poverty inci-
dence in the decade prior to the mid-1970s. Note
that the data in Figure 13 present two series, as
in 1996 the official poverty line was revised
upwards. As is evident in the data, changes in
poverty incidence follow economic growth quite
closely. Thus the one major interruption to the
process of declining poverty occurred during the
AFC, when poverty incidence rose by more than
one-third. Poverty then fell quite quickly as
growth resumed.
486 ECONOMIC RECORD
There is a very large literature on the interpre-
tation and quality of these data. 27 Collecting
accurate data from a sprawling archipelago with
inaccessible regions, pronounced seasonality in
incomes, a large informal sector, and still quite
extensive non-monetised activities is extremely
difficult. During the crisis years, the statistical
agency struggled to maintain the quality of its
field enumerations. The extremes of the distribu-
tion tend to be underestimated, hence probably
understating poverty incidence (and also inequal-
ity; see Leigh & van der Eng, 2009). Income is
recorded less accurately than expenditure, which
is a problem in measuring income inequality (and
for inter-country comparisons), but not for expen-
diture-based poverty estimates. The wealth data
are more limited still. Nevertheless, in spite of
these and other caveats, the overall conclusion of
a major decline in poverty incidence has not been
seriously questioned.
Having established this basic proposition, a
host of analytical, measurement and philosophi-
cal issues arises. We examine some of these
briefly.
First, poverty lines are arbitrary constructs, at
best based on some scientific estimate of mini-
mum human physical needs. Moreover, commu-
nity living standards change, especially over
several decades and in the presence of rapid
growth. A poverty line suited to the Indonesia of
the 1960s, characterised by very widespread
destitution, is not suitable for current-day mid-
dle-income Indonesia. This is clearly illustrated
in Figure 13, when the upward revision in the
official poverty line in 1996 increased the head-
count poverty estimate by 50 per cent.
Second, expenditures of most Indonesians are
clustered around the poverty line, that is, they are
either ‘near poor’ or ‘poor’. This has two impli-
cations. One is that the poverty estimates are
highly sensitive to the selection of the poverty
line. For example, according to World Bank
estimates (http://www.povertydata.worldbank.
org), in 2015 the headcount poverty estimate
was 6.8 per cent if using the ‘international
poverty line’ (of $1.90 per day, PPP) but 31.4
27
In addition to the material cited below, illustrative
of this literature, and also containing extensive cita-
tions to earlier research, see Suryahadi et al. (2011),
Priebe (2014), Sumner and Edward (2014), De Silva
and Sumarto (2014), Yusuf et al. (2014). See also
Ravallion (2016) for a comprehensive general overview
and analysis of the poverty literature.
DECEMBER
per cent if using the ‘lower middle-income
poverty line’ (of $3.20). Another implication is
that there is extensive churning in poverty inci-
dence. That is, relatively minor interruptions to
earnings or unexpected windfalls can push people
out of and into poverty. For example, using panel
data, Suryahadi et al. (2011, pp. 74–6) showed
that 53.3 per cent of the people who were classed
as poor in 2008 moved out of poverty in 2009,
while almost half the people who were classified
as poor in 2009 were not poor in 2008.
Third, at the household level the correlates of
high poverty incidence are well known (see the
references cited in footnote 27) and generally
accord with international experience. Key corre-
lates of higher poverty incidence include female-
headed households, the occupation and education
of household heads, the agriculture sector, and
the occurrence of unanticipated shocks such as
loss of earnings, serious health events and major
natural disasters. As would be expected also,
poverty is generally (but not always) significantly
higher in poorer regions, especially in eastern
Indonesia.
Fourth, with regard to policy, the data clearly
show that the most important poverty alleviation
strategy has been high growth. But the type of
growth also matters. Poverty appears to have been
more responsive to growth when it has been
labour-intensive, and therefore more inclusive,
for example in labour-intensive manufacturing
and most agriculture. Broad-based education
programs have been crucial, as discussed shortly.
In addition, since the AFC the government has
begun to introduce various targeted social safety
net (SSN) programs (World Bank, 2016). For
poor families, these have included subsidised
rice, scholarships, health insurance and condi-
tional cash transfers. The operation of these
schemes has been facilitated by the rapid
advances in mobile and internet banking, and
local-level democracy. The programs are still
modest in scale, effective monitoring is still
work-in-progress and targeting outcomes are
mixed. But program quality is improving and
lessons are being learned.
Education and health indicators have also
improved significantly over this period. This can
be seen by comparing changes in the same set of
socioeconomic indicators for the four countries
presented in Table 1 for around 1960 for 2015,
that is, approximately two generations later
(Table 2). The average Indonesian can now
expect to live more than two decades longer,
© 2018 Economic Society of Australia
2018 A SURVEY OF THE INDONESIAN ECONOMY
T ABLE 2
Comparative Socioeconomic Conditions, 2015
GDP per Years of
capita Trade schooling,
(constant (% of for those aged
Country 2010 $) GDP) 15 and above
Indonesia 3,834 41.9 7.6
Malaysia 10,878 134.2 10.4
Philippines 2,640 63.0 8.4
Thailand 5,775 126.8 8.0
adults receive six additional years of schooling
(resulting in near universal adult literacy), and
babies are now six times less likely to die before
their first birthday. Indonesia is closing in on its
middle-income neighbours even as they too have
achieved significant improvements in social indi-
cators. In its social outcomes, therefore, the
Indonesian record more closely resembles that
of East Asia than most of South Asia or the
Islamic world, where major education deficien-
cies persist.
So much for the positive story. Major gaps
remain. In education, Indonesia has prioritised
quantity over quality.28 Given the historical
backlog and the limited resources, choices had
to be made, and there is a strong case on both
equity and efficiency grounds for going universal
at lower quality over higher quality but incom-
plete coverage of the population. Most interna-
tional test comparisons (such as the TIMMS and
PISA) rank Indonesia poorly among middle-
income economies. A similar outcome is evident
in universities, where the emphasis on quantity
has resulted in no Indonesian universities being
ranked within the top 100 for Asia. Again,
history matters, as there were estimated to be
only 2,000 university graduates at the time of
independence.
The socioeconomic correlates of education
performance highlight another major challenge.
In 2009, 29 per cent of 16–18-year-olds in the
poorest household expenditure quintile were
enrolled in senior secondary school, compared
to 73 per cent in the richest quintile. Moreover,
the latter would almost always attend schools of
superior quality (Suharti, 2013, p. 215). There are
28
This paragraph draws substantially on Suryadarma
and Jones (2013), the most comprehensive recent study
of education in Indonesia.
© 2018 Economic Society of Australia
487
Years of Life Infant
schooling, expectancy mortality
for those aged at birth (deaths per
25 and above (years) 1,000 population)
7.3 68.6 25.0
9.8 74.5 6.8
8.2 68.0 23.2
7.3 74.1 11.2
also large regional differences in these percent-
ages, ranging from 48 to 87 per cent. One
important policy initiative was the 2002 consti-
tutional amendment that requires central and
local governments to spend at least 20 per cent
of their budgets on education. With the financial
resource constraint much reduced, the issue now
is to translate the funding into better educational
outcomes.
A similar story is evident in the health sector.
As noted, from a very low base health indicators
have improved rapidly. However, public policy
has not been as effective as in education. Public
health facilities have until recently been rudi-
mentary. Public spending has remained very
small, although it is now beginning to rise in
response to a constitutional mandate of 5 per cent
budget spending. Moreover, the spending has not
been directed as efficiently and equitably as in
education. The construction of modern hospitals
caters mainly to better-off urban residents, while
there has been underinvestment in rural and
preventative health facilities. The 5 per cent
budget requirement will at least remove some of
the financial limitations. However, supply-side
constraints (e.g. the training of medical person-
nel) and the political preferences for ‘high-end’
medicine will still need to be addressed. As with
primary and secondary education, much of the
delivery is now undertaken by the more than 500
local governments, resulting in spatial outcomes
that are highly variable.
There are also pronounced gender gaps in
education, the labour market and societal power
structures, although these gaps are smaller than
for practically any other Muslim country, and
Indonesia does not have the severe gender birth
disparities evident in China and India. There is
very little difference in literacy rates, while
enrolment rates at all levels of schooling across
488 ECONOMIC RECORD
genders are narrowing. In 2010 males received on
average about 0.8 more years of schooling than
females (Suharti, 2013, p. 18). At the tertiary
level, female enrolments now actually exceed
those of males, although in some conservative
regions restrictions on the ability of women to
access preferred education opportunities (and
later employment) remain substantial. Female
labour force participation has also risen since
the 1960s, especially during the manufacturing
export boom, when younger women were drawn
into the factories in large numbers. However, it
has stalled in recent years (Manning, 2014).
Although there are few formal barriers to
female workforce participation, the gender earn-
ings gap remains significant (Taniguchi & Tuwo,
2014). In the formal labour market, the hourly
female wage is 70–80 per cent that of males. In
the informal sector, reliable hourly data are
limited, but the gap is likely to be larger. The
usual set of factors appear to be the main
explanation: occupational specialisations, greater
mobility freedom for males, disrupted employ-
ment patterns for women owing to childbirth and
childrearing, and, as informal extended family
arrangements weaken, limited formal childcare
support facilities. Broader societal power struc-
tures reinforce these outcomes: only 20 per cent
of the members of the national parliament are
women, while only two women are in the 2016
Forbes list of the 50 richest Indonesians, ranked
at 37 and 44.
Inequality as measured by the Gini ratio was
broadly stable and relatively moderate during
much of the Suharto era (Figure 13). It was low
prior to the growth take-off in the late 1960s. The
colonial era plantation and mining enclaves had
been to a large extent nationalised, there was little
high-end service and manufacturing activity,
while most of the workforce was engaged in
smallholder rice and estate crop production, along
with petty trade and other informal service sector
activities. Then rapid growth, much of it rural and
agricultural, ensured a reasonably egalitarian
growth path for the ensuing 20 years. This
occurred in spite of both the centralised political
and patronage regimes and the 1970s oil boom.
The major 1980s fiscal adjustment to lower oil
prices also had little impact on inequality, as the
expenditure reductions were carefully targeted
(Thorbecke, 1991). In addition, the export-
oriented strategy was creating many new
employment opportunities in labour-intensive
industries.
DECEMBER
As Figure 13 shows, inequality began to
increase in the 1990s as major business conglom-
erates, many linked to the Suharto family,
extended their reach. By the mid-1990s, Indone-
sia had the highest level of business concentration
in the nine developing East Asian economies for
which Claessens et al. (2000) were able to
obtain data.29 Firm-level seller concentration
was also high (Bird, 1999). In addition, high-
income urban-based services were growing very
fast, especially following the 1988 financial
deregulation. The AFC reversed this trend sig-
nificantly, but temporarily. This was a crisis in
the modern urban economy. The rural economy
held up quite well, resulting in reverse migration
out of the cities. Export-oriented agriculture
benefited from the very large exchange rate
depreciation.
Inequality began to rise again quickly during
the democratic era, with the Gini ratio increasing
by almost 10 percentage points in the first decade
of the twenty-first century, one of the fastest
increases in developing Asia over this period
(Warr, 2015; World Bank, 2016). The database
does not permit a full decomposition and identi-
fication of the factors that explain this increase.
But the following are likely to have been the key
drivers. First, as noted below, the new labour
market regulations increased inequality in labour
market outcomes. Second, the public sector
commitment to egalitarian outcomes arguably
(and unintentionally) weakened, especially as
the earlier, equalising education expenditures
diminished in importance. The very small social
expenditures were insufficient to compensate for
these changes. Third, the changing economic
structure, particularly the growth of more capi-
tal-intensive services sectors, further increased
income disparities. Fourth, the second commodity
boom had more unequalising distributional
effects than the first one. The former was almost
entirely in the oil and gas sector, and the major
beneficiary, apart from foreign shareholders, was
the government, which, as noted, distributed the
proceeds reasonably equitably. In the second
boom, occurring approximately over the period
2004–14, the major beneficiaries were in the
private sector, particularly the owners of factors
29
They found that the top 10 corporations owned 58
per cent of corporate assets; six of the top 10 were very
closely connected to the Suharto family.
© 2018 Economic Society of Australia
2018 A SURVEY OF THE INDONESIAN ECONOMY
of production in coal and palm oil, the proceeds
of which were lightly taxed. 30 ,31
Accompanying this rapid socioeconomic devel-
opment have been profound demographic
changes (Jones, 2015). Indonesia, Java in partic-
ular, was seen as a ‘Malthusian’ candidate in the
1960s. A leading demographer of that era
described Java as ‘asphyxiating for want of land’
(Keyfitz, 1965). According to one gloomy but
influential study, population pressures were then
seen to be reinforcing the deep-seated poverty in
rural Java (Penny & Singarimbun, 1973). In fact,
Indonesia has subsequently managed its demo-
graphic challenges quite effectively. Fertility and
mortality rates have both fallen quickly, resulting
in a largely stable population growth rate. The
Suharto regime in particular pioneered an inno-
vative family planning program that overcame
resistance from conservative religious quarters
(Hugo et al., 1987). The result is that the country
continues to enjoy a demographic dividend, a
window that will remain open for about the next
15 years, providing sufficient time for govern-
ments to introduce pension, healthcare and other
reforms in preparation for an ageing society.
Labour market structure and conditions reflect
these various social, economic and demographic
changes, as mediated by the policy environ-
ment.32 The rapid structural change observed
above had its counterpart in the labour market,
with labour shifting out of low productivity
agriculture into industry and services. The labour
market transitions have occurred more slowly, as
is usually the case, resulting in widening inter-
sectoral labour productivities. The labour market
has also become more formal, more urban, older,
and (slowly) more female. Interregional labour
market mobility has also increased as transport
and communication facilities have improved.
People have moved to Java in search of better
30
See World Bank (2016) for discussion of some of
these issues.
31
In passing, comprehensive Indonesian wealth
statistics and the distribution of wealth are not yet
available. On the basis of fragmentary evidence, and as
part of its international wealth comparisons project,
Oxfam has estimated that the wealth holdings of the
four richest people in the country are similar to that of
the poorest 100 million citizens.
32
This and the following paragraph draw on the
extensive writings of Chris Manning; see, for example,
Manning (1998) on the Suharto era, Manning (2014) on
the democratic era, and the references cited therein.
© 2018 Economic Society of Australia
489
education and employment opportunities, and to
resource-rich regions off-Java in search of higher
wages. Unlike China and to a lesser extent India,
there are very few formal restrictions on labour
mobility.
Democracy ushered in a significant change in
the labour market policy regime. During the
Suharto era the labour market was relatively
unregulated and independent trade unions were
suppressed. But productivity and wages grew
quite strongly, especially during the 1980s when
the country embarked on a labour-intensive,
export-oriented industrialisation strategy. Over
this period, the Indonesian experience was similar
to that of its high-growth authoritarian neigh-
bours. The resultant labour market flexibility also
resulted in poverty falling less than might have
otherwise been the case during the AFC: the
labour market impacts fell more heavily on price
(i.e. real wages) than on quantity (i.e. employ-
ment). But democracy then led to increased
labour market populism. Although there was a
welcome increase in worker freedoms, regulated
minimum wages rose rapidly, and the government
introduced among the most onerous severance
pay requirements in developing Asia. The result
has been much slower formal sector employment
growth, increased dualism between the relatively
well-off and protected formal sector alongside the
vast informal sector characterised by low earn-
ings and insecurity, and increased wage inequal-
ity. It has also arguably delayed the transition
from Lewis-style surplus labour to the tighter
labour market conditions evident in Indonesia’s
more advanced neighbours. In addition, for the
first time Indonesia became a significant labour
exporter, with about 5 million people working
abroad, mostly in unskilled and semi-skilled
occupations. Mismatches between the labour
market and the education system have further
exacerbated these the problems of underemploy-
ment and wage inequality, with the rapid growth
of high school and diploma graduates, many of
indifferent quality and with employment expec-
tations that are unlikely to be realised.
IX Environmental Dimensions 33
Indonesia is a classic case of the environmen-
tal Kuznets curve in action. Like its neighbours,
over time most environmental indicators have
33
Thanks go to Paul Burke and Budy Resosudarmo
for helpful discussions on this section.
490 ECONOMIC RECORD DECEMBER

F IGURE 14
Deforestation and Emissions Indicators

Emissions Totals: Total GHG Emissions Including Land-Use Change

and Forestry (MtCO₂e)
Anthropogenic Suphur Dioxide Emissions: Total (Gigagrams of SO2)

Forest area (sq. km)

3,000 1,500,000
Emission measures

Forest area (sq. km)

2,500 1,250,000
2,000 1,000,000
1,500 750,000
1,000 500,000
500 250,000
0 0
1990 1995 2000 2005 2010 2015
Year

regressed (Resosudarmo, 2005). The factors
driving these outcomes are well known. Soci-
ety’s implicit preferences have been for rising
material welfare and for not fully internalising
the cost of the externalities created by environ-
mental degradation. Moreover, institutions are
not yet strong enough to represent the public
interest (including the global public interest) in
ameliorating the destruction of the ‘commons’,
including the nation’s vast forest and marine
resources, upland soil cover, and urban (and
peri-urban) air quality.
Environmental deterioration is everywhere evi-
dent, and some of it is already having a major
impact on everyday life in the country. The
capital city, Jakarta, is experiencing ever more
serious flood problems during the rainy season,
owing to rising sea levels and excessive pumping
of groundwater. By some estimates, as much as
one-quarter of the current city area will be subject

to regular and severe flooding by 2030 ( Alvarez & 34
Resosudarmo, 2017). Fatal mudslides are occur-
ring frequently, as a result of upland deforestation
and unchecked population settlements. The forest
cover is receding rapidly, often replaced by
mono-crop agriculture (much of it the booming
palm oil plantings), accompanied by unsustain-
able pesticide and fertiliser use. The loss of forest
cover is also leading to a loss of flora and fauna
species that are unique to Indonesia. 34 There is
rampant and largely unchecked overexploitation
of the country’s vast maritime resources, espe-
cially its fisheries. The fragile (and, for aquacul-
ture and tourism, potentially very productive)
marine ecology is also being destroyed. Urban air
quality is not as poor as that of the major Chinese
and Indian cities, but it too is deteriorating,
exacerbated by the almost complete reliance on
petrol-driven (and often poorly maintained) engi-
nes, in buses, cars and motorcycles. 35 Notwith-
standing the frequent flood events, paradoxically,
increasing areas of Indonesia are experiencing
‘water stress’ as a result of extended drought
periods, poor management of river systems, and
rapid, unplanned urbanisation.
These environmental problems have global
ramifications. During prolonged dry seasons
Indonesia’s forests rank first in the world in
endemic birds and mammals and sixth in endemic
amphibians (Alisjahbana & Busch, 2017, p. 122).
35
See, for example, Asian Development Bank
(ADB) (2010) and http://www.who.int/phe/health_to
pics/outdoorair/databases/cities/en/for some compara-
tive data. Among three major developing Asia capitals,
Beijing, Delhi and Jakarta, the latter is perhaps fortu-
nate owing to its immediate coastal location and the
absence of winter heating requirements.

© 2018 Economic Society of Australia

2018 A SURVEY OF THE INDONESIAN ECONOMY
caused by El Ni~no weather events, the cut-and-
burn agricultural technologies that have been
employed for millennia result in massive fires,
exacerbated further by the ongoing deep-level
burning of peat lands. Particularly during years of
prolonged dry seasons, Indonesia is actually the
world’s fourth largest CO 2 emitter. The palm oil
boom in recent decades has accelerated this
process. In these periods also, air quality becomes
so toxic as to be injurious to health. At its most
serious, it has disrupted civil aviation, including
in neighbouring Malaysia and Singapore. Indone-
sia is therefore a significant player in interna-
tional climate negotiations. Thus far it has
adopted the position common to most developing
countries, of being a willing participant in these
negotiations within the context of phased-in
global emission reductions, and seeking compen-
sation from richer countries in exchange for
improved forest management (Seymour & Busch,
2016). 36
Measuring these impacts is not easy. There are
few reliable long-term environmental statistics.
As an illustration, Figure 14 shows trends in three
key indicators: deforestation, CO 2 emissions and
sulphur emissions.
One approach to quantifying this environmen-
tal deterioration is by attempting to measure some
sort of ‘green national income’, that subtracts
from GDP an estimate of the depletion of non-
renewable mining and energy resources, the
degradation of otherwise renewable resources
such as forests and fisheries, and the loss of other
environmental amenities. An early set of esti-
mates prepared by Repetto et al. (1989) found
quite a large divergence between GDP growth and
an alternative ‘green growth’. However, the
estimates were highly sensitive to prices used,
the depletion rates, and the discovery of new
mineral resources.37
In the final analysis, improved environmental
outcomes will be achieved when community pref-
erences force policy-makers in that direction.
There is some slow progress. Mass rail transit in
36
Indonesia was an early signatory to REDD (http://
www.un-redd.org/), an international instrument to
incentivise emission reductions, through improved
forestry management. It signed a $1 billion LOI with
the Norwegian government in 2010, but implementation
progress to date has been very slow (Alisjahbana &
Busch, 2017, pp. 124–8).
37
See Nurkholis et al. (2007) for a more recent set of
tentative estimates.
© 2018 Economic Society of Australia
491
Jakarta and other major cities is belatedly under
construction. The rising incidence of catastrophic
environmental events is elevating the political
importance of the issues. Indonesia already has a
very active environmental movement, and the
democratic space to mobilise public support.
Satellite and other surveillance technologies
enable environmental activists to better monitor
forest loss and maritime exploitation. The gradual
phasing out of fossil fuel subsidies will encourage
greater fuel efficiency. At the margin, international
assistance can make a contribution. Sustainable
forest practices are gaining more support. Greater
environmental awareness in richer neighbours,
most importantly China and Japan, has potentially
powerful demonstration effects.
X Institutions and Governance
What role have institutions and systems of
governance played in these generally good devel-
opment outcomes? What light does the Indone-
sian evidence shed on the notion that ‘institutions
rule’ (Rodrik, 2003)? Is there any evidence that
they have led (or lagged) economic development?
How has the transition from authoritarian rule to
democracy impacted on processes and outcomes?
The first point to note is that Indonesia’s formal
institutions are generally weak and embryonic.
During the authoritarian era, political scientists
observed that ‘Suharto was the institution’
(Mackie & McIntyre, 1994; MacIntyre, 2003),
in the sense that he or his close associates were
the ultimate arbiters on any major policy issue or
commercial decision. During the democratic era,
political power has obviously been more diffused,
and decentralised, and therefore the formal insti-
tutions of the state have become more important.
But although the policy-making modalities have
differed, in both periods the country’s political
leadership has been more or less committed to
economic development, reinforced by the politi-
cal imperative to achieve rising living standards
that are broadly comparable to those in the high-
growth neighbours. The result has been, for most
of the five decades, a commitment to prudent
macroeconomic management, to moderately open
commercial policies, and to social progress.
Social scientists have struggled with attempts
to measure institutional quality, let alone estab-
lish causal relationships, especially in developing
countries. Table 3 presents some of the most
widely used comparative indicators of institu-
tional quality, for Indonesia, the two Asian
giants, and two middle-income neighbours. They
492 ECONOMIC RECORD DECEMBER

T ABLE 3
Comparative Institutional and Governance Indicators

Indicator China India Indonesia Philippines Thailand

CPI (rank/no. of countries)

1995 40/41 35/41 41/41 36/41 34/41
2016 79/176 79/176 90/176 101/176 101/176
LPI (rank/no. of countries)
2007 30/150 39/150 43/150 65/150 31/150
2016 27/160 35/160 63/160 71/160 45/160
EODB (rank/no. of countries)
2004 72/140 133/140 131/140 109/140 30/140
2018 78/190 100/190 72/190 113/190 26/190
WGI-GE (rank/no. of countries)
1996 105/184 85/184 141/184 102/184 73/184
2016 68/209 90/209 98/209 101/209 71/209
WGI-V&A (rank/no. of countries)
1996 177/201 73/201 160/201 84/201 80/201
2016 190/204 85/204 102/204 101/204 162/204

Source: Transparency International, Corruption Perceptions Index (CPI); World Bank, Logistics Performance Index (LPI); World
Bank, Ease of Doing Business (EODB); World Bank, World Governance Indicators – Voice and Accountability (WGI-V&A);
World Bank, World Governance Indicators – Government Effectiveness (WGI-GE).

are at best indicative, based mainly on subjective
survey material, and in some cases regarded as
somewhat political exercises, especially in the
case of the World Bank’s Ease of Doing Business
(EODB) results. These indicators are not avail-
able for the longer time period covered by this
survey. We therefore present them for the longest
time period available for each series. This gen-
erally coincides with Indonesia’s transition from
authoritarian to democratic rule, and so at least it
enables some sort of comparative assessment of
progress during the democratic era. The number
of countries surveyed varies over time, and so the
results are reported as ranks (x/y).
The major conclusion from these data is that
Indonesia’s ranking more or less corresponds to
that of its relative per capita income. That is, most
of the governance indicators suggest that institu-
tional quality has neither lagged nor led the
country’s economic development. But there is
considerable variation over time, which is broadly
indicative of the changes that have occurred during
Indonesia’s transition from authoritarian to demo-
cratic rule. For example, the Corruption Percep-
tions Index (CPI) data conclude that the country is
now considered to be relatively less corrupt than
was the case in the mid-1990s, although other
comparisons dispute this conclusion. Indonesia’s
logistics quality (Logistics Performance Index,
LPI) continues to lag most countries, and there has
been no improvement over the past decade. The
country is obviously now more democratic (World
Governance Indicators – Voice and Accountabil-
ity, WGI-V&A), although the political system is in
its infancy. Government effectiveness (WGI-GE)
has improved somewhat, but it is well behind
China, and also India. Controversially, the busi-
ness environment (EODB) has reportedly
improved, and is even ahead of China. Each of
these indicators, and many more like them, could of
course be unpacked in much greater detail. But
together they provide at least an indicative com-
posite picture.
These are highly aggregated measures. To gain
insights into how institutions have shaped policy
processes and development outcomes, it is useful
to briefly examine some case study material. We
select three important areas for illustrative pur-
poses.
The first area concerns policy reform. At a
general level, there are the common elements in
both periods that are widely discussed in the
literature (see, for example, Krueger, 2002): actual
or potential economic crises force or embolden
leaders to reform; the power of ‘good ideas’ may
triumph; and idiosyncratic factors including, in
particular, the occasional ascendancy of reform-
oriented personalities. But there is also a clear

© 2018 Economic Society of Australia

2018 A SURVEY OF THE INDONESIAN ECONOMY
distinction between the authoritarian and demo-
cratic eras in some policy reform processes. The
response to commodity boom and bust episodes
illustrates the differences.
During Indonesia’s two major commodity
booms of the last half-century, the forces of
economic nationalism intensified, resulting in
more inward-looking commercial policy and
greater direct government economic intervention
(in state enterprises, subsidised credit and so on).
As the boom faded, the response differed between
the periods. During the 1980s, the ‘technocrats’
(the key economy ministers, mostly from the
University of Indonesia) were able to persuade
the president that, without major reform, an
economic crisis was imminent. Suharto was in
‘supreme control’, and thus the key objective of
the technocrats was to persuade him of the case
for reform. Once achieved, reform proceeded
quickly and effectively. This was a case of ‘low
politics’, in the words of Soesastro (1989). 38 The
technocrats avoided ‘grand ideological debates’,
which they probably would have lost anyway.
Reform has been more difficult as the second
commodity boom ended, even with a somewhat
insulated technocracy. The legislature is now
considerably more powerful relative to the exec-
utive. Reformers have to engage in public debates
and win over (mostly sceptical) constituencies,
often proceeding in a succession of small steps.39
The result has been slow and often incoherent
reform (Patunru & Rahardja, 2015). The conclu-
sion from these two episodes is that reform is
therefore faster under authoritarian rule. But this
is not a generalisable proposition: unchecked
authoritarian rule brings with it the risk of major
policy mistakes, and in any case the reforms
could be undone if they were perceived to be
undertaken by a regime that is lacking in political
legitimacy, in the event that the regime is
overthrown.
The second issue concerns the persistence of
corruption, defined in the narrow sense of the use
of public office for private gain.40 As Table 3
shows, this has been a major challenge for
38
See also Azis (1994) and Basri and Hill (2004).
39
The ‘small steps’ argument is advanced in a
reflections piece by former finance minister Chatib
Basri (2017).
40
That is, we put aside the broader issue of ‘rent-
seeking’, defined in the Krueger (1974) sense of
lobbying for personal gain through protection, political
donations, infrastructure expenditures and so on.
© 2018 Economic Society of Australia
493
Indonesia throughout its history. Its form has
also changed in response to different governance
structures and policy regimes. For example, it
was heavily concentrated around Suharto and his
circle during the authoritarian era, 41 while during
the democratic era it has been more widely
diffused. Also, the 2001 decentralisation reform
transferred considerable administrative and finan-
cial authority to subnational governments, and
corrupt practices therefore similarly devolved,
particularly in resource-rich regions. Moreover,
the major 1980s trade reforms had cleaned up
much of the trade regime, especially through the
removal of (more corruption-prone) non-tariff
barriers, with the consequence that corruption
tended to migrate from the traded to the non-
traded sectors. Corruption has also tended to
proliferate during commodity booms since in
times of prosperity there is less pressure on
governments for accountability.
Nevertheless, while corruption has had adverse
distributional consequences, and it corrodes trust
in institutions, contrary to popular opinion
Indonesia’s economic progress over the past
half-century indicates that corruption has not
fundamentally held the country back. As Table 3
illustrates, Indonesia is not an outlier with regard
to corruption. The coexistence of rapid economic
development and corruption is of course a general
Asian story, and a feature of both authoritarian
and democratic regimes. The central explanation
for the coexistence is not that corruption is
conducive to growth 42 but that these regimes
have enough ‘growth-promoting’ policies to
overcome its negative effects.
The literature on corruption in Indonesia more
or less parallels that for other countries, taking
account of country-specific features of the type
noted above. 43 An analytical framework that
combines simple supply and demand analysis
41
McLeod (2011), for example, refers to the
‘Suharto franchise’ to draw attention to this concentra-
tion. Fisman (2001) creatively attempted to show the
value of the Suharto connection by examining relative
share price movements in Suharto-linked companies
during the AFC.
42
Although there is a literature, of some relevance to
Indonesia, arguing that it might be in some cases, in
overcoming stifling regulations, on the premise that (to
paraphrase Milton Friedman) ‘corruption is the intru-
sion of the market into a controlled economy’.
43
See Kis-Katos and Schulze (2013) for a compre-
hensive survey of corruption in Southeast Asia with
much emphasis on Indonesia.
494 ECONOMIC RECORD
with the country’s political economy framework
sheds light on the issue. The country’s complex
regulatory regime, evident in Table 3, both
empowers office-holders and incentivises firms
to short-circuit the system. A poorly remunerated
civil service with lifetime employment enhances
the likelihood that bureaucrats will be open to
bribery. Jobs in the more lucrative sections of the
civil service, such as customs and the police,
continue to be ‘purchased’. Institutions that are
designed to protect the public interest, including
the legal system, are still rather weak. Therefore
the likelihood of detection and the penalties for
malfeasance are generally low. Civic morality at
the top also matters – the more there are senior
officials and political leaders living lavish life-
styles on officially modest incomes, the greater
the temptation for their underlings to follow their
example.44
Democracy has had three major effects on
corruption in Indonesia. First, there are the
distributional effects noted above. Second,
Indonesia now has an independent anti-corruption
commission (known by its acronym, KPK) that
has the power to undertake both investigations
and prosecutions. It is the most popular institu-
tion in the country. While it has achieved much,
including the imprisonment of cabinet ministers
and heads of subnational governments, its work
has been hamstrung by limited resources and
frequent attempts by senior politicians to under-
mine its authority. 45 Third, while introducing at
least some public accountability, democracy has
also created greater uncertainty in government–
business relations. The proposition that ‘the only
thing worse than organized corruption is disor-
ganized corruption’ (Shleifer & Vishny, 1993) is
relevant to Indonesia’s commercial environment.
Under Suharto, and abstracting from moral con-
siderations, the rules of engagement were gener-
ally very clear: who to pay, how much, and for
what benefit. The greater political fluidity under
democracy has introduced considerable uncer-
tainty with regard to all three parameters. The
44
This analysis is of course greatly simplified.
There are many dedicated Indonesian civil servants
living on modest incomes. For an illustration of how
Indonesian bureaucrats may employ their monopoly
power for extortion, see Olken and Barron (2009). See
McLeod (2005) on the slow pace of bureaucratic
reform.
45
See Butt (2011) for an analysis of the work of the
KPK.
DECEMBER
consequence is that there is much greater uncer-
tainty in business, especially for long-term, cap-
ital-intensive projects such as infrastructure and
mining.46
A third example is the struggle to define a
workable division between state and market. That
is, what are the public and collective goods that
only a government can provide, or for which it
can establish a satisfactory framework in which
they can be provided, and what is best left to the
market? Corruption persists partly because
Indonesia’s legal system is still in its infancy. A
substantial proportion of the legal code still
originates from the colonial era and is written in
Dutch, a language now intelligible to less than 1
per cent of the population. Most firms, especially
larger foreign ones, try to avoid the legal system
anyway. 47 The level of criminality and the resort
to private security providers suggest that the
police force is unable to adequately discharge its
responsibilities (Wilson, 2015). The rapid envi-
ronmental degradation indicates that the ‘protec-
tion of the commons’ is still weak.
Yet in other areas there has been more interven-
tion than could be justified on efficiency or equity
grounds. Industry policy is a case in point.
Successive administrations in both the authoritar-
ian and democratic eras have attempted to imple-
ment selective industry policy (i.e. non-neutral
inter-industry and inter-firm incentives) through a
plethora of instruments: import protection, direc-
ted credit, a sizeable state enterprise sector, fiscal
incentives, regulatory provisions, preferential bid-
ding arrangements, and much else. As noted above,
Indonesia has industrialised rapidly, especially
during the Suharto era. But the overwhelming
evidence is that this dynamism has been the result
of the general factors that have driven growth in all
sectors. In fact, selectivity has either had no impact
or possibly a negative one. The sectors and firms
that have received government largesse have not
achieved higher rates of productivity growth,
superior export performance or exceeded any other
benchmark indicator (Hill, 1996). In fact, the most
detailed study of the political economy determi-
nants of inter-industry protection concluded that
‘Governments may not be very good at picking
46
See Wells and Ahmed (2007) for an authoritative
case study of the complexities of FDI in capital-intensive
projects in Indonesia before and after the AFC.
47
See Deinla (2017) for a comparative Southeast
Asian analysis.
© 2018 Economic Society of Australia
2018 A SURVEY OF THE INDONESIAN ECONOMY
winners, but losers are good at picking govern-
ments’ (Basri, 2001). 48
XI Conclusions
Over the past half-century Indonesia has been
one of the world’s major development success
stories. Facing highly unfavourable initial condi-
tions, and defying notions of path dependence, it
started to grow quickly from the late 1960s, and it
has by and large maintained this dynamism, apart
from the AFC years and albeit at a slower growth
rate in the twenty-first century. At the three key
junctures in its history, 1945, 1966, and 1998, the
overwhelmingly gloomy prognostications were
proved to be mistaken. Per capita incomes have
risen more than sixfold over 50 years, poverty has
fallen rapidly, and social indicators have
improved significantly. This achievement has
broader ramifications, for the 630 million people
of Southeast Asia, in which Indonesia is the
dominant power, and the Muslim world, where
there are relatively few cases of dynamic
economies and democratic polities.
Of course, numerous caveats need to be
attached to this conclusion. Although it warrants
inclusion in the high-growth club, it has not
grown as fast as some of its neighbours, notably
China and the Asian NIEs. Since the AFC it
appears to have transitioned to a lower equilib-
rium growth rate, one that is resulting in much
slower poverty eradication and job creation.
Inequality has risen appreciably this century,
compounding the social problems. In addition,
the government’s power to tax is insufficient to
provide the goods and services the community
expects, while most environmental indicators
have deteriorated, some perilously so. The coun-
try’s institutions are at an embryonic stage of
development and remain vulnerable to capture.
The analytical explanations for these outcomes
are largely conventional. That is, Indonesia began
to grow quickly once it opened up to the world
economy, adopted prudent macroeconomic poli-
cies, ended conflict, achieved political stability,
provided a workable business environment,
48
According to this research, industries in which
Suharto cronies were prominent were the most reliable
predictor of above-average import protection. There has
been no comparable study for the democratic era, but if
anything industry policy has become more popular
politically, and populist (Patunru & Rahardja, 2015).
Rock (2017) is the major academic dissenter to these
conclusions.
© 2018 Economic Society of Australia
495
invested in the supply side (especially education
and, earlier, infrastructure), and ensured reason-
ably broad-based social progress. 49 The interna-
tional environment has also been fairly benign for
most of the period. The economic take-off period
occurred during the Cold War era, when US
economic and diplomatic support was assured
and generous. This also coincided with Japan’s
period of exceptionally rapid growth, combined
with strong economic complementarities and
close diplomatic relations between the two coun-
tries. From the 1970s onwards, Indonesia was
surrounded by high-growth and mostly stable
neighbours. In the twenty-first century the China
boom led to huge windfall gains for Indonesia
through the effects on its terms of trade. To be
sure, volatility in international capital and com-
modity markets has posed significant challenges
for Indonesian policy-makers but, with the excep-
tion of the AFC, they have learnt to manage these
external shocks effectively.
Indonesia’s political economy has underpinned
and reinforced its economic dynamism. Suharto’s
controversial accession to power in 1966 created
the imperative for economic growth as his prin-
cipal source of political legitimacy. The revolu-
tion of rising expectations created during his
three-decade rule was a factor in the country’s
surprising swift navigation through its twin crises
of 1997–8. In the four elections since Suharto’s
demise the commitment to continued economic
growth has been adopted by all presidential
candidates. The bipartisan commitment to
macroeconomic prudence appears secure, and it
should be adequate to prevent a recurrence of
economic crises. Meanwhile, in this, one of the
world’s most ethnically and geographically
diverse nation states, democratic space and
local-level autonomy are able to peacefully
mediate most of the country’s social and political
tensions.
Finally, although it is a single-country obser-
vation, the Indonesian experience is also of
relevance in exploring the interrelationships
between democracy and economic development.
At the most general level, the literature suggests
that democracy is ‘good’ for development (Ace-
moglu et al., 2014), although there are plenty of
counter-examples, most especially from East
49
That is, the variables that are consistent with Sala-
i-Martin’s (1997) ‘two million regressions’. See Tem-
ple (2003) and Hill and Hill (2006) for attempts to
apply the growth econometrics literature to Indonesia.
496 ECONOMIC RECORD
Asia, and dissenters in the literature. Indonesia’s
experience indicates that growth has been lower,
inequality higher and corruption largely
unchanged during the democratic era. Of course,
democratic freedoms are inherently important,
and these inferences are far too casual for
rigorous empirical verification. There are also
much bigger issues than these, none more so than
the survival of the Indonesian nation state.
Can any causal inferences be drawn between the
transition to democracy and economic policy-
making, and hence economic outcomes? A number
have been suggested in this paper. First, macroe-
conomic rules have been instituted in both periods
which, while different, have had a similar effect in
‘insulating’ macro policy from political pressures.
But in other respects there are significant differ-
ences. For one thing, policy reform is slower,
owing to the larger number of ‘veto players’ in the
policy-making processes. But reforms, once
enacted, are more likely to be secure since they
have greater democratic legitimacy.
Second, policy time horizons are shorter, as
determined by political cycles. The most obvious
example is public sector infrastructure expendi-
ture, which as a percentage of GDP is now about
half that of the Suharto era, even allowing for the
time required to restore fiscal health in the wake
of the AFC (McCawley, 2015). Governments
appear to be reluctant to commit resources to
projects that will benefit future administrations,
while under decentralisation inter-jurisdictional
coordination has become more complicated.
Third, perhaps the most puzzling difference is
the sharp increase in inequality under democracy,
when the majority of the population, who are still
poor or near-poor, now have a voice, and
rudimentary SSNs have been instituted. As sur-
veyed above, this outcome does not appear to
have been the result of a deliberate strategy on the
part of leaders in the democratic era. Rather it has
been an accidental outcome of the interaction
between the labour market, the education system,
and commodity booms, an outcome that policy-
makers thus far have been unable to correct
through compensating fiscal and other measures.
REFERENCES
Acemoglu, D. and Robinson, J.A. (2012), Why Nations
Fail: The Origins of Power, Prosperity and Poverty.
Crown, New York, NY.
Acemoglu, D., Naidu, S., Restrepo, P. and Robinson,
J.A. (2014), ‘Democracy Does Cause Growth’,
DECEMBER
NBER Working Paper No. 20004. National Bureau
of Economic Research, Cambridge, MA.
Alisjahbana, A. and Busch, J. (2017), ‘Survey of Recent
Developments’, Bulletin of Indonesian Economic
Studies, 53 (2), 111–36.

Alvarez, J.C. and Resosudarmo, B.P. (2017), ‘The Cost
of Floods in Developing Countries’ Megacities: A
Hedonic Price Analysis of the Jakarta Housing
Market, Indonesia’, unpublished paper.
Arndt, H.W. (1984), The Indonesian Economy: Col-
lected Papers. Chopmen Publishers, Singapore.
Arndt, H.W. and Hill, H. (eds) (1999), Southeast Asia’s
Economic Crisis: Origins, Lessons, and the Way
Forward. Allen & Unwin, Sydney.
Asian Development Bank (ADB) (2010), Air Quality in
Asia: Status and Trends. ADB, Manila.
Aswicahyono, H.H. and Hill, H. (2016), ‘Is Indonesia
Trapped in the Middle?’, in Hutchison, F.E., Das,
S.B. (eds), Asia and the Middle-Income Trap. Rout-
ledge, London; 101–25.
Athukorala, P.C. (2014), ‘Global Production Sharing and
Trade Patterns in East Asia’, in Kaur, I., Singh, N. (eds),
Oxford Handbook of the Economics of the Pacific. Oxford
University Press, New York, NY; 333–61.
Azis, I.J. (1994), ‘Indonesia’, in Williamson, J. (ed),
The Political Economy of Policy Reform. Institute for
International Economics, Washington, DC; 385–416.
Barlow, C. (1997), ‘Growth, Structural Change and
Plantation Tree Crops: The Case of Rubber’, World
Development, 25, 1589–607.
Basri, M.C. (2001), ‘The Political Economy of Manu-
facturing Protection in Indonesia, 1975-1995’ (doc-
toral dissertation), Australian National University.
Basri, M.C. (2012), ‘Indonesia’s Role in the World
Economy: Sitting on the Fence’, in Reid, A. (ed),
Indonesia Rising: The Repositioning of Asia’s Third
Giant. Institute of Southeast Asian Studies, Singa-
pore; 28–48.
Basri, M.C. (2017), ‘Reform in an Imperfect World:
The Case of Indonesia’, Asian Pacific Economic
Literature, 31 (2), 3–18.
Basri, M.C. and Hill, H. (2004), ‘Ideas, Interests and
Oil Prices: The Political Economy of Trade Reform
during Soeharto’s Indonesia’, World Economy, 27
(5), 633–56.
Bevan, D., Collier, P. and Gunning, J.W. (1999), The
Political Economy of Poverty, Equity and Growth:
Nigeria and Indonesia. Oxford University Press,
New York, NY.
Bird, K. (1999), ‘Concentration in Indonesian Manu-
facturing, 1975-1993’, Bulletin of Indonesian Eco-
nomic Studies, 35 (1), 43–73.
Blustein, P. (2001), The Chastening. Public Affairs,
New York, NY.
Booth, A. (1992), ‘Income Distribution and Poverty’,
in Booth, A. (ed), The Oil Boom and After:
Indonesian Economic Policy and Performance in
the Soeharto Era. Oxford University Press, Singa-
pore; 323–62.
© 2018 Economic Society of Australia
2018 A SURVEY OF THE INDONESIAN ECONOMY
Booth, A. (1998), The Indonesian Economy in the
Nineteenth and Twentieth Centuries: A History of
Missed Opportunities. Macmillan, London.
Booth, A. (2016), Economic Change in Modern
Indonesia: Colonial and Post-Colonial Compar-
isons. Cambridge University Press, Cambridge.
Booth, A. and McCawley, P. (eds) (1981), The Indone-
sian Economy during the Soeharto Era. Oxford
University Press, Kuala Lumpur.
Borsuk, R. and Chng, N. (2014), Liem Sioe Liong’s Salim
Group: The Business Pillar of Suharto’s Indonesia.
Institute of Southeast Asian Studies, Singapore.
Boughton, J.M. (2012), Tearing down the Walls: The
International Monetary Fund, 1990–1999. Interna-
tional Monetary Fund, Washington, DC.
Burke, P. and Kurniawati, S. (2018), ‘Electricity
Subsidy Reform in Indonesia: Demand-Side Effects
on Electricity Use’, Working Papers in Trade and
Development No 2018/01, Arndt Corden Department
of Economics, Australian National University.
Butt, S. (2011), ‘Anti-Corruption Reform in Indonesia:
An Obituary?’, Bulletin of Indonesian Economic
Studies, 47 (3), 381–94.
Claessens, S., Djankov, S. and Lang, L.H.P. (2000), ‘The
Separation of Ownership and Control in East Asian
Corporations’, Journal of Financial Economics, 58,
81–112.
Cole, D.C. and Slade, B.F. (1996), Building a Modern
Financial System: The Indonesian Experience. Cam-
bridge University Press, Cambridge.
Corden, W.M. (2007), ‘The Asian Crisis: A Perspective
after Ten Years’, Asian Pacific Economic Literature,
21 (2), 1–12.
De Silva, I. and Sumarto, S. (2014), ‘Does Economic
Growth Really Benefit the Poor? Income Distribu-
tion Dynamics and Pro-Poor Growth in Indonesia’,
Bulletin of Indonesian Economic Studies, 50, 227–
42.
Deinla, I. (2017), The Development of the Rule of Law
in ASEAN: The State and Regional Integration.
Cambridge University Press, Cambridge.
Djojohadikusumo, S. (1984), ‘Recollections of My
Career’, Bulletin of Indonesian Economic Studies,
22 (3), 27–39.
Fane, G. and Condon, T. (1996), ‘Trade Reform in
Indonesia, 1987-1995’, Bulletin of Indonesian Eco-
nomic Studies, 32 (3), 33–54.
Fisman, R. (2001), ‘Estimating the Value of Political
Connections’, American Economic Review, 91 (4),
1095–102.
Garnaut, R. (2015), ‘Indonesia’s Resources Boom in
International Perspective: Policy Dilemmas and
Options for Continued Strong Growth’, Bulletin of
Indonesian Economic Studies, 51 (2), 189–212.
Gelb, A. and Associates (1987), Oil Windfalls: Bless-
ing or Curse?. Oxford University Press, New York,
NY.
Higgins, B. (1968), Economic Development, 2nd edn.
W.W. Norton, New York, NY.
© 2018 Economic Society of Australia
497
Hill, H. (1996), ‘Indonesia’s Industrial Policy and
Performance: “Orthodoxy” Vindicated’, Economic
Development and Cultural Change, 45 (1), 147–74.
Hill, H. (1997), Indonesia’s Industrial Transformation.
Institute of Southeast Asian Studies, Singapore.
Hill, H. (2000), The Indonesian Economy, 2nd edn.
Cambridge University Press, Cambridge.
Hill, H. (ed) (2014), Regional Dynamics in a Decen-
tralized Indonesia. Institute of Southeast Asian
Studies, Singapore.
Hill, H. and Hill, S. (2006), ‘Growth Econometrics in the
Tropics: What Insights for Southeast Asian Economic
Development?’, Singapore Economic Review, 51 (1),
313–44.
Hill, H. and Vidyattama, Y. (2016), ‘Regional Devel-
opment Dynamics in Indonesia before and after the
‘Big Bang’ Decentralization’, Singapore Economic
Review, 61 (1), 1–26.
Hugo, G.J., Hull, T.H., Hull, V.J. and Jones, G.W.
(1987), The Demographic Dimension in Indonesian
Development. Oxford University Press, Singapore.
Ito, T. (2007), ‘Asian Currency Crisis and the Interna-
tional Monetary Fund, 10 Years Later: Overview’,
Asian Economic Policy Review, 2 (1), 16–49.
Jones, G.W. (2015), ‘The Population of Southeast Asia’,
in Coxhead, I. (ed), Routledge Handbook of Southeast
Asian Economies. Routledge, London; 201–29.
Joshi, V. (2017), India’s Long Road: The Search for
Prosperity. Oxford University Press, Oxford.
Keyfitz, N. (1965), ‘Indonesia’s Population and the
European Industrial Revolution’, Asian Survey, 10,
503–14.
Kis-Katos, K. and Schulze, G.S. (2013), ‘Corruption in
Southeast Asia: A Survey of Recent Research’, Asian
Pacific Economic Literature, 27 (1), 79–109.
Krueger, A.O. (1974), ‘The Political Economy of the
Rent-Seeking Society’, American Economic Review,
64 (3), 291–303.
Krueger, A.O. (ed) (2002), Economic Policy Reform:
The Second Stage. University of Chicago Press,
Chicago, IL.
Krugman, P. (1994), ‘The Myth of Asia’s Miracle’,
Foreign Affairs, 73 (6), 62–78.
Legge, J.D. (2003), Sukarno: A Political Biography.
Archipelago, Singapore.
Leigh, A. and van der Eng, P. (2009), ‘Inequality in
Indonesia: What Can We Learn from Top Incomes?’,
Journal of Public Economics, 93 (1–2), 209–12.
Lewis, B. (2014), ‘Twelve Years of Fiscal Decentral-
ization: A Balance Sheet’, in Hill, H. (ed), Regional
Dynamics in a Decentralized Indonesia. Institute of
Southeast Asian Studies, Singapore; 135–55.
Lewis, B.D. and Smoke, P. (2017), ‘Intergovernmental
Fiscal Transfers and Local Incentives and Responses:
The Case of Indonesia’, Fiscal Studies, 38 (1), 111–
39.
Lindblad, J.T. (2015), ‘Foreign Direct Investment in
Indonesia: Fifty Years of Discourse’, Bulletin of
Indonesian Economic Studies, 51 (2), 217–37.
498 ECONOMIC RECORD
MacIntyre, A. (2003), The Power of Institutions:
Political Architecture and Governance. Cornell
University Press, Ithaca, NY.
Mackie, J.A.C. and McIntyre, A. (1994), ‘Politics’, in
Hill, H. (ed), Indonesia’s New Order: The Dynamics
of Socio-Economic Transformation. Allen & Unwin,
Sydney, NSW; 1–53.
Manning, C. (1998), Indonesian Labour in Transition:
An East Asian Success Story?. Cambridge University
Press, Cambridge.
Manning, C. (2014), ‘Labour Market Regulation and
Employment during the Yudhoyono Years in Indone-
sia’, in Athukorala, P.C., Patunru, A.A., Resosu-
darmo, B.P. (eds), Trade, Development and Political
Economy in East Asia. Institute of Southeast Asian
Studies, Singapore; 153–72.
Marks, S.V. and Rahardja, S. (2012), ‘Effective Rates
of Protection Revisited for Indonesia’, Bulletin of
Indonesian Economic Studies, 48 (1), 57–84.
McCawley, P. (2015), ‘Infrastructure Policy in Indone-
sia, 1965-2015: A Survey’, Bulletin of Indonesian
Economic Studies, 51 (2), 263–85.
McLeod, R. (2005), ‘The Struggle to Regain Effective
Government under Democracy in Indonesia’, Bulletin of
Indonesian Economic Studies, 41 (3), 367–86.
McLeod, R. (2011), ‘Institutionalized Public Sector
Corruption: A Legacy of the Suharto Franchise’, in
Aspinall, E., van Klinken, G. (eds), The State and
Illegality in Indonesia. KITLV Press, Leiden; 45–63.
McLeod, R.H. and Garnaut, R. (eds) (1998), East Asia
in Crisis: From Being a Miracle to Needing One?.
Routledge, London.
Mietzner, M. (2014), ‘Indonesia’s Decentralization:
The Rise of Local Identities and the Survival of the
Nation State’, in Hill, H. (ed), Regional Dynamics in
a Decentralized Indonesia. Institute of Southeast
Asian Studies, Singapore; 45–67.
Milanovic, B. (2005), ‘Half a World: Regional Inequal-
ity in Five Great Federations’, Journal of the Asia-
Pacific Economy, 10 (4), 408–45.
Mortimer, R. (ed) (1973), Showcase State: The Illusion
of Indonesia’s ‘Accelerated Modernization’. Angus
and Robertson, Sydney, NSW.
Naughton, B.J. (2006), The Chinese Economy: Transi-
tions and Growth. MIT Press, Cambridge, MA.
Nurkholis , Resosudarmo, B.P. and Hartono, D. (2007),
‘‘Internalisasi Faktor Degradasi LH dan Deplesi
SDA dalam Perhitungan PDB di Indonesia Tahun
2000-2005’ [Internalization of Environmental
Degradation and Natural Resource Depletion in the
Calculation of Indonesian GDP for 2000-2005]’,
Jurnal Ekonomi Lingkungan, 23, 1–22 (in Indone-
sian).
Olken, B.A. and Barron, P. (2009), ‘The Simple
Economics of Extortion: Evidence from Trucking
in Aceh’, Journal of Political Economy, 117 (3),
417–52.
Palmer, I. (1978), The Indonesian Economy since 1965.
Frank Cass, London.
DECEMBER
Pangestu, M., Rahardja, S. and Ing, L.Y. (2015), ‘Fifty
Years of Trade Policy in Indonesia: New World
Trade, Old Treatments’, Bulletin of Indonesian
Economic Studies, 51 (2), 239–61.
Patunru, A.A. and Rahardja, S. (2015), Trade Protec-
tionism in Indonesia: Bad Times and Bad Policy.
Lowy Institute, Sydney, NSW.
Patunru, A.A., Pangestu, M. and Basri, M.C. (eds)
(2018), Indonesia in the New World: Globalisation,
Nationalism and Sovereignty. Institute of Southeast
Asian Studies, Singapore.
Penny, D.H. and Singarimbun, M. (1973), Population
and Poverty in Rural Java: Some Economic Arith-
metic from Sriharjo. Cornell International Agricul-
tural Development Monograph 41, Cornell
University, Ithaca, NY.
Perkins, D.W. (2013), East Asian Development: Foun-
dations and Strategies. Harvard University Press,
Cambridge, MA.
Pisani, E. (2014), Indonesia etc.: Exploring the Improb-
able Nation. W.W. Norton, New York, NY.
Priebe, J. (2014), ‘Official Poverty Measurement in
Indonesia since 1984: A Methodological Review’,
Bulletin of Indonesian Economic Studies, 50, 185–205.
Ravallion, M. (2016), The Economics of Poverty:
History, Measurement and Policy. Oxford University
Press, Oxford.
Reid, A. (1993), Southeast Asia in the Age of Commerce,
1450–1680. Yale University Press, New Haven, CT.
Repetto, R., Magrath, W., Wells, M. and Rossini, F.
(1989), Wasting Assets: Natural Resources in the
National Income Accounts. World Resources Insti-
tute, Washington, DC.
Resosudarmo, B.P. (ed) (2005), The Politics and
Economics of Indonesia’s Natural Resources. Insti-
tute of Southeast Asian Studies, Singapore.
Rock, M.T. (2017), Dictators, Democrats, and Devel-
opment in Southeast Asia: Implications for the Rest.
Oxford University Press, New York, NY.
Rodrik, D. (ed) (2003), In Search of Prosperity:
Analytical Narratives on Economic Growth. Prince-
ton University Press, Princeton, NJ.
Sala-I-Martin, (1997), ‘I Just Ran Two Million Regres-
sions’, American Economic Review, 87(2), 178–83.
Seymour, F. and Busch, J. (2016), Why Forests? Why
Now?. Center for Global Development, Washington,
DC.
Shleifer, A. and Vishny, R.W. (1993), ‘Corruption’,
Quarterly Journal of Economics, 108 (3), 599–617.
Sjahrir, B.S., Kis-Katos, K. and Schulze, G.G. (2014),
‘Administrative Overspending in Indonesian Dis-
tricts: The Role of Local Politics’, World Develop-
ment, 59, 166–83.
Soejachmoen, M. (2012), ‘Why is Indonesia Left
Behind in Global Production Networks?’ (doctoral
dissertation), Australian National University.
Soesastro, M.H. (1989), ‘The Political Economy of
Deregulation in Indonesia’, Asian Survey, 29 (9),
853–68.
© 2018 Economic Society of Australia
2018 A SURVEY OF THE INDONESIAN ECONOMY
Stiglitz, J.E. (2002), Globalization and its Discontents.
W.W. Norton, New York, NY.
Suharti (2013), ‘Trends in Education in Indonesia’, in
Suryadarma, D., Jones, G.W. (eds), Education in
Indonesia. Institute of Southeast Asian Studies,
Singapore; 15–52.
Sumner, A. and Edward, P. (2014), ‘Assessing Poverty
Trends in Indonesia by International Poverty Lines’,
Bulletin of Indonesian Economic Studies, 50, 207–
25.
Suryadarma, D. and Jones, G.W. (eds) (2013), Educa-
tion in Indonesia. Institute of Southeast Asian
Studies, Singapore.
Suryahadi, A., Raya, U.R., Marbun, D. and Yumna, A.
(2011), ‘Accelerating Poverty and Vulnerability
Reduction: Trends, Opportunities and Constraints’,
in Manning, C., Sumarto, S. (eds), Employment,
Living Standards and Poverty in Contemporary
Indonesia. Institute of Southeast Asian Studies,
Singapore; 68–89.
Szirmai, A. (2015), The Dynamics of Socio-Economic
Development. Cambridge University Press, Cam-
bridge.
Taniguchi, K. and Tuwo, A. (2014), ‘New Evidence on
the Gender Wage Gap in Indonesia’, Economics
Working Papers no. 408, Asian Development Bank,
Manila.
Temple, J. (2003), ‘Growing into Trouble: Indonesia after
1966’, in Rodrik, D. (ed), In Search of Prosperity:
Analytical Narratives on Economic Growth. Princeton
University Press, Princeton, NJ; 152–84.
Thee K.W. (ed) (2003), Recollections: The Indonesian
Economy, 1950s–1990s. Institute of Southeast Asian
Studies, Singapore.
Thorbecke, E. (1991), ‘Adjustment, Growth and Income
Distribution in Indonesia’, World Development, 19
(11), 1595–614.
© 2018 Economic Society of Australia
499
Timmer, C.P. (2015), ‘The Dynamics of Agricultural
Development and Food Security in Southeast Asia:
Historical Continuity and Rapid Change’, in Cox-
head, I. (ed), Routledge Handbook of Southeast Asian
Economies. Routledge, London; 89–113.
Van der Eng, P. (2002), ‘Indonesia’s Growth Perfor-
mance in the Twentieth Century’, in Maddison, A.,
Prasada Rao, D.S., Shepherd, W.F. (eds), The Asian
Economies in the Twentieth Century. Edward Elgar,
Cheltenham; 143–79.
Warr, P. (2015), ‘The Drivers of Poverty Reduction’, in
Coxhead, I. (ed), Routledge Handbook of Southeast
Asian Economies. Routledge, London; 303–26.
Wells, L.T. and Ahmed, R. (2007), Making Foreign
Investment Safe: Property Rights and National
Sovereignty. Oxford University Press, Oxford.
Wilson, I. (2015), The Politics of Protection Rackets in
Post-New Order Indonesia: Coercive Capital,
Authority and Street Politics. Routledge, London.
Woo, W.T., Glassburner, B. and Nasution, A. (1994),
Macroeconomic Policies, Crises, and Long-Term
Growth in Indonesia, 1965–90. World Bank, Wash-
ington, DC.
World Bank (1993), The East Asian Miracle: Economic
Growth and Public Policy. World Bank, Washington,
DC.
World Bank (2008), The Growth Report: Strategies for
Sustained Growth and Inclusive Development. Com-
mission on Growth and Development, World Bank,
Washington, DC.
World Bank (2016), Indonesia’s Rising Divide: Why
Inequality is Rising, Why it Matters, and What Can
be Done. World Bank, Washington, DC.
Yusuf, A.A., Sumner, A. and Rum, I.A. (2014),
‘Twenty Years of Expenditure Inequality in Indone-
sia, 1993–2013’, Bulletin of Indonesian Economic
Studies, 50, 243–54.
 Bulletin of Indonesian Economic Studies

ISSN: 0007-4918 (Print) 1472-7234 (Online) Journal homepage: https://www.tandfonline.com/loi/cbie20

An Evaluation of Some Key Economic Policies

Ross H. McLeod & Sitta Rosdaniah

To cite this article: Ross H. McLeod & Sitta Rosdaniah (2018) An Evaluation of Some
Key Economic Policies, Bulletin of Indonesian Economic Studies, 54:3, 279-306, DOI:
10.1080/00074918.2018.1548245

To link to this article: https://doi.org/10.1080/00074918.2018.1548245

Published online: 11 Dec 2018.

Submit your article to this journal

Article views: 309

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=cbie20
 Bulletin of Indonesian Economic Studies, Vol. 54, No. 3, 2018: 279–306

Survey of Recent Developments

AN EVALUATION OF SOME KEY ECONOMIC POLICIES

Ross H. McLeod Sitta Rosdaniah

Australian National University Indonesian Ministry of State Owned Enterprises

For roughly the past three years, Indonesia’s growth has maintained its slight upward
trend while inflation has been broadly on target. The budget deficit has been kept
under control and government debt is low relative to that of many other countries.
Nevertheless, this sound macroeconomic performance is now under threat because the
government’s aversion to depreciation of the currency has led it to introduce unnecessary
contractionary fiscal measures—including cutbacks of spending on infrastructure on
which output growth depends. Along with many other currencies, the rupiah has
depreciated significantly throughout much of 2018—despite the attempts of Bank
Indonesia, the central bank, to prevent this—partly because of rapidly growing imports,
but also because of a reversal of portfolio capital inflow. This reversal reflects various
factors, including Bank Indonesia’s policy, until recently, of pushing interest rates
down while global interest rates have been rising; fear that rapid depreciation in poorly
managed economies such as Turkey and Argentina may prove contagious; market
recognition that expansion of the current account deficit is likely to require depreciation
beyond that desired by the central bank; and concern that the forthcoming presidential
election creates conditions in which it will be difficult to pursue sound economic policies.
This survey discusses the rather unexpected currency depreciation and associated
selling pressure in the stock exchange in detail before going on to provide evaluations
of, and suggestions in relation to, a range of microeconomic policies, including
infrastructure development; the provision of subsidised loans to small businesses;
the fiscal aspects of decentralisation; minimum wages; and management of the
bureaucracy. To a large extent, most of these policies can be characterised as reflecting
a distrust of market forces, including a reluctance to adopt market-mimicking processes
in the public sector—specifically, in relation to the funding of local governments,
and to the staffing of the bureaucracy. From another perspective, the adoption of a
wide variety of objectives for economic policy reflects the absence of any consistent
analytical framework, making it highly likely that policies will turn out to be mutually
inconsistent and, in particular, that they will conflict with the ultimate economic
objective of increasing living standards, especially those of the poor.

Keywords: crisis, eminent domain, small business, fiscal transfers, minimum wages,
bureaucracy
JEL classification: E58, F32, G01, H54, H77, J45

ISSN 0007-4918 print/ISSN 1472-7234 online/18/000279-306 © 2018 ANU Indonesia Project

http://dx.doi.org/10.1080/00074918.2018.1548245
280 Ross H. McLeod and Sitta Rosdaniah

INTRODUCTION
‘Indonesia, you’re amazing!’ So read the front page banner headline of the national
daily, Kompas, on 3 September, 2018, following the successful staging of the 18th
Asian Games over the previous fortnight. So great was the euphoria that the
president immediately announced Indonesia’s intention to bid for the Olympic
Games in 2032. A few months earlier, Bambang Brodjonegoro, the head of the
national planning agency, Bappenas, had painted a similarly optimistic picture
of Indonesia’s economic future. He said that he expected per capita income to
grow from $3,377 to $28,934 in the 30-year period from 2015 to 2045, provided
the economy could achieve an average GDP growth rate of 6.4% per annum.1
Growth at that rate was last seen, briefly, in 2011, so this was more aspiration than
prediction, but Bappenas clearly believes this is not out of the question, having
made similar claims at least as far back as 2011.2 Thus, it claims that Indonesia can
achieve ‘advanced country’ status by moving into the World Bank’s high-income
country classification by 2034.
These kinds of claims are typically accepted at face value, but it is instructive
to look at them carefully. The country classification is actually based on gross
national income (GNI), so the starting figure in 2015 was actually a little higher, at
$3,430 per capita. Assuming the same 6.4% growth rate for GNI, continuation of
recent population growth of around 1.1% per annum translates to 5.2% growth in
GNI per capita, implying that the latter would rise to just $15,886 in 2045—little
more than half the stated level—and that high-income status in fact would not be
achieved until about 2040. To achieve this status by 2034 would actually require
GNI per capita growth of 7.4% per annum, corresponding to GDP growth at about
8.6%—considerably higher than the rates achieved during the Soeharto years, let
alone the more modest rates of the last decade and a half. More realistically, if
Indonesia can achieve an average GDP growth rate equal to the recent level (5.3%
p.a.), it will take 32 years rather than 16 for it to become an advanced economy. To
achieve Bappenas’s imagined rate of 6.4%, much less the president’s target of 7%,
will require significant modification of a wide range of current policies.

AN EVALUATION OF SOME KEY CURRENT POLICIES

In this survey we therefore evaluate several key economic policies to see if they
support the aspiration to accelerate the growth rate and thus improve average
living standards as quickly as possible. In reality, there is a plethora of other
economic policy objectives: boosting exports, manufacturing, agriculture,
fisheries, education, health, small business, infrastructure, public housing,
development of outer and backward regions, and wages; stabilising the exchange
rate; reducing the current account deficit; preventing increases in interest rates,
and food, fuel, and electricity prices; reducing inequality and instability; and
so on. Most of these objectives reflect superficial beliefs that these things are
inherently desirable, or that the people involved in them are especially deserving,

1. Presentation by Bambang Brodjonegoro at Australian National University, 9 March 2018.

2. Bappenas ‘has outlined an economic master plan that aims to lift the country to high-
income status within 15 years’ (EIU 2011).
An Evaluation of Some Key Economic Policies 281

or that they are necessary if living standards are to improve. By contrast, there
is much less support for imports, ‘non-productive’ sectors (i.e., services such
as wholesale and retail trade), large-scale business, and higher prices. What is
lacking here is any coherent framework for deciding what constitutes sound
policy. The inevitable consequence is that policies are looked at in isolation
without considering potential conflicts between them.
The purpose of subjecting economic policies to critical scrutiny is to be able to
suggest changes to these policies—or perhaps to suggest their abandonment—if
they conflict with what we consider to be the ultimate economic objectives of
maximising the material well-being of the public and promoting the interests of
the poor in particular. Space limitations allow us to consider just a few of these
policies, each of which, no matter how well meant, seems to conflict with one or
both of these goals.
The proliferation of economic policy objectives (in Indonesia and elsewhere)
reflects a lack of debate as to the appropriate economic role of government. In
our view, basic neoclassical economic theory provides sensible guidelines on the
logical delineation of the respective roles of government and the private sector in
economic activity. It starts from the observation that the private sector is perfectly
capable of efficiently providing most of the goods and services desired by the
public (given the finite productive capacity of the economy). But it also shows that
there are specific cases of ‘market failure’ where it cannot do so—most importantly
in the provision of public goods such as national security, domestic law and order,
and protection against natural disasters, and when there are externalities involved
such as air and water pollution.3 We argue that these crucially important areas of
market failure are where governments should place their primary focus.
In short, we argue that the economic role of government is to complement the
private sector in cases of market failure rather than to usurp its role or unnecessarily
interfere with its activities. In reality, insufficient attention is paid to areas in which
the private sector cannot or will not generate socially optimal outcomes, while a
great deal of effort is expended in counterproductive interference with, or even
displacement of, the private sector. The low esteem in which the public holds the
entire legal system—from police officers to prosecutors and judges—is strong
evidence that this field has been badly neglected by successive governments.
The enormous damage and loss of life caused by frequent volcanic eruptions,
earthquakes, and tsunamis suggest that much more effort needs to be put into
improving preparedness for natural disasters. Ongoing problems with flooding in
key cities is another area in which increased government effort would seem very
likely to be socially beneficial. By contrast, a current concern is that in its haste to
expand infrastructure rapidly, the government has been appointing state-owned
construction companies directly rather than putting projects out to tender, thus
forgoing the opportunity to identify private sector firms willing and able to do
the work more cost-effectively. In general, it is not clear that anything is gained
by having state enterprises compete with private sector firms, or by having the
bureaucracy tell private firms how or what to produce.

3. Limited space precludes discussion here of market failure arising from natural monopolies.
282 Ross H. McLeod and Sitta Rosdaniah

The constitutional requirement (Art. 33 (2)) that branches of production that

serve basic needs are to be controlled by the state implies that the private sector
cannot be relied on to provide basic needs and that the government can be expected
to do better. Both theory and empirical evidence suggest strongly that private sector
outcomes will be superior, however, except in cases of market failure. The latter
are the branches of production in which government intervention can make an
important contribution. Other sectors supplying basic needs such as food, clothing,
and shelter are not obviously in need of much, if any, government intervention.
The continued focus of the national planning agency on economy-wide
planning epitomises the allocation of government resources to areas where they
are not needed; such planning is the antithesis of trusting markets to allocate
productive resources efficiently. The private sector, which accounts for about 90%
of all goods and services consumed domestically, will make its own decisions
about where to invest, regardless of what Bappenas sees in its crystal ball. The
agency, meanwhile, could best apply its analytical resources to the remaining
10% share purchased or produced by the government on behalf of the public—
for example, by undertaking cost-benefit analyses of potential infrastructure
improvements and extensions.

Macroeconomic developments and policies

We begin by looking at the conduct of macroeconomic policy. The deceleration of
GDP growth evident in the last years of the second Susilo Bambang Yudhoyono
administration has been halted (table 1), but there seems little reason to expect
annual growth much above the current (Q3 2018) 5.2% per annum in the near
future. In particular, although investment spending grew quite rapidly from mid-
2017 through the first quarter of 2018, its average growth rate was noticeably lower
in the next two quarters—albeit still exceeding that of GDP by some margin. The
three key tradables sectors—manufacturing, agriculture, and mining—continue
to be outperformed by non-tradables (mainly services). Inflation has remained on
target at low levels (3%–4% p.a.) since the one-off effect of the jump in domestic
fuel prices in November 2014 is no longer a factor.
The budget deficit grew quite rapidly during the second Yudhoyono
administration—both absolutely and relative to GDP—but under the Joko
Widodo (Jokowi) administration the ratio to GDP has stabilised at around 2.5%,
which is within the legal maximum (3% of GDP) (table 2). Central government
debt outstanding began to accelerate noticeably from mid-2015—especially the
rupiah-denominated part, which now comprises almost 60% of the total—but
even though the ratio of public sector debt to annual GDP has been rising steadily
over the past four years, it remains low by international standards, at about 29%.
On the other hand, from January to October 2018, the rupiah depreciated by
about 14% against the US dollar, while share prices declined by about 12% (figure
1). The current account deficit (CAD) grew to 3.4% of GDP in the third quarter
of 2018, while international reserves held by the central bank fell from $132
billion to $115 billion between January and October. Somewhat surprisingly, and
despite repeated attempts by government spokespeople to assure the public that
the economic fundamentals are currently sound, many commentators prefer to
emphasise Indonesia’s purported vulnerability to destabilisation—if not crisis—
as evidenced by this second group of indicators.
An Evaluation of Some Key Economic Policies 283

TABLE 1  GDP Growth (% p.a. year on year)

Jun-17 Sep-17 Dec-17 Mar-18 Jun-18 Sep-18

GDP 5.0 5.1 5.2 5.1 5.3 5.2

By expenditure
Household consumption 4.9 4.9 5.0 4.9 5.1 5.0
Government consumption –1.9 3.5 3.8 2.7 5.2 6.3
Investment 5.3 7.1 7.3 7.9 5.9 7.0
Exports 2.8 17.0 8.5 6.1 7.6 7.5
Imports 0.2 15.5 11.8 12.8 15.3 14.1
By sector
Manufacturing 3.5 4.8 4.5 4.6 3.8 4.3
Agric., forestry, and fisheries 3.2 2.8 2.2 3.3 4.8 3.6
Mining and quarrying 2.1 1.8 0.1 0.9 2.6 2.7
Wholesale and retail tradeᵃ 3.5 5.2 4.5 5.0 5.2 5.3
Accommodation and restaurants 5.7 5.7 5.5 5.5 5.7 5.9
Information and communication 11.1 8.8 9.0 8.5 5.9 9.0
Construction 6.9 7.0 7.2 7.4 5.7 5.8
Transportation and storage 8.8 8.9 8.2 8.6 8.7 5.6
Servicesᵇ 6.1 6.5 5.5 5.3 5.0 5.4
Governmentc 1.3 2.9 6.4 5.4 6.3 7.3
Utilitiesᵈ –2.1 4.9 2.5 3.3 7.3 5.6

Source: BPS (Statistics Indonesia) through CEIC.

Note: The sectors named above include the following sub-sectors: a repair of motor vehicles; ᵇ financial
and insurance activity; real estate; business services; other services; c public administration, defence
and social security; education services; health care and social work; ᵈ electricity, gas, and water supply;
sewerage, waste, and recycling.

Figure 1 shows a period of more than a decade, allowing us to put current trends
into historical perspective. The recent depreciation is milder than that during the
18-month period to September 2015 and very modest by comparison with the rapid
depreciations associated with the global financial crisis (commencing in Q3 2008)
and the ‘taper tantrum’ (Q3 2013). Similarly, the share price declines in 2008 and
2015 were more severe than in 2018. In short, when seen in historical perspective,
concerns about the recent deterioration in these financial indicators appear to be
unduly alarmist.

Explaining the Perturbations

There have been several contributors to the depreciation of the rupiah and the
decline of share prices. First, years of expansionary monetary policy in the US after
the 2008 global financial crisis (GFC) ended in January 2015, causing interest rates in
this key financial market to move slowly but steadily upward. Bank Indonesia (BI)
chose not merely to ignore this new trend for more than three years but to reduce
its own policy rates by even more than its US counterparts raised theirs, resulting
284 Ross H. McLeod and Sitta Rosdaniah

TABLE 2  Budget Deficit and Government Debt

Deficit (Rp trillion) Deficit : GDP (%) Debt : GDP (%)

2010 46.8 0.7 23.7

2011 84.4 1.1 22.3
2012 153.3 1.8 22.7
2013 211.7 2.2 23.0
2014 226.7 2.1 23.8
2015 298.5 2.6 26.4
2016 308.3 2.5 26.7
2017 341.0 2.5 28.6

Source: BPK (Audit Agency) and BPS through CEIC.

in a significant reduction in Indonesia’s relative attractiveness as an investment

destination. The narrowing gap between rupiah and dollar interest rates eventually
began to reduce net foreign portfolio investment, and the consequent reduction
in total savings available for investment began to push domestic market-based
interest rates upward.
The central bank was slow and reluctant to respond, belatedly increasing its
policy rates—by considerably smaller amounts—from around May 2018 (figure 2).4
By this time, BI was intervening as a seller in the foreign exchange market, shedding
international reserves in an attempt to limit depreciation of the rupiah. The effect
of this was to reduce liquidity and thus keep market interest rates higher than BI’s
policy rates, but the central bank chose to offset this effect, partially at least, by
intervening as a buyer in the market for government bonds (Wall Street Journal, 5
September 2018). Unwillingness to allow domestic interest rates to move in line
with the global financial markets worked against restoration of portfolio capital
inflows, which fell from $8.1 billion in the second quarter of 2017 to –$1.1 billion
in the first quarter of 2018, ensuring continued pressure on the exchange rate.
Second, a few developing countries (Turkey, Argentina, and South Africa, in
particular) have been experiencing rapid currency depreciation (largely because of
poor macroeconomic management). This has revived painful memories of the Asian
financial crisis (AFC), which was triggered by a large, unexpected, and ultimately
contagious depreciation of the Thai baht. Presumably some of the recent capital
outflow reflects herd-mentality behaviour based on a fear that currency weakness
in these other developing countries could spread to Indonesia, notwithstanding
its lack of close financial links with them. Awareness, on the part of sophisticated
investors, of the central bank’s demonstrated reluctance to allow interest rates to
rise to levels consistent with exchange-rate stability would only have compounded
this behaviour.

4. The lack of close correlation between policy rates and market rates suggests that monetary
conditions are in fact mainly determined by BI’s transactions in the foreign exchange and
bond markets, and by cash flows through the government’s account at BI; policy rates do
not accurately reflect the stance of monetary policy.
An Evaluation of Some Key Economic Policies 285

FIGURE 1  Exchange rate and Composite Share Price Index

Rp/$ Share prices

16,000 8,000

14,000 7,000
Exchange rate
12,000 6,000

10,000 5,000

8,000 4,000
Share price index

6,000 3,000

4,000 2,000

2,000 1,000

0 0
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Sources: BI and Indonesian Stock Exchange through CEIC.

Third, non-oil and gas imports have been growing rapidly since late 2016 and
by much more than exports, suggesting the need for depreciation in the absence
of any reason to expect a change in this trend. Moreover, there is a move to greater
protectionism in the global trading environment—to which Indonesia has been
making its own contribution (Patunru and Rahardja 2015)—and, indeed, a threat
of trade wars between the US and some of its major trading partners, leading to
an anticipation of disruption of Indonesia’s export markets and possible further
increases in the CAD. Such concerns also tend to generate speculative capital outflows.
Finally, there are concerns about the potentially destabilising impact of the 2019
presidential election campaign. There is a vicious circle at work here, with the
president’s opponents playing on fears of rising food prices and emphasising the
return of the exchange rate to levels last seen during the disastrous AFC—even
though the recent rate of depreciation has been quite modest, as we have just noted.
Such tactics help to create a crisis mentality that itself encourages capital outflow,
with the potential to make fear of depreciation a self-fulfilling prophecy.

Policy Responses
The central bank has sold off more than 10% of its international reserves in order
to limit the extent of depreciation. For its part, the government has focused mainly
on trying to reduce the CAD through fiscal policy, most significantly by delaying
billions of dollars of infrastructure projects with a heavy import content, mainly in
286 Ross H. McLeod and Sitta Rosdaniah

FIGURE 2  Money Market and Policy Interest Rates (% p.a.)

10
Call money: 1 month
9 BI lending
facility
8

4
BI deposit facility
3

0
Jun-15 Dec-15 Jun-16 Dec-16 Jun-17 Dec-17 Jun-18

Source: BI through CEIC Asia Database.

the electricity sector. The worrying implication is that it is now consciously sacrificing
timely completion of its ambitious electricity-generation expansion program and the
boost to economic growth this could have provided. The public may find it hard to
understand this policy reversal in a context in which government spokespeople have
been claiming that the economic fundamentals are sound.
On the revenue side of the budget, the government has made much of its newly
imposed increases in ‘taxes on imports’, but these are unlikely to have a significant
impact. The increases apply only to a narrow range of ‘non-essential’ consumer goods,
and consumer goods in total account for only a small proportion of total imports, in
any case. Moreover, the taxes are not customs duties but are withholding taxes paid
by importers; since these payments will be credited against their future income tax
liabilities, little more will be achieved than to modestly and temporarily interrupt
importers’ cash flow. In addition, the government has announced its intention to
require greater substitution of biodiesel for traditional fuel, thus reducing the need to
import corresponding amounts of the latter, but the logistical difficulties of achieving
much progress on this front in the short run appear to be considerable.5

5. Basri and Patunru (2006, 307–8) provide an extended discussion of arguments against
the production of ‘bio fuels’.
An Evaluation of Some Key Economic Policies 287

Various other possibilities for reducing capital outflow have also been discussed,
such as introducing measures to induce exporters to keep half their earnings in
the country and convert them to rupiah—ignoring the fact that businesses do not
generate revenue in order to accumulate money in the bank. For the most part, these
suggestions seem more likely to strengthen the crisis mentality and thus exacerbate
capital outflow than to turn it around. To the extent that various mooted tax breaks
will be available only to exporters, and perhaps only to exports of particular products
or in certain minimum volumes, the result would be akin to re-imposing the multiple
exchange rate system Indonesia wisely abandoned decades ago (Arndt 1966, 8).

Growth Versus Stability?

Of particular concern is the notion that there is a trade-off between stability and
growth (Bank Danamon 2018; Basri 2016 and 2018b; Rajah 2018; Juoro 2018), and that
in view of these threats to stability it is necessary to sacrifice growth to some extent.
By contrast, we argue that there is no such trade-off—provided the policy response
is appropriate—and that there is therefore a danger that Indonesia will unnecessarily
settle for slower growth than it could otherwise achieve.
How, then, can policymakers respond to increases in the CAD and/or reductions in
capital inflow without endangering stability? Our advice is that, aside from ensuring
‘good communication with business communities’ (Basri 2018a, 27) and the general
public, they should continue to maintain slow and steady growth of base money (at
a rate of around 10% p.a.) as the nominal anchor for monetary policy, avoiding the
temptation to try to control other monetary variables—in particular, the exchange
rate and interest rates. In short, they should trust market processes to adjust these
key prices to changing circumstances rather than try to prevent them from doing so.
If indeed foreign and domestic investors collectively decide to devote less of their
savings to Indonesia, a matching reduction in the CAD will be required. In turn,
this can be achieved without government intervention simply by allowing market
forces to find a new exchange rate that balances the supply of, and demand for,
foreign exchange. This will involve some depreciation of the rupiah, making it more
profitable to produce tradables and more expensive to import them. A new, smaller
CAD will therefore emerge, balancing the new, smaller capital inflow. In addition,
the reduction of total savings available to finance investment will result in higher
domestic interest rates that reflect the now higher rates in global financial markets.
The policy recommended here roughly matches that followed during the GFC,
when the authorities largely trusted the financial markets, allowing exchange rates
and interest rates to move freely. Indonesia weathered that storm easily, with only a
slight and short-lived deceleration in growth (Kuncoro et al. 2009). Notwithstanding
this experience, the currently dominant view in Jakarta presumes severe consequences
for the economy if the exchange rate is allowed to depreciate freely and if strong
measures are not taken to control the CAD. A key proponent of this view is former
finance minister Chatib Basri. In his discussion of Indonesia’s response to the 2013
taper tantrum, he argues that the central bank

could not use the exchange rate alone to solve the problem of external imbalances. The
memory of the trauma from the AFC led to worries of the rupiah falling too steeply
because many feared a repeat of the AFC (Basri 2018a, 26).
288 Ross H. McLeod and Sitta Rosdaniah

The underlying argument here appears to be that during the AFC the
authorities mistakenly allowed the exchange rate to be market-determined.
The consequences were disastrous: an astonishingly large depreciation of the
rupiah, very high inflation and interest rates, the collapse of the banking sector
at enormous cost to the public, and strongly negative growth. The desire to avoid
a repeat is perfectly understandable, but to attribute these dire consequences to
the decision to float the currency is facile.
Prior to the AFC, BI’s nominal anchor for monetary policy was a de facto
commitment to a steady rate of depreciation of the rupiah (Basri 2018a, 26). The
decision to float the currency in August 1997 was sound, but no replacement
anchor was adopted, so monetary policy became incoherent during the following
8–10 months. The consequence, as Basri (citing an IMF study) points out, was
‘a loss of monetary control, which … caused further drops in the rupiah’ (2018a,
25). To be precise, base money doubled in eight months, resulting in surging
prices and a loss of competitiveness that could only be offset by depreciation.
The expectation of depreciation generated rapid capital outflow, making the
expectation self-fulfilling.
The loss of monetary control was a consequence of BI’s response to opportunistic
behaviour on the part of the banks, which rapidly increased their lending to
favoured entities wanting to pay down dollar-denominated debt or to purchase
dollar-denominated assets (McLeod 2014, 222). When the recipients of these
loans used the proceeds to purchase dollars—ultimately from BI—the banks’
deposits at BI fell correspondingly, rendering the banks illiquid and in need
of central bank support if they were to continue to operate. BI’s willingness to
restore lost liquidity amounted to financing speculation against its own currency.
If it had stayed out of the foreign exchange market there would have been no
loss of liquidity from the banking system, only a reallocation of reserves among
banks as their customers bought and sold foreign exchange among themselves. It
would have been a simple matter to identify banks that became illiquid because
of financing currency speculation, and to bring this activity under control using
BI’s authority as prudential regulator. To the extent it was judged necessary to
restore the liquidity of such banks, BI could have borrowed the funds from other
banks with liquidity surpluses, thus avoiding any impact on base money.
In short, the severity of the AFC for Indonesia was a consequence of policy
responses on the part of the authorities. The implication is that current external
imbalances can be dealt with by allowing the markets to determine exchange rates
and interest rates, provided base money growth is kept under control. Moreover,
there is no need to change fiscal settings to reduce the CAD. The supposed need
to sacrifice growth for stability is illusory.

Infrastructure
The Jokowi administration set out a long list of strategic national projects in
Presidential Regulation 3/2016, most recently revised in Presidential Regulation
56/2018. The current version of this list (table 3) includes more than 200 projects
and a further three programs, with an estimated total cost of Rp 4,150 trillion (about
$280 billion). Overall responsibility for realising these projects and programs
is in the hands of the specially created Committee for Acceleration of Priority
Infrastructure Delivery (the Indonesian acronym for which is KPPIP), rather than
 TABLE 3  National Strategic Projects

Financing Rp trillion % No. of projects

Central/reg. gov. budgets 428 10

State and reg. gov. enterprises 1,273 31
Private sector 2,449 59
Total 4,150 100
Sectors
Energy 1,212 29 11
Electricity program (generation,
transmission, sub-stations,
distribution) 1,036 25
Roadsᵃ 670 16 69
Industrial estates/special economic
418 10 29
zones
Railwaysᵇ 394 9 16
Other 420 10 97
Airports (new/refurbish) 7
Harbours (new/expansion) 10
Maritime and fisheries integrated centre 1
Town water supply 7
Sewerage system 1
Dams and irrigation 57
Flood control embankment 1
Broadband rings 2
Other hi-techc 2
Oil refineries (new/expansion) 3
Smelters 6
Gas (pipes, storage facilities, liquefaction, extraction) 7
Electricity from waste 1
Tourism infrastructure (transport, electricity,
1
and water supply)
Education (university campus) 1
Housing 3
Aircraft manufacturingᵈ 2
Economic redistribution program 1

Sources: KPPIP (2018). Laporan Pelaksanaan Proyek Strategis Nasional, Komite Percepatan Penyediaan
Infrastruktur Prioritas (June); Presidential Regulation 56/2018 (appendix).

Notes:
ᵃ of which toll roads 64 (4,598 km), non-toll roads 5 (620 km; length of one project not stated)
ᵇ of which inter-city 9, intra-city 7
c
including ‘technopark’ and multifunction satellite
ᵈ two types of aircraft.
290 Ross H. McLeod and Sitta Rosdaniah

the coordinating ministry for economic affairs or the national planning agency.
The committee is responsible for ‘coordinating the decision-making processes
to encourage settlement of issues arising from the lack of effective coordination
between the various stakeholders’.6 The name of the committee is somewhat
confusing in that the ‘strategic projects’ include many that would not normally be
thought of as public sector infrastructure, such as oil refineries; mineral smelters;
aircraft manufacturing; natural gas extraction, liquefaction, and storage; a (non-
state) university campus; an enormous housing program (‘one million houses’);
and a number of programs aimed to help the poor in various ways.
A striking feature of the current list is that implementation depends on the
private sector for no less than 59% of the required financing. Enterprises owned
by the central and regional governments are expected to provide a further 31%,
leaving a mere 10% to come from government budgets. Attempts by previous
governments to involve the private sector in infrastructure through public-
private partnerships have had very little success, so a significant change in
approach would seem necessary. Also, getting government-owned enterprises
involved may not always be easy. In particular, if state-owned oil and natural
gas corporation Pertamina is expected to build new oil refineries, the government
may need to stop requiring it to sell fuel at low prices without providing adequate
compensation. One of the important implications of relying so heavily on profit-
oriented firms, whether private or government owned, is that this will require
them to impose cost-covering user charges when the infrastructure in question
comes into operation. This would be entirely sensible, but it remains to be seen
whether these firms will be persuaded that this is the government’s intention.
There is no easy way to evaluate the performance of the committee because
the list does not include a timeline for completion of most projects/programs. In
the case of toll roads, for example, three roads totalling 61 kilometres had been
completed by the end of 2017, compared with a target list of about 64 roads
totalling 4,600 kilometres to be completed sometime in the indefinite future. In its
reports, for the most part, the committee tends to focus on the number of projects
completed or at particular stages of development, such as planning, procurement,
or construction. Since there are vast size differences in the various projects, this
approach is not very helpful. An exception to these comments is the 35.8 gigawatt
(GW) expansion program for electricity generation targeted for completion in
mid-2019.7 By mid-2018, 6% of the targeted new generation capacity had been
installed, while 47% was under construction; construction had not even started
on the remaining 47%, so clearly the target will not be reached by 2019.8

6. https://kppip.go.id/en/about-kppip/.
7. ‘Presiden Jokowi Luncurkan Program ‘35 Ribu MW Listrik untuk Indonesia’ (President
Jokowi launches '35 GW electricity program for Indonesia') Voice of America, 4 May 2015.
https://bit.ly/2PPOf13.
8. Introduction of the much-discussed fast train connecting Jakarta and Bandung is running
well behind schedule, according to Antara (18 May 2018), with only 20% of the needed land
yet resumed. https://bit.ly/2RrM9S6.
An Evaluation of Some Key Economic Policies 291

The Land Resumption Issue9

One of the issues often mentioned in relation to delays in creating new public
sector infrastructure is difficulty in the resumption of land,10 and the relevant laws
and regulations have been modified several times in attempts to overcome this
difficulty.11 We conjecture that the main underlying problem is with the incentives
facing the relevant parts of the bureaucracy: specifically, there are potential corrupt
gains from delaying the process (Rosdaniah 2015, 167–8). In the remainder of this
sub-section, we briefly outline a possible approach to redesigning the institutional
aspects of compulsory land acquisition.
Government has an important role to play in managing property rights when
a proposed change in land use is for some socially beneficial purpose, such as the
construction of new toll roads or railways. It is not feasible to rely on voluntary
market purchases of the land required, because each individual landowner is in
a monopoly position and is therefore able to appropriate a significant share of
the supernormal profit expected from such projects—thus removing most of the
incentive to undertake them. In principle, the last landowner to sell would be able
to capture the entire remaining supernormal profit because the road cannot be
completed without this last strip of land. Likewise, for non-toll roads, the entire
net social benefit could be captured by existing landowners if the government had
to rely on voluntary land sales. For this reason, governments need to exercise their
power of eminent domain to compulsorily gain control of the land in question.
Surprisingly, the various laws and regulations on compulsory land acquisition
make no reference to the government’s power to resume land at its discretion for
public purposes, as stated in Article 33 (3) of the Constitution: ‘land … is controlled
by the state and used for the maximum benefit to the people’. Under present
arrangements, landowners are effectively invited to lodge formal complaints
about proposed resumption of their land, and there is a lengthy process by
which such complaints are negotiated with a number of representatives of the
government. In fact, the only significant practical issue that needs to be resolved in
each individual case is the amount of compensation to be paid. It follows that the
land resumption process need not delay infrastructure projects: land can be taken
over and compensation paid almost immediately. Landowners dissatisfied with
the amount of compensation could then appeal to the courts to have the amount
adjusted without this holding back commencement of the project in question.
We therefore propose the establishment of a new land valuation agency and an
associated special court for adjudicating compensation claims to deal solely with
the compensation aspect of land resumption for public infrastructure purposes.
This agency would recruit experts in valuing land and the fixed assets located
thereon, and build up its in-house systems, database, and expertise over time. It
would aim for compensation payments based on estimated market value plus a
generous ‘compulsion margin’ (say, around 10%–20%) to reflect the fact that the
value of land to its owner will typically be somewhat higher than its current market
value, and that there will be unavoidable costs of relocation and disruption.

9. This section draws heavily on Rosdaniah (2015).

10. Laporan Pelaksanaan Proyek Strategis Nasional KPPIP June 2018, 14.
11. Most recently by Presidential Regulation 148/2015.
292 Ross H. McLeod and Sitta Rosdaniah

The sole objective of the agency would be to appraise the value of land being
compulsorily acquired, relying as much as possible on price information generated
by voluntary market transactions in comparable land. It would focus on rapid
appraisal by experienced professionals rather than relying on the current system
of negotiations between landowners and the bureaucracy: such ‘negotiations’ are
essentially meaningless when one party has the power to determine the final outcome
unilaterally and they are an open invitation to corrupt behaviour (analogous to
negotiations on income tax liabilities between taxpayers and tax officials).
Once the appropriate level of compensation has been determined—and the land
resumed—landowners would be able to appeal to the special court regarding the
amount. Other than this court, no other government entity would have any role
in determining the compensation amount, and the only parties involved in the
appeal would be the landowner and the agency. The court’s finding would depend
solely on the landowner’s ability to convince it that the agency’s analysis was
faulty and, to ensure that landowners only made genuine appeals, the process
could potentially result in either upward or downward adjustment of the amount
initially offered. Those who wished to appeal the compensation amount might
receive an initial payment of, say, 15% less than the provisionally determined
amount, potentially forfeiting this margin if the court adjusted the final amount
downward sufficiently.
The proposed agency and court would need to meet very high standards of
competence and integrity. Because of the obvious scope for ‘capture’ by rent-seeking
firms and individuals, it would be essential for its officials to be remunerated at
levels broadly commensurate with those in the private real estate sector, where the
necessary expertise currently resides, and from where most would be recruited. It
would also be essential for its top management to comprise individuals of known
high integrity. The highly respected Corruption Eradication Commission (KPK),
along with the special Corruption Crimes Court (Tipikor), could serve as a model.

Small Business Finance

Subsidised credit for small business has been a staple of Indonesian economic
policy since early in the New Order, reflecting a view that while micro, small and
medium enterprises play a crucial role in the economy, their progress is hampered
by lack of access to bank loans, mainly because they do not own substantial tangible
assets that can be used as collateral (Tambunan 2018, 319).
Since 2007, government policy in relation to this issue has taken the form of the
Loans to the People’s Businesses (Kredit Usaha Rakyat) program. In 2018 it was
planned to channel Rp 120 trillion of loans to small firms, with a highly subsidised
interest rate of 7% per annum.12 By comparison, prime lending rates were around
10% per annum in mid-2018. Not only is the interest rate extremely low, but
borrowers are also not required to provide collateral. Loans to finance working
capital requirements may be for up to three years, while those for investment are
for up to five years.

12. Coordinating Minister for Economic Affairs Regulation 11/2017. The rate had been 9%
in 2016 and 12% in 2015.
An Evaluation of Some Key Economic Policies 293

TABLE 4  Subsidised Loans to People’s Businesses (Kredit Usaha Rakyat)

Loans Borrowers

Total Total Average size

(Rp billion) % (thousand) % (Rp million)

State banks 84,111 96.2 3,277 98.6 25.7

BRI 60,222 68.9 2,994 90.1 20.1
Bank Mandiri 11,833 13.5 179 5.4 66.0
BNI 11,471 13.1 84 2.5 136.5
BRI Syariah 536 0.6 20 0.6 26.9
BTN 48 0.1 0 0.0 256.3
Prov. gov. banks 2,744 3.1 29 0.9 95.3
Private banks 576 0.7 18 0.6 31.2
Total 87,430 100.0 3,325 100.0 26.3

Source: Coordinating Ministry for Economic Affairs. http://kur.ekon.go.id/upload/doc/Penyaluran%20

KUR%20Tahun%202018-31%20Agustus%202018.pdf

Note: The data reflect cumulative totals for loans made in calendar 2018 through 31 August.

Some idea of the size of the program can be gained from table 4, although the
picture is incomplete because the data refer only to loans disbursed in 2018, rather
than to the amounts outstanding. Although there are nominally about 40 banks
involved in the program, lending is totally dominated by three state banks—and
by one in particular, BRI, which has a very wide network of village-based offices
and a great deal of experience in small-scale lending. By contrast, involvement
on the part of private banks and provincial government banks is negligible. The
average loan size is Rp 26 million (around $1,700). The program’s website claims
that there were about 58 million micro, small, and medium enterprises (MSMEs)
in 2013; even if the number of borrowers on the books were to increase to 5 million
by the end of 2018, it is clear that, even after decades, programs such as KUR in
Indonesia reach only a small proportion of target firms.
Perhaps because of the failure of subsidised lending programs to have much
impact, the government also tries by various means to force banks to lend to
MSMEs, from requiring preparation of detailed MSME business plans to mandating
higher loan to deposit ratios and threatening to impose greater minimum reserve
requirements for non-compliance (Rosengard and Prasetyantoko 2011, 291).
A recent study of small business financing in Indonesia observed that

SMEs perceive the lack of access to finance and the cost of loans to be greater obstacles
to development and high profits than do large firms. [Several studies] show that finance
shortages restrict development of small firms more than large firms. These general
findings indicate an unmet demand for finance. (IFC 2017, 105).

Firms’ self-interested perceptions do not provide a sound basis for efficiency-

improving economic policy, a prerequisite for which is the identification of
market failure. The alleged ‘unmet demand for finance’ is not in this category.
294 Ross H. McLeod and Sitta Rosdaniah

FIGURE 3  Indirect Bank Funding of Small Business

Loan
Bank Manufacturer/wholesaler

Trade/supplier credit

Small business

Indeed, this label is not even an accurate representation of reality, since small firms
actually have a range of financing options. McLeod (1992) notes, for example, that
small firms do not need to own capital to be in business. Many productive assets,
particularly land and buildings, can be rented instead: in effect, firms can borrow
physical capital instead of funds, thus obviating the presumed need for bank loans.
High profitability—necessary if the firm is to grow through reinvestment—does
not depend on the ownership of such assets but on how effectively they are used.
Moreover, the perception that banks do not lend to small firms is greatly
exaggerated in an economic sense. In reality, they do so indirectly through larger
firms, which themselves use part of their own borrowings from banks to supply
goods and services to smaller firms on credit—whether for just a week or two
to allow small traders to purchase trading stock, or for a year or more to allow
small firms to make longer-term investments in equipment, vehicles, and so on
(figure 3). The ubiquity of such financing arrangements reflects the recognition by
banks of their limited ability to compete in this segment of the finance market.13
Whether small firms obtain bank finance directly or indirectly is of little consequence.
What matters is that the most competitive (i.e., most efficient) financing technique
for any given circumstance comes to the fore.
It is argued that the intention of the KUR program is to improve the
competitiveness of micro, small and medium enterprises and cooperatives so that
they can graduate into sustainable, larger-scale businesses.14 If a single firm gains
access to cheaper finance, its profit will of course increase, but if similar firms also
obtain cheaper finance, temporarily elevated profits will be quickly competed away,
just as they would if the cost of any other commonly used input fell. The argument
is therefore an example of the fallacy of composition. It is unrealistic to expect
that a large proportion of small firms can ever become large, because they are
competing in the same markets. Those whose owners have inferior entrepreneurial
skills will not go far, even if given access to cheap finance. If small firms are losing

13. Rosengard and Prasetyantoko (2011, 276) assert that ‘lack of access to credit for viable
MSMEs wishing to borrow is a significant constraint to their development’, but they make
no mention of indirect lending through larger firms.
14. http://kur.ekon.go.id/gambaran-umum (authors’ translation).
An Evaluation of Some Key Economic Policies 295

out in competition with large ones (for example, corner stores competing with
supermarkets), the likely explanation is lack of economies of scale rather than lack
of access to finance, in which case steady displacement of less efficient small firms
from the industry in question is socially beneficial.
A notable aspect of the data on lending under KUR is the predominance of
lending to finance working capital for trading activity. By contrast, there is very
little for the purpose of investment in fixed capital, which would indicate an effort
to transform the nature of the business in question. Rather, the loans do nothing
more than reduce the cost of a particular input, temporarily increasing profitability
as any subsidy would be expected to do. In short, there is no persuasive evidence
of market failure in relation to the financing of small business. The seemingly
never-ending procession of subsidised credit schemes favouring small businesses
represents an unnecessary waste of effort and resources.

Decentralisation15
The basic rationale for decentralisation—or multi-level government—is ‘to bring
government closer to the people’ (Oates 1999, 1,120). This is no mere cliché.
Subnational governments located close to the populations they serve are best
able to determine the nature and extent of the local services desired by their
constituents and the amount they are willing to pay for these services; regular
elections give voters the opportunity to choose political parties considered most
likely to fulfil these desires. But Indonesia’s decentralisation arrangements fall
well short of maximising these potential efficiency gains, for two main reasons.
First, local governments (LGs, comprising kabupaten and kota, or districts
and municipalities) have been made responsible for delivering both education
and health care services, but in these fields they respond less to the desires of
their constituents and more to the commands of the central government, which
determines the key objectives of education and health-care policy. For example,
the laws on education and health were enacted by the central government,
and there is a national curriculum and a national requirement for a minimum
nine years of schooling for Indonesia’s children. The employment conditions
of teachers and health-care professionals are also centrally determined. To a
large extent, LGs operate, in effect, as contractors to the central government in
these key fields.
Whether outsourcing education and health care to LGs yields superior outcomes
is open to question: economies of scale are important to achieve efficiency, but
many kabupaten have tiny populations. However, taking this choice as given, the
key related decision is how these activities are to be funded. At present, LGs have
to use block grant transfers from the centre to meet most of the cost of providing
these services. The per capita amount of this funding varies enormously according
to the size of the population of the LG in question (figure 4), but it is unrelated
to demographic composition. Some LGs therefore have far less, per capita, to
spend than others on education and health, making it virtually impossible to
achieve anything like uniform national outcomes in these fields. In addition, any

15. This section draws heavily on Fadliya (2015).

296 Ross H. McLeod and Sitta Rosdaniah

FIGURE 4  Local Government Total Revenue Per Capita

(Rp million) and Population (million)

Revenue
60

50

40

30

20

10

0
0 1 2 3 4 5 6
Population

Source: World Bank, Indonesia Database for Policy and Economic Research.

LG accused of failing to deliver services of the desired quality can always deflect
blame to the centre for providing insufficient funding.
An obvious quasi-market solution to these problems is to formalise LGs’ role
as quasi-contractors, and to ‘pay’ them accordingly for services rendered at
predetermined rates per unit—just as construction contractors are paid specific
unit rates for laying pipelines, building roads, and so on. For example, LGs
could be paid specified amounts annually for each student enrolled in primary,
junior secondary, and senior secondary school, and for each member of several
demographic components of the population for whom they provide health care,
such as infants, young children, teenagers, women of child-bearing age, persons
of working age, and the elderly. The aggregate flow of funds to all LGs combined
would not change, but the total would be allocated equitably across jurisdictions
in proportion to the quantity of services provided. At present, there is no such link.
Indeed, each LG is constitutionally obligated to spend at least 20% of its budget
on education and 5% on health, but since the size of LG budgets varies over an
extraordinarily wide range in per capita terms, this implies vastly different levels
of expenditure per capita on these services.
Starting from the premise that each local government acts as agent for the
representative individual within its jurisdiction, it seems desirable that fiscal
An Evaluation of Some Key Economic Policies 297

TABLE 5  Gross Violations of Horizontal and Vertical

Equity Principles (selected examples)

Transfer per capita

Jurisdiction HDI Poverty (%) (Rp thousand)

Horizontal equity
Tana Tidung 10.2 54,938
Klungkung 10.3 1,303
Mamberamo Raya 61 43,907
Paniai 61 4,374
Vertical equity
Maybrat 35.6 18,065
Natuna 3.8 18,076
Nduga 49 7,614
Kepulauan Mentawai 70 7,649

Source: World Bank, Indonesia Database for Policy and Economic Research.

transfers are broadly consistent with the principles of horizontal and vertical
equity (i.e., equal treatment of equals and more generous treatment of the less
well-off). With enormous variation in per capita revenues, it is unsurprising
to find gross violations of these principles. Table 5 shows two pairs of cases
in which jurisdictions that are equally well-off on the basis of either poverty
incidence or their human development indices (HDIs) receive vastly different
per capita transfers from the centre, along with other cases in which two pairs
of jurisdictions with very different levels of either poverty or HDI receive similar
per capita transfers. The pattern in figure 4 looks as it does, resulting in these
violations of standard equity principles, for reasons explained in the appendix.
The second major shortcoming of current decentralisation arrangements is
in areas of activity where LGs act on behalf of their constituents rather than
the central government—in particular, in the provision of public goods, and
in dealing with externalities. Public goods such as stormwater drainage and
flood control are local rather than national, and the same can often be said of
externalities such as air and water pollution. LGs are much better able than
the central government to discover what is important to their constituents, and
therefore to provide the appropriate amount and type of these kinds of services.
An essential aspect of this process, however, is estimating how much constituents
value the package of services provided—and thus how much they are prepared
to pay for them in taxes—because only when they are confronted with the cost
of provision (as with private goods and services provided by the market) does
their expressed demand for services have any real meaning.
The problem here is the decision to fund almost all LG expenditures from the
central government budget. Since LGs therefore have very little need to raise funds,
they are in no position to weigh up the cost of providing services against what their
constituents are willing to pay for them. The alternative approach is to require LGs
298 Ross H. McLeod and Sitta Rosdaniah

to generate their own revenue, most obviously by taxing land and buildings, as
well as by imposing cost-based user charges where feasible. Land and buildings
taxes were originally pre-empted by the central government, and although more
recently these revenues have been handed to local governments, the authority of
the latter to set tax rates is greatly circumscribed.16 Requiring LGs to fund their
own provision of public goods and their efforts to deal with externality problems,
while at the same time giving them the freedom to set their own rates of taxation
on land and buildings, would force them to provide a package of these kinds of
services for which local citizens would be prepared to pay.
These two market-mimicking suggestions regarding fundamental changes to
the fiscal relationship between the central government and LGs would improve
the efficiency and equity of decentralised governance by forcing the latter to be
financially self-sufficient. To the extent it is desired to use transfers to LGs as an
additional instrument of national income redistribution in favour of the poor,
however, it would be sensible to retain a much-simplified untied transfer, with
the per capita amount calculated to increase as per capita income declines and
as poverty increases, along the lines proposed by Fadliya (2015, 33). A simplified
version consistent with the principles of horizontal and vertical equity is

  Pj  Pa    
T j  Ta 1  1
 Pa 
2 C j  Ca 
 
where T = per capita transfer; P = poverty incidence; C = per capita consumption
(as proxy for income); α1 > 0 = adjustment factor for poverty; α2 > 0 = marginal
rate of redistribution; j = 1, 2, … n denotes LGs; a denotes national average
values; and Ta = per capita total amount of funds available for transfer. A
jurisdiction with average levels of poverty and per capita consumption receives
the average per capita transfer. Better-off jurisdictions receive less (for given
poverty levels), and those that have higher than average poverty receive more
(for given consumption levels).

Minimum Wages
Article 88 of Labour Law 13/2003 and its elucidation stipulate that employees are
entitled to an income sufficient to provide a ‘decent standard of living’, interpreted
as ‘the ability to meet the reasonable living needs of workers and their families,
including food and drink, clothing, housing, education, health, recreation, and
saving for retirement’. This closely reflects the official (BPS) definition of poverty:
‘the inability to fulfil basic food and non-food needs’.17 To achieve its goal, the
government implements various wage policies, including fixing minimum wages
(Art. 89 (2)). As Bird and Manning (2008, 917) put it: ‘An active minimum wage
policy has been adopted partly as a tool to reduce poverty’. Minimum wage (MW)
policy thus appears to reflect a belief—or perhaps a pretence—that it is possible

16. Law 28/2009 imposes a maximum rate of 0.3% for the land and buildings tax (Art. 80).
17. https://www.bps.go.id/subject/23/kemiskinan-dan-ketimpangan.html.
An Evaluation of Some Key Economic Policies 299

to legislate poverty out of existence. The notion that Indonesia’s workers could
enjoy increasing incomes in the absence of MWs seems never to be considered,
notwithstanding that many countries have never introduced them (Manning
and Roesad 2007, 74).
In the past when the MW was being reset, there was often debate about how
much employers could ‘afford to pay’. Firms have little interest in this elusive
concept, however. Their actual behaviour depends instead on a simple cost-
benefit comparison. Specifically, no rational employer will recruit or retain a
worker at a MW of $2 a day if that worker can only generate revenue of $1 a day,
so the impact of a MW increase is for firms to reduce their employment of such
workers: the law can determine the price of labour but not the quantity employed.
Thus, although some workers gain from the higher wage, others lose access to
these kinds of jobs, ending up working for less than they would otherwise in
areas where the MW is not enforced.18 Thus, to describe MW increases during
a certain period as ‘a fairly good outcome for labour’ (Manning and Pratomo
2018, 172) seems highly misleading, because the focus is only on workers who
gain from the policy, ignoring those who lose. MW policy is not really about
redistributing the economic pie between labour and capital: the long-term impact
on the owners of capital is minimal, since the rate of return to capital in Indonesia
is broadly determined in the global financial market. Rather, its purpose is to
prevent low-skilled labour from competing with more highly skilled labour, to
the advantage of workers in the latter group lucky enough to work in areas where
the MW is enforced.19
Rapid increases in MWs in recent years have seen their levels rising close
to, or even above, average wages—and far above the poverty line—in many
jurisdictions (table 6). There were at least two main reasons for this. First,
responsibility for setting MWs passed from the central government to the
provinces—and on to districts and municipalities if these third-tier jurisdictions
desired to move higher than the province minima (Manning and Roesad 2007, 75).
Second, adjustment of MWs in line with objectively measured changes in the cost
of living was discontinued in favour of bargaining between unions, employers,
and governments.20
Promises of large MW increases were attractive to voters, as the only losers
seemed to be the relatively few owners of firms in the formal sector (presumed
to be well-off), while the potential beneficiaries seemed to be all those employees
with wages below the promised new minima. Few would have understood that
this policy would displace workers from the formal sector where MWs could
actually be implemented, forcing them to compete against workers in agriculture

18. Empirical support for this argument can be found in studies such as Suryahadi et al.
(2003), Bird and Manning (2008), and Merdikawati (2018).
19. Allen (2018, 208) asserts that ‘the general purpose of the minimum wage is to militate
against unduly low pay for unskilled and low-skilled workers’, seemingly confusing the
nominal purpose of MWs with their actual impact.
20. The system later returned to an objective, formula-driven process (Hamilton-Hart and
Schulze 2016, 282), but this did nothing to overturn the huge MW increases that had already
occurred.
300 Ross H. McLeod and Sitta Rosdaniah

TABLE 6  Minimum Wages Relative to Average Wages

and Poverty Lines across Provinces

Minimum wage : average wage Minimum wage : poverty line

Minimum Maximum Average Minimum Maximum Average

2005 2.6 4.7 3.6

2006 2.7 5.1 3.8
2007 2.7 5.3 4.0
2008 0.5 0.8 0.6 2.8 4.8 3.7
2009 0.4 0.8 0.6 2.7 4.9 3.7
2010 0.5 0.9 0.6 2.5 4.8 3.6
2011 0.4 0.9 0.6 n.a. n.a. n.a.
2012 0.4 0.9 0.6 2.7 5.6 3.7
2013 0.4 0.9 0.6 2.5 5.5 3.6
2014 0.5 1.0 0.7 2.7 6.3 4.0
2015 0.6 1.1 0.8 2.9 7.2 4.3
2016 0.5 1.1 0.8 3.1 7.4 4.6
2017 0.4 1.0 0.8 3.3 7.9 4.9
2018 0.5 1.2 0.9 3.3 7.9 4.9

Sources: BPS and Ministry of Manpower and Transmigration, through CEIC Asia Database.

Note: Within any province, individual districts and municipalities may set higher minimum wages, so
the actual maximum ratios are correspondingly higher than those shown in the table.

and the informal sector, who would have their already low incomes reduced as
a consequence. Most workers in the latter group, who comprise a majority of
the workforce, were not protected by the MW—either because they were self-
employed (i.e., non-wage earning) or because high administrative costs made
enforcement infeasible in their workplaces.21 The policy therefore conflicts with
the objective of reducing poverty.
We conjecture that additional support for higher MWs came from the bureaucracy.
MWs set far above free market levels create incentives for bribery and opportunities
for bureaucratic extortion, since firms that use relatively low-skill labour intensively
have much to gain from non-enforcement. Wide gaps between union demands and
employer offers created opportunities for sub-national governments to determine
outcomes (Allen 2018, 206) closer to the former, thus providing greater incentives
for employers to bribe regulators.
Poverty reduction and modernisation of the economy are likely to be held
back needlessly by MWs, which reduce the profitability—and therefore slow
the expansion—of large, low-skill, labour-intensive firms in the formal sector,
condemning Indonesia to continued heavy reliance on small-scale, traditional
activity such as agriculture and petty services. It was the rapid expansion of the
more modern sectors in the late 1980s and early 1990s that allowed the economy

21. Allen (2018, 216–22) provides an extended discussion of non-compliance with MWs.
An Evaluation of Some Key Economic Policies 301

to grow at annual rates of around 7%—well above the 5% rate of recent years—
contributing to a rapid decline in poverty. Arguably, MW policy has played an
important part in killing Indonesia’s golden goose.

Managing the Bureaucracy: The New Law on the Civil Service

Although Indonesia’s Corruption Eradication Commission (KPK) continues to be
busy, recording many new arrests, prosecutions, and imprisonments of corrupt
officials, corruption is still rife in the bureaucracy. Thus, Indonesia ranked only 96
out of 180 countries in Transparency International’s 2017 Corruption Perceptions
Index, scoring just 36 on a scale of 0–100.22 But another problematic characteristic
of the bureaucracy receives little attention by comparison: namely, its competence.
For the government to optimise its contribution to economic growth, it will need
to greatly improve the competence of its bureaucracy—a task that seems to have
been largely overlooked hitherto. The current anti-corruption approach, focused
on detection and punishment, is unlikely to contribute much to the objective of
achieving an effective civil service.
Since private-sector personnel management practices work well, it makes sense
to consider adopting similar processes within the bureaucracy, even though it is not
motivated by profit. As agent of the public, it should have the same commitment
to efficiency and to delivering services of the quality that its ‘customers’—the
public—want, as do private sector firms. That said, it is precisely the fact that
the bureaucracy has a monopoly position—with no threat of it being ‘forced out
of business’ by more efficient competitors—that permits poor human resource
management practices to persist almost indefinitely.
Until recently management of human resources in Indonesia’s bureaucracy
differed greatly from common practice in the private sector. In particular, when
filling positions, the pool of eligible applicants was extremely limited except at the
levels of new school or university graduates. Above these levels, hardly anybody
was hired from outside the civil service, or even from outside the department in
question. Moreover, priority in promotion nearly always reflected seniority rather
than capability or merit. Change began to become apparent, however, following
the enactment of Law 5/2014 on the Civil Service. The law emphasises the need to
get rid of corruption and political intervention, and to improve capacity to properly
deliver services to the public. It notes that recruitment, placement, and promotion
processes in the past failed to match individuals’ skills to those needed for each
position. It expresses the intention to apply merit principles to human resource
management, and it requires the government to pay fair and reasonable salaries to
its employees, consistent with their workloads, responsibilities, and job risks (Art.
79). Of particular interest is the stipulation (Art. 60, 61, 95) that every Indonesian
citizen has an equal opportunity to apply to become a civil servant.23
As a codification of general principles for managing the civil service, the new law is
a big improvement on past practice, but there is one glaring weakness. The reference
to fair and reasonable salaries consistent with workloads, responsibilities, and job

22. https://www.transparency.org/news/feature/corruption_perceptions_index_2017.
23. Implementing this would require a substantial change to the prevalent organisational
culture.
302 Ross H. McLeod and Sitta Rosdaniah

risks is unhelpful. There is no simple way to measure such things as workloads,

responsibilities, and job risks, much less to attach a monetary value to them, so any
salary structure established on this basis is necessarily subjective and arbitrary. Based
on the view that nobody should be made financially worse off by choosing a career
in the civil service, it seems preferable to interpret ‘fair and reasonable’ to mean
‘comparable with salaries available to individuals with similar skill sets within the
private sector’. It is no simple matter to compare the skill requirements of positions
in the private and public sectors, so a certain amount of subjectivity is unavoidable.
But job-seekers themselves will compare jobs in the two sectors and come to their
own conclusions as to whether the corresponding civil service salaries are indeed
‘fair and reasonable’, so it is important to get this right.
This has not been done thus far. Figure 5 compares remuneration in the government
and private sectors at various management levels. There are four echelons of
management in the bureaucracy, and we have taken these to be roughly comparable
with junior managers, senior managers, directors, and chief executive officers (CEOs)
within the private sector. In both sectors there is a range of salary levels, so we present
low and high levels for both.24
Clearly, private sector managerial-level salaries far exceed those in the civil service,
especially at higher levels of management: the government fails to provide ‘fair and
reasonable’ salaries to its high-level employees in our preferred interpretation of the
term. The obvious conclusion to be drawn from these comparisons is that opening up
applications for jobs within the civil service at these levels to all citizens is unlikely
to have any significant impact, because potential applicants from the private sector
would have to make huge financial sacrifices if they were to move between sectors.
One consequence of ignoring market salaries for managerial-level skills—and thus
effectively severely limiting the pool of potential strong applicants for civil service
positions—is illustrated by the attempt, initiated by the Megawati Sukarnoputri
government, to reform accounting practices within the public sector (McLeod and
Harun 2014). While such reform was long overdue, making it effective required much
more than the mere introduction of new laws. Implementing these laws required
high-level accounting skills that were largely absent in the civil service, so the reform
turned out to be much less successful than hoped. The key obstacle was inability to
recruit qualified and experienced accountants in the large numbers required to effect
this reform, given that these kinds of skills attract quite high salaries in the private
sector. Generalising this example to the civil service as a whole, we can say that
the organisational competence of the bureaucracy depends directly on its ability to
recruit from as large a pool of applicants as possible, in order to maximise the chance
of appointing individuals who are both competent and of high integrity.

24. Remuneration data for the civil service can be found in several presidential regulations
(e.g., 156/2014). The amounts include the basic salary plus a performance allowance
that varies across government institutions. Information on private sector salaries was
compiled from a number of commercial sources, including Kelly Services Indonesia
(https://www.kellyservices.co.id/), Robert Walters (https://www.robertwalters.co.id), and
Michael Page (https://www.michaelpage.co.id/). The existence of such sources clearly
indicates the importance to employers and job applicants of up-to-date information on
private sector salaries.
An Evaluation of Some Key Economic Policies 303

FIGURE 5  Remuneration in the Private Sector and the Civil Service (Rp million/month)

Junior manager Priv. low

Priv. high
Echelon 4
Gov. low

Senior manager Gov. high

Echelon 3

Director

Echelon 2

CEO

Echelon 1

0 50 100 150 200 250 300

Source: Refer to footnote 24.

A second consequence relates to the behaviour of individuals within the

bureaucracy. Indonesia’s civil servants are well aware of the significant financial
sacrifice implied by choosing a career in the civil service rather than the private
sector. They can hardly be blamed for being constantly on the lookout for ways to
generate additional income so that they can enjoy a standard of living similar to that
of their private sector peers. One example of this is the case of university lecturers,
who typically limit their time spent with students in order to participate in much
more lucrative consultancy projects, speaking engagements, and the like. Within
the bureaucracy more generally, numerous perfectly legal ways have been found to
generate additional income (the introduction of a wide variety of special allowances
for attending meetings and conferences, participating in study tours, projects, and
the like, and—for a lucky few—appointment as commissioners of state-owned
enterprises). In addition, however, many officials choose to involve themselves
in corruption. That this is a common practice is evident in colloquial references
to ‘wet’ (basah) parts of the bureaucracy: those prized for their opportunities to
engage in corruption.
In short, the attempt to reform the bureaucracy evidenced by Law 5/2014 on the
Civil Service and its implementing regulations is unlikely to succeed in its aims
because of failure to appreciate the importance of salaries in the market for skilled
labour. To make substantial progress in improving the capacity of the civil service,
while at the same time cutting back on corruption, it will be necessary not only
to extend eligibility for high-level positions to the entire population, but also to
make this meaningful by adopting a salary structure that is competitive with that
in the private sector.
304 Ross H. McLeod and Sitta Rosdaniah

APPENDIX
The formula for calculating transfers to each LG is dominated by a central
component that aims to allocate funds in proportion to various factors that together
form a proxy for so-called ‘fiscal need’, each of which is given a weighting to reflect
its desired relative importance.25 This component is expressed as

5
xn
T  F an
n 1 Xn

where αn are the weights (summing to 1) for allocation factor n, xn and Xn are the
individual and either total or average values of the allocation factors,26 respectively,
and F is the total pool of funds available. Taking x₁ and X₁ to refer to population—
now denoted p and P, respectively—and expressing the transfer in per capita terms,
we obtain

T a1 F 5 x
t   F  an n
p P p n2 X n

t F 5
xn
p

p2
a
n2
n
Xn
0

That is, per capita transfers decline, asymptoting to zero, as population increases,
all other things equal.27 Two jurisdictions equal in all respects other than population
receive different per capita transfers. A better-off jurisdiction that has a small, well-
off population will receive a higher transfer than a sufficiently large one whose
population is relatively poor.

25. See Martinez-Vazquez and Boex (2001, 22–23) for a full exposition of this kind of transfer.
The Indonesian system of transfers is more complex than described here (see Fadliya 2015).
26. Other than population, the allocation factors are area, the Human Development Index,
regional GDP per capita, and the Construction Price Index.
27. Boadway and Shah (2009: 362–3) appear to endorse the same kind of system—making
no reference to the implications for per capita transfers.
An Evaluation of Some Key Economic Policies 305

REFERENCES
Allen, Emma. 2018. ‘Labor Market Policies and Institutions in Indonesia’. In Indonesia:
Enhancing Productivity through Quality Jobs, edited by Edimon Ginting, Christopher
Manning, and Kiyoshi Taniguchi. Manila: Asian Development Bank.
[Arndt, H.W.] 1966. ‘Survey of Recent Developments’. Bulletin of Indonesian Economic Studies
2 (5): 1–21.
Bank Danamon. 2018. 'Indonesia Economic Briefing'. 20 August 2018.
Basri, Muhammad Chatib. 2016. 'The Fed’s Tapering Talk: A Short Statement’s Long Impact
on Indonesia', Ash Center Occasional Papers. Cambridge, MA: Harvard Kennedy School.
Basri, Muhammad Chatib. 2018a. ‘Twenty Years after the Asian Financial Crisis’. In Realizing
Indonesia’s Economic Potential, edited by Luis E. Breuer, Jaime Guajardo, and Tidiane Kinda.
Washington, DC: International Monetary Fund.
———. 2018b. ‘Waspada’ (Be on the Alert). Kompas, 15 August.
Basri, Muhammad Chatib and Arianto A. Patunru. 2006. ‘Survey of Recent Developments’.
Bulletin of Indonesian Economic Studies 42 (3): 295–319.
Bird, Kelly, and Chris Manning. 2008. ‘Minimum Wages and Poverty in a Developing
Country: Simulations from Indonesia’s Household Survey’. World Development 36 (5):
916–33.
Boadway, Robin, and Anwar Shah. 2009. Fiscal Federalism: Principles and Practice of Multiorder
Governance. New York, NY: Cambridge University Press.
EIU (Economist Intelligence Unit). 2011. 'Strong Growth Takes Indonesia to Middle-Income
Status'. EIU. 2 March 2011. http://country.eiu.com/article.aspx?articleid=1027866487&Co
untry=Indonesia&topic=Economy&subtopic=Recent+developments&subsubtopic=Econ
omic+performance:+Strong+growth+takes+Indonesia+to+middle-income+status)
Fadliya. 2015. ‘Fiscal Aspects of Decentralisation in Indonesia’. PhD thesis. Canberra, ACT:
Australian National University.
Hamilton-Hart, Natasha, and Günther G. Schulze. 2016. ‘Taxing Times in Indonesia: The
Challenge of Restoring Competitiveness and the Search for Fiscal Space’. Bulletin of
Indonesian Economic Studies 52 (3): 265–95.
IFC (International Finance Corporation). 2017. 'Serving the Needs of Indonesian SMEs:
Main Findings'. Washington, DC: International Finance Corporation, World Bank Group.
http://documents.worldbank.org/curated/en/556861495102658074/pdf/115101-WP-ID-
SME-Banking-Study-Main-Findings-PUBLIC.pdf
Juoro, Umar. 2018. ‘Jaga Ekonomi dari Tekanan Eksternal’ (Safeguard the Economy from
External Pressure), Kompas. 30 August 2018.
Kuncoro, Mudrajad, Tri Widodo, and Ross H. McLeod. 2009. ‘Survey of Recent Developments’.
Bulletin of Indonesian Economic Studies 45 (2): 151–76.
Manning, Chris, and Devanto Pratomo. 2018. ‘Labour Market Developments in the Jokowi
Years’. Journal of Southeast Asian Economies 35 (2): 165–84.
Manning, Chris, and Kurnya Roesad. 2007. ‘The Manpower Law of 2003 and its Implementing
Regulations: Genesis, Key Articles and Potential Impact’. Bulletin of Indonesian Economic
Studies 43 (1): 59–86.
Martinez-Vazquez, Jorge, and Jameson Boex. 2001. 'The Design of Equalization Grants:
Theory and Applications'. Atlanta, GA: Andrew Young School of Policy Studies, Georgia
State University.
McLeod, Ross H. 1992. ‘The Financial Evolution of Small Businesses in Indonesia’. In Informal
Finance in Low-Income Countries, edited by Dale W. Adams and Delbert A. Fitchett., 249–64.
Boulder, CO: Westview Press.
306 Ross H. McLeod and Sitta Rosdaniah

———. 2014. ‘The Ill-fated Currency Board Proposal for Indonesia’. In Trade, Development,
and Political Economy in East Asia: Essays in Honour of Hal Hill, edited by Prema‐chandra
Athukorala, Arianto A. Patunru, and Budy P. Resosudarmo. Singapore: ISEAS.
McLeod, Ross H., and Harun Harun. 2014. ‘Public Sector Accounting Reform at Local
Government Level in Indonesia’. Financial Accountability & Management 30 (2): 238–58.
Merdikawati, Nurina. 2018. ‘The Impact of District Minimum Wage on Employment by
Firms in Indonesia’, mimeo.
Oates, Wallace E. 1999. ‘An Essay on Fiscal Federalism’. Journal of Economic Literature 37 (3):
1120–49.
Patunru, Arianto, and Sjamsu Rahardja. 2015. ‘Trade Protectionism in Indonesia: Bad Times
and Bad Policy’. Lowy Institute. Last modified 30 July 2015. https://www.lowyinstitute.
org/publications/trade-protectionism-indonesia-bad-times-and-bad-policy
Rajah, Roland. 2018. ‘Indonesia’s Economy: Between Growth and Stability’. Lowy
Institute. Last modified 15 August 2018. https://www.lowyinstitute.org/publications/
indonesia-economy-between-growth-and-stability
Rosdaniah, Sitta. 2015. ‘Nominal and Actual Objectives of Economic Policy Making in
Indonesia’, PhD dissertation. Canberra, ACT: Australian National University.
Rosengard, Jay K., and Agustinus Prasetyantoko. 2011. ‘If the Banks Are Doing So Well, Why
Can’t I Get a Loan? Regulatory Constraints to Financial Inclusion in Indonesia’. Asian
Economic Policy Review 6: 273–96.
Suryahadi, Asep, Wenefrida Widyanti, Daniel Perwira, and Sudarno Sumarto. 2003.
‘Minimum Wage Policy and Its Impact on Employment in the Urban Formal Sector’.
Bulletin of Indonesian Economic Studies 39 (1): 29–50.
Tambunan, T.H. 2018. ‘The Performance of Indonesia’s Public Credit Guarantee Scheme
for MSMEs: A Regional Comparative Perspective’. Journal of Southeast Asian Economics
35 (2): 319–32.
econstor
A Service of

zbw
Leibniz-Informationszentrum
Wirtschaft

Make Your Publications Visible.

Leibniz Information Centre
for Economics

Timmer, Charles Peter

Working Paper
Pro-poor growth in Indonesia: Challenging the
pessimism of Myrdal's Asian Drama

WIDER Working Paper, No. 2018/103

Provided in Cooperation with:

United Nations University (UNU), World Institute for Development
Economics Research (WIDER)

Suggested Citation: Timmer, Charles Peter (2018) : Pro-poor growth in Indonesia: Challenging
the pessimism of Myrdal's Asian Drama, WIDER Working Paper, No. 2018/103, ISBN
978-92-9256-545-9, The United Nations University World Institute for Development Economics
Research (UNU-WIDER), Helsinki,
http://dx.doi.org/10.35188/UNU-WIDER/2018/545-9

This Version is available at:

http://hdl.handle.net/10419/190151

Standard-Nutzungsbedingungen: Terms of use:

Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Documents in EconStor may be saved and copied for your
Zwecken und zum Privatgebrauch gespeichert und kopiert werden. personal and scholarly purposes.

Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle
Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich
machen, vertreiben oder anderweitig nutzen.
Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen
(insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten,
gelten abweichend von diesen Nutzungsbedingungen die in der dort
genannten Lizenz gewährten Nutzungsrechte.
You are not to copy documents for public or commercial
purposes, to exhibit the documents publicly, to make them
publicly available on the internet, or to distribute or otherwise
use the documents in public.
If the documents have been made available under an Open
Content Licence (especially Creative Commons Licences), you
may exercise further usage rights as specified in the indicated
licence.

www.econstor.eu
WIDER Working Paper 2018/103

Pro-poor growth in Indonesia

Challenging the pessimism of Myrdal’s Asian Drama

C. Peter Timmer*

September 2018
Abstract: This chapter addresses the unrelenting pessimism in Asian Drama about Indonesia’s
development prospects. This pessimism was based on two key realities: the poor level of
governance demonstrated by the Sukarno regime (partly a heritage of Dutch colonial policies) and
the extreme poverty witnessed in rural areas. Using historical and modern data on the Indonesian
economy, the chapter explains the policy approach that resulted in three decades of rapid, pro-
poor growth during the Suharto regime. The Asian financial crisis in 1998 caused the Suharto
regime to fall and introduced democratically elected governments. After a decade of stagnation,
economic growth returned to the rapid rate seen during the first three decades of the Suharto
regime, but it is no longer pro-poor.

Keywords: governance, Indonesia, pessimism, poverty, pro-poor growth

JEL classification: B25, I32, N55, O53

Acknowledgements: I would like to thank Deepak Nayyar for extremely helpful guidance as we
went through several iterations of this paper. I also acknowledge comments from other authors in
this project, although I was not able to cope with all of them. Mistakes and opinions remain my
own.

*Harvard University, Cambridge, MA, USA, email: ptimmer63@gmail.com.

This study has been prepared within the UNU-WIDER project ‘Asian Transformations: An Inquiry into the Development of
Nations’.
Copyright © UNU-WIDER 2018
Information and requests: publications@wider.unu.edu
ISSN 1798-7237 ISBN 978-92-9256-545-9
Typescript prepared by Gary Smith.
The United Nations University World Institute for Development Economics Research provides economic analysis and policy
advice with the aim of promoting sustainable and equitable development. The Institute began operations in 1985 in Helsinki,
Finland, as the first research and training centre of the United Nations University. Today it is a unique blend of think tank, research
institute, and UN agency—providing a range of services from policy advice to governments as well as freely available original
research.
The Institute is funded through income from an endowment fund with additional contributions to its work programme from
Finland, Sweden, and the United Kingdom as well as earmarked contributions for specific projects from a variety of donors.
Katajanokanlaituri 6 B, 00160 Helsinki, Finland
The views expressed in this paper are those of the author(s), and do not necessarily reflect the views of the Institute or the United
Nations University, nor the programme/project donors.
1 Introduction

Two powerful, vicious circles drove Gunnar Myrdal’s pessimism about the development prospects
of South and Southeast Asia. The first was the pervasive poor governance that he termed the ‘soft
state’, usually controlled by corrupt, rent-seeking elites who were uninterested in economic policies
that might bring widespread gains in welfare. The second was the debilitating nutritional status of
most of the rural population, too poor and unproductive to even put enough food on the tables
for their families, thus reducing their productivity even further. These downward spirals, visible in
many countries in the early-to-mid-1960s when Myrdal was researching and framing Asian Drama,
were especially ‘dramatic’ in Indonesia. No one at the time seemed to doubt his judgement that
the country was the showcase example for development pessimism. This chapter tries to explain
why that pessimism turned out to be unfounded.

When the First Five-Year Plan of the Suharto government (REPELITA I), which ran from 1969
to 1974, was drafted, the entire economic team was aware of Myrdal’s pessimism about
development prospects in South Asia broadly (including Southeast Asia), and his especially harsh
judgement about the prospects for Indonesia. The Plan anticipated growth in per capita incomes
of about 3 per cent per year, and this was widely viewed as optimistic. But the subsequent half-
century proved that far more was possible. Per capita economic growth of nearly 5 per cent per
year was sustained over more than three decades, pulling tens of millions of Indonesians above
the nutrition-based poverty line. The trauma of the Asian financial crisis in 1998 hit Indonesia
hard, both economically and politically. But it recovered from this as well, in the context of a
surprisingly smooth transition from military dictatorship to widespread participatory democracy.
Much of the story of the democratic era is necessarily about politics and governance, and this
paper focuses on these and the accompanying economic trade-offs.

Three themes run through this paper, interconnected yet discrete enough to merit their own
headings: (1) the story of economic growth and its accompanying structural transformation of
what was a largely rural economy in 1968; (2) the record on poverty alleviation that was mostly a
result of that economic growth, but which was enhanced by a conscious policy of connecting poor
households to the growing economy; and (3) the political transformation that began well before
the 1998 fall of the Suharto regime, and which continues to this day.

The main message from this experience is succinct but powerful: the poor in Indonesia have been
very closely connected to economic growth in the country, benefiting differentially when the
economy was growing rapidly, and suffering disproportionately when the economy is not growing,
or suffers a major crisis, as in 1998. Of all the countries discussed by Myrdal, Indonesia’s record
from the late 1960s to the mid-1990s was one of the most ‘pro-poor’, and from the late 1990s to
the mid-2000s one of the most traumatic. But once again the country recovered from chaos to
build a successful economy with a rising middle class and aspirations of regional leadership.

The paper closes with a review of the economic record of the various democratically elected
presidents since 1999, and a prognosis of likely progress for the next quarter-century. Indonesia
will celebrate the centenary of its independence in 2045.

1
2 Understanding Indonesia’s pathway out of poverty

Myrdal was fully aware of the historical significance of the structural transformation in Western
societies as the route out of poverty for agrarian populations, although he apparently did not think
the process was relevant to the Asian societies he was observing in the mid-1960s. 1 The structural
transformation of poor, agrarian economies into rich, industrial- and service-based economies
located largely in urban areas is arguably the only sustainable pathway out of poverty. Surplus
labour from agricultural households migrates to more productive off-farm employment,
generating a process of economic growth. Productivity growth in the agricultural sector is also
needed as a way to feed workers in the cities and keep labour productivity in agriculture from
falling too far behind labour productivity in urban areas. A large gap in incomes between the two
sectors creates severe political tensions, and we will see these play out in Indonesia after 1998.

2.1 The dynamics of structural transformation

Two variables drive the structural transformation story: the share of agriculture in gross domestic
product (GDP) (AgGDPshr) and the equivalent share of the agricultural labour force in total
employment (AgEMPshr). The difference between these two shares measures the gap in labour
productivity between the agricultural and non-agricultural sectors (AgGAPshr). AgEMPshr is
nearly always larger than AgGDPshr, and thus labour productivity in agriculture is lower than in
non-agriculture (Timmer 2015; Timmer and Akkus 2008).

Figure 1 shows these three basic variables for Indonesia from 1880 to 2016. The common pattern
of structural transformation is readily apparent—both sectoral employment and contribution to
GDP decline relatively smoothly after the economic chaos of the Great Depression and the
Second World War. The gap between these two variables also tends to decline (absolutely; by
definition it tends to be negative), but remains negative. Even most advanced countries did not
reach parity between rural and urban incomes until late in their development. The United States,
for example, reached parity about 1980 (Gardner 2002).

1
The term ‘structural transformation’ does not appear in the 35-page index to the three-volume set (Myrdal 1968).

2
Figure 1: Long-run patterns of structural transformation in Indonesia, 1880–2016

90

70

50

30

10

-10
InGDPpc
-30

-50

-70

AgGDPshr AgEMPshr AgGAPshr

Note: the data used in Figure 1 for measuring structural transformation include four variables, of which the first
three are:
• AgEMPshr: percentage share of agricultural employment in total employment;
• AgGDPshr: percentage share of agricultural GDP in total GDP;
• AgGAPshr: percentage share of agricultural employment in total state employment minus the
percentage share of agricultural GDP in total GDP (AgGDPshr – AgEMPshr).
Values for the third variable are typically negative because the share of the labour force working in agriculture is
nearly always larger than the contribution of those workers to GDP. This definition makes Figure 1 easy to
interpret.
The fourth variable is the logarithm of per capita GDP for each decade (lnGDPpc) (see Table 4 for source and
details). The decadal data used in Figure 1 are presented in Table 1.
Source: Author’s illustration based on updated data from Timmer (2017) and Bolt et al. (2018).

Table 1: Decadal data used in Figure 1

1880 1890 1900 1910 1920 1930 1940

AgGDPshr 44.1 42.5 41.4 40.4 36.7 33.1 32

AgEMPshr 76.3 76.3 76.4 76.5 76.8 77.1 75
AgGAPshr –32.2 –33.8 –35.5 –36.1 –40.1 –44.0 –43
GDPpc 1,147 1,134 1,267 1,392 1,551 1,876 1,944
lnGDPpc 7.045 7.034 7.144 7.238 7.347 7.537 7.573

3
 1950 1960 1970 1980 1990 2000 2010 2016

AgGDPshr 35.5 31.0 27.8 21.4 18.5 15.7 13.9 13.5

AgEMPshr 72.2 69.9 64.7 56.4 55.9 45.3 39.1 31.8
AgGAPshr –36.7 –38.9 –36.9 –35.0 –37.4 –29.6 –25.2 –18.3
GDPpc 1,410 1,776 2,074 3,283 4,414 5,664 8,425 10,911
lnGDPpc 7.251 7.482 7.637 8.097 8.393 8.642 9.039 9.298

Source: Author’s illustration based on updated data from Timmer (2017) and Bolt et al. (2018).

Labour productivity in agriculture has two components: the physical yield of commodities on the
land at the disposal of the farm household, and the value of that output in home consumption or
in the market. Agricultural labour productivity, and farm incomes, can be raised by policies
affecting output prices, as well as by investments that raise physical yields of crops and livestock.
Of course, markets also affect agricultural commodity prices and significantly influence the cost
of policy choices to alter them. For tradable commodities such as rice, it is much easier for policy
to affect the domestic market price than the farm yield. High rice prices help close the gap between
rapidly rising urban incomes and lagging rural incomes. They also encourage domestic rice self-
sufficiency, which been the politically articulated approach to poverty reduction and food security
in Indonesia since the mid-1970s.

2.2 Initial conditions: where does the path begin?

Booth (2016) makes a powerful argument that the initial conditions for much of modern
Indonesia’s economic development were laid down by Dutch colonial policies in the nineteenth
and early twentieth centuries. Certainly Myrdal, in writing Asian Drama in the mid-1960s, was
acutely aware of the differential impact of British, French, and Dutch colonial legacies on the
development paths of their former colonies, and the Dutch come out rather badly in his
assessment. But does that matter 50 years on? Are the relevant ‘initial conditions’ the reality that
Myrdal was observing in the mid-1960s, which was quite grim, or does the story need to start much
earlier?

As Booth emphasizes, there is no neat answer to this question. The title of her book indicates that
her preference is to start with the colonial legacy. 2 A short paper does not have that luxury, but
Figure 1 does provide useful historical perspective. Indonesia suffered greatly from the Depression
and Second World War—the level of per capita GDP was still depressed (by 10 per cent) in 1950
compared with 1920. Compounding the problem, Indonesia missed much of the growth spurt
experienced in other parts of Asia between 1950 and 1970—per capita GDP was only 15 per cent
higher in Indonesia after those two decades (and almost all of that growth came between 1967 and
1970, after the fall of Sukarno), whereas Japan, for example, had per capita GDP that was more
than 400 per cent higher.

There will be more colonial history in subsequent sections, but the basic starting point for this
paper is the mid-1960s, when Myrdal was writing and when much of the development profession
was forming its first impressions of Indonesia. Based on his review of the Dutch colonial and

2
The full title is Economic Change in Modern Indonesia: Colonial and Post-Colonial Comparisons.

4
Sukarno era economic record, he is unrelenting in his pessimism for Indonesia’s future economic
prospects, as the following brief quotes indicate: 3

It was in their refusal to permit any significant advance toward responsible

government that the Dutch, together with the French, stood out in comparison
with the British as the guardians of a reactionary imperialism. …. Holland’s
stubborn refusal to accede to the moderate demands of indigenous nationalists, as
well as its practice of reserving even the less elevated positions in the colonial
administration to Dutch and Eurasian residents, drove the forces of nationalism
into conspiratorial channels and militant agitation. (p. 164)

Without collective cabinet responsibility for major policy directives, government

proved grossly ineffective and unstable. There were six cabinets between 1945 and
1948, seven between 1949 and 1957 …. Corruption, graft and fraud on a colossal
scale were conspicuous features of public life. (p. 367)

Furthermore, it has been suggested that Indonesian ‘national income per head in
the middle of the fifties was below the 1939 level, … probably below the 1929
level and may even be below the level of 1919,’ [citing Benjamin Higgins] and the
evidence shows a sharp decline since 1957. As things look at the beginning of
1966, there seems to be little prospect of rapid economic growth in Indonesia. (p.
489)

Of course, the next 50 years played out quite differently.

3 The evolution of the Indonesian economy

Indonesia is the original home of the dual economy. Boeke’s experience during the Dutch colonial
administration of Java led him to identify two types of economic agents—‘rational’ and
‘traditional’—with almost entirely separate spheres of economic activity (Boeke 1946). Lewis
(1954) built his Nobel-Prize-winning model of the dual economy with unlimited supplies of labour
on the behaviour of such agents. Indonesia’s history has modern lessons.

3.1 The historical setting for Indonesia’s growth experience 4

Indonesia is an archipelago nation, with thousands of islands. The soils are mostly volcanic on
Java, and they support intensive rice cultivation. The laterite soils on the Outer Islands are most
productive in tree crops and natural rainforest. Indonesia’s agriculture is dominated by wet- and
dry-season rice cultivation and a variety of tree crops for export. There has been long-term
experience with international trade, and Muslim traders established Islam peacefully by the
sixteenth century.

Under Dutch colonial rule, which started in the fifteenth century and ended with independence in
1945, the trade and tax regime favoured Dutch extraction of income, except for a brief period at
the beginning of the twentieth century, when Dutch public opinion supported a more
developmental approach to the colony, known as the ‘ethical policy’. However, this policy

3
The quotes are from Volume I of Asian Drama (Myrdal 1968).
4
Much of this early history is drawn from Timmer (1975).

5
collapsed with world prices for export commodities in the 1920s, and Indonesia experienced
especially poor economic management during the Great Depression. The Dutch forced the
Netherland East Indies to stay on the Gold Standard well after their regional competitors,
including the Japanese, devalued. The colonial authorities did build a significant network of
irrigation canals, roads, ports and shipping facilities, and railroads. There was, however, very little
investment in education of the local population. Only 3.5 per cent of the population was attending
school in 1939, compared with 13.3 per cent in 1995. The historical record suggests there was
severe poverty in the mid-nineteenth century, which fell gradually until the 1920s. Poverty
increased rapidly from the Great Depression until the end of the Second World War.

The Sukarno government took control after declaring independence in 1945, and put ‘politics in
command’ after 1958. It severely neglected agriculture and adopted an ‘inward looking’
development policy; the result was economic and political chaos by the mid-1960s. Incomes fell
and the hyper-inflation in 1965–66 had an impact on virtually everyone. The post-war recovery
had helped reduce poverty, but the poverty rate increased rapidly as inflation soared and the
economy collapsed. Probably 70 per cent of the population was ‘absolutely poor’ by 1966. Average
food energy intake was about 1,600 kilocalories per day. Hunger was widespread.

3.2 Pro-poor growth in historical perspective

Thanks to the painstaking historical research of Pierre van der Eng (1993a, 1993b, 2000, 2002), it
is possible to construct a long-run indicator of how well the poor have fared since 1880 (see Table
2). Van der Eng’s time series data from 1880 to 1990 can be divided into five main epochs: Dutch
colonial exploitation of the Indonesian economy (1880–1905); the ‘ethical policy’ era when efforts
were made by the Dutch to improve living standards of native Indonesians (1905–25); the
tumultuous period during the Great Depression, Pacific War, and fight for independence (1925–
50); the Sukarno era (1950–65), whose ‘guided economy’ after 1958 created the economic turmoil
that was observed by Myrdal; and President Suharto’s ‘New Order’ regime in the 1965–90 period,
which is the main focus of this paper. The original van der Eng data end in 1990 and the results
reported here depend on that dataset.

Three sets of calculations for each epoch are shown in Table 2. The first column shows the trend
growth rate of incomes per capita (YPC), as estimated from a semi-logarithmic time trend for the
respective time period. These growth rates vary widely. There was a sharp deterioration over the
quarter-century of economic chaos from 1925 to 1950, which contrasts equally sharply with strong
growth in both the ‘ethical policy’ era under the Dutch and, strikingly, in the Suharto era. Over the
entire time period for which van der Eng reports these data, per capita incomes rose 0.89 per cent
per year.

6
Table 2: Long-run patterns of pro-poor growth in Indonesia 5

Time period Growth rates (%) Income Index of pro-

elasticity for poor growth
KCAL (IPPG) 6

YPC KCAL
Dutch colonial exploitation, 1880–1905 0.33 –0.34 0.051
0.165 0.05
‘Ethical policy’ under the Dutch, 1905–25 1.63 1.39 0.878
2.805 4.57
Depression, the Pacific War, and the fight for –2.42 0.78 0.333
independence, 1925–50
1.064 –2.57
The Sukarno era, including the ‘guided economy’ 1.46 0.68 0.509
period, 1950–65
1.626 2.37
The ‘New Order’ regime of Suharto, 1965–90 3.45 2.10 0.595
1.901 6.56
The long-run averages, 1880–1990 0.89 0.22 0.313
1.00 0.89

Source: Timmer (2004b).

The second column in Table 2 shows the similarly estimated time trend for intake of food energy,
as measured from food balance sheet data on kilocalories (Kcal) consumed per capita per day, on
average for each year from 1880 to 1990. During two epochs this time trend was negative, which
indicates a decline in nutritional status on average for the whole society and the strong likelihood
of significant increases in hunger among the poor. During such episodes, poverty was rising. A
sharply positive trend in food energy intake, however, as during the ‘ethical policy’ era and the
early Suharto era, suggests that income growth was reaching the poor and improving their access
to food. Over the entire time period, the trend in food energy intake was just 0.2 per cent per year.

The relationship between the variables underlying these two trends is also reported in Table 2. The
third column reports the average income elasticity of demand for food energy (KCAL), which is
estimated from the annual data for each epoch. Importantly, the pattern of coefficients is similar
for the income elasticities and the rate of change in food energy intake. The logic connecting the
two is straightforward. Engel’s law suggests that the income elasticity of demand for food (of
which energy is an important component for the poor) is a declining function of income level.
When income growth includes the poor, their higher income elasticities for food energy raise the
income elasticity observed on average. It is thus possible to infer what is happening to the poor

5
Details of the regressions are shown in Timmer (2004b), which also provides a full explanation of the analytical
relationship between the overall incidence of poverty and the average income elasticity of demand for food energy.
6
The IPPG is calculated as the product of the growth rate in per capita income times the ‘standardized’ income
elasticity of demand for food energy (KCAL), where the base income elasticity is the value for the entire time period
from 1880 to 1990 (0.313). Growth rates are calculated as least squares time trends of logarithmic values of incomes
per capita (YPC) and average daily per capita food energy intake per capita (KCAL). The ‘top’ value for the income
elasticity of demand for food energy for each epoch is estimated as a constant elasticity value from a double logarithmic
function. The ‘bottom’ value re-scales this estimated value, with the 1880–1990 average of 0.313 equal to 1.000. As
an example, the IPPG value of 6.56 for the Suharto era from 1965 to 1990 results from the OLS-estimated rate of
growth in per capita income of 3.45 per cent, times the ‘standardized’ income elasticity of 1.901. This standardized
value is computed by scaling up the OLS-estimated income elasticity for the period of 0.595 from the historical base
income elasticity of 0.313. Thus 0.595/0.313 = 1.901, and 3.45 × 1.901 = 6.56.

7
during long-run periods of economic growth (or decline) by analysing these changes in food energy
intake (Timmer 1996).

This approach works, of course, only for those societies in which the poor wish to increase their
food energy intake when their incomes increase—that is, when they are still on the rising part of
the Engel curve. This is the case for Indonesia. Even in 2002, at least the bottom half of the income
distribution had significantly positive Engel elasticities for food energy. This half of the population
subsisted on less than US$2 per day (World Bank 2003a, 2003b).

Finally, Table 2 carries this inference process to its logical conclusion, by constructing a crude
‘index of pro-poor growth’. The scale is somewhat arbitrary, but it is based on an analytical
relationship between the overall incidence of poverty and the observed, average income elasticity
of demand (see the derivation and proof in Timmer 2004b). The income elasticity of food energy
for the entire period from 1880 to 1990, estimated to be 0.313, is used as the long-run base, scaled
to 1. It is multiplied by the long-run growth rate in per capita incomes, 0.89 per cent per year, to
generate the long-run average IPPG of 0.89. The income elasticity for each separate epoch is then
scaled relative to the long-run average, and multiplied by the growth rate in per capita incomes to
generate the IPPG for each epoch. Note that the IPPG incorporates both the growth and the distributional
dimensions of pro-poor growth, and this index is thus a country-specific version of equation 1 in
‘Concept paper on operationalizing pro-poor growth’ (World Bank 2004).

As shown in Table 2, the IPPG has varied dramatically over time, from –2.53 during the 1925–50
epoch, to 4.57 during the ‘ethical policy’ era of 1905–25. The index is surprisingly high during the
Sukarno era, when economic policy is widely regarded to have been a disaster. But a combination
of a modest recovery from the quarter-century of depression and wars, with average per capita
incomes rising 1.5 per cent per year, and a large average income elasticity for food energy, suggest
that what growth there was did actually reach the poor.

The strongest pro-poor growth has been from 1965 to 1990. The data analysed in Table 2 only
carry the story to 1990, which is well before the Asian financial crisis in 1997/98. The quarter-
century from 1965 to 1990 has an IPPG of 6.56, which is more than seven times the long-run
average and nearly half again as large as during the next best epoch, from 1905 to 1925. Clearly,
something quite outside earlier historical experience was going on during the first two and a half
decades of the Suharto era. What made this era so pro-poor?

3.3 The ‘New Order’ government of Suharto: why was growth so ‘pro-poor’?

In the early years of the Suharto government, pre-OPEC (1966–73), there was a need to establish
stability and consolidate political power. In this process, there was an important role for the food
logistics agency (Bulog) in stabilizing rice prices, and for donor assistance, especially the provision
of food aid. Major investments were made to stimulate agriculture: irrigation rehabilitation, the
introduction of high-yielding varieties of rice from the International Rice Research Institute (IRRI),
fertilizer imports and distribution, and the BIMAS programme of extension and farm credits.
Because median farm size was less than 1 ha, rice intensification had widespread benefits, although
larger farmers (those who cultivate about 1 ha of land) benefited the most in the early years (Afiff
and Timmer 1971).

Macroeconomic stability was achieved through a balanced budget and donor-provided foreign
borrowing, with all proceeds going to the Development Budget (Hill 1996; World Bank 1968).
Poverty fell rapidly as the economy stabilized and grew 5–6 per cent per year in per capita terms,
and as food production and overall food supplies rose sharply. Still, absolute poverty was thought
to be about 60 per cent in 1970. The first official poverty estimates, based on the 1976 National

8
Socio-Economic Survey (SUSENAS), indicate a national poverty rate of about 40 per cent (see
Table 3).

Even for an oil exporter, coping with high oil prices (1973–83) is not the luxury it might seem. To
be sure, there was rapid expansion of the economy as the role of the state expanded, but much of
this expansion was in inefficient public-sector investments. Accompanying the real appreciation
of the rupiah was declining profitability of tradable goods production, especially in agriculture
(Warr 1984). During the mid-1970s there was a growing sense of income inequalities and severe
poverty in rural areas, although the regional and commodity dimensions of the poverty masked its
economic roots.

Table 3 shows estimates of income distribution and the share of the population below the poverty
line, using data that are from a variety of sources, including from official Indonesian government
statistical sources. These historical data provide comparisons of poverty and inequality from
several methodological approaches and update the key data to 2016. Further historical data are
illustrated at the sectoral level in Figure 1, which charts the structural transformation of the
economy from 1880 to 2016.
Table 3: Measures of inequality and poverty in Indonesia, 1932–2016

Year Gini coefficient (a) Share of the population below the

Rural Urban Total (50 per cent of median) BPS
Indonesia People
1932 0.52 0.57 0.56 26.3 16.4 –
1939 0.58 0.53 0.60 – – –
–
1942 0.60 0.53 0.60 27.6 20.0 –
–
1953 0.51 0.49 0.55 25.4 21.1 –
1959 0.46 0.47 0.51 23.9 22.3 –
–
1975 0.16 0.31 0.28 9.4 12.2 –
1976 – – – – – – 40.1

1980 0.13 0.27 0.24 5.4 8.0 28.6

1985 0.19 0.20 0.24 5.4 8.9 –
1987 – – – – – – 17.4

1990 0.15 0.20 0.24 6.2 11.1 15.1

1993 0.16 0.27 0.31 10.2 19.1 13.7
1999 0.17 0.23 0.32 11.7 23.9 –

2005 0.16 0.19 0.34 13.3 29.8 16.0

2008 0.12 0.22 0.37 15.4 36.2 15.2

2012 – – 0.42 (c) – – 12.0 (c)

2014 – – 0.42 (c) – – 11.0 (c)
2016 – – 0.38 (c) – – 10.5 (c)

Source: Author’s illustration based on: (a) van Leeuwen and Foldvari (2016: 386, table 5); (b) van Leeuwen and
Foldvari (2016: 391, table 6); (c) World Bank (2017: 30–31).

Income distribution deteriorated sharply between 1976 and 1978, confirming the growing
anxieties. The technocrats took a highly original strategic approach to what was then diagnosed as

9
‘Dutch disease’, with a devaluation in November 1978 that came as a big surprise to financial
markets. After this, tradable goods production rapidly recovered, especially in agriculture. Poverty
rates after 1978 fell, driven by a significant recovery in the share of income garnered by the bottom
40 per cent of the distribution (Papanek 2004). 7

By the early 1980s the oil boom was over. It became necessary to restructure the economy for a
world of low commodity prices in world markets, which is the basic story from 1983 to 1993.
Agriculture continued to grow, and rice prices were stable (this stability amounted to protection
against the low prices in world markets). The government pursued aggressive exchange rate
protection via further devaluations in 1983 and 1986 (Hill 1996; Thorbecke 1995). Massive
investments in rural infrastructure from earlier oil revenues began to pay off in higher production
and lower transactions costs for marketed goods (and improved labour mobility). Industrial output
surged in the latter part of the period, led by labour-intensive manufactured exports.

The manufacturing sector contributed 29.2 per cent of the growth in GDP between 1987 and
1992, a radical increase from the 10 per cent contributed during the recovery from 1967 to 1973
(Hill 1996, 2000). Large-scale and sustained economic deregulation led to better incentives for
exports, and these were matched by incentives for foreign direct investment (FDI). Manufactured
exports responded even faster than policy makers had hoped, and contributed almost half of all
exports by 1992, up dramatically from the 3 per cent in 1980. The fortuitous ‘push’ in FDI from
Japan and the ‘pull’ from the attractive climate in Indonesia thus allowed manufactured exports to
play a significant role in employment generation by the end of the 1980s.

There was also a boom in the non-tradable economy. National income accounts, however, are not
kept according to this distinction–hence the data are more impressionistic. 8 Nonetheless, just as
the export economy was booming in the late 1980s and early 1990s, and overall GDP was growing
by nearly 7 per cent per year, roughly half of that growth was made up of non-tradable goods and
services. According to the Mellor model of poverty reduction (Mellor 2000, 2017), production of
non-tradable goods and services, especially in rural areas, provides the economic link between
higher incomes from both agriculture and manufacturing wages, and serves to pull people out of
underemployment in rural areas—and out of poverty.

The Mellor model stresses the role of producing rural non-tradables that are locally consumed—
processed foods, construction, trade, and small-scale manufactures—as the ‘ladder’ for
underemployed workers in agriculture to begin the climb to modern jobs at higher wages. In most
poor, rural economies this non-tradable sector is demand-constrained. That is, expanding it, and the number
of jobs it creates, does not depend on better access to capital or to management skills, but on
greater purchasing power among local consumers. Thus Mellor emphasizes rising profitability of
agriculture—through higher productivity, not higher prices. Higher prices for agricultural output,
especially food, do not contribute much to added demand for non-tradable goods and services

7
A sharp debate, totally confidential, took place in mid-1978 between advisers concerned about ‘Dutch disease’ and
its impact on rural incomes and poverty, who argued for a significant devaluation of the rupiah, and the trade and
industry advisers, who argued for a revaluation because of foreign exchange surpluses (Timmer 1994). The devaluation
in November 1978 set the exchange rate at Rp 625 per US dollar, instead of revaluing to what the market expected,
which was Rp 380 per US dollar.
8
It is hard to define the components of the non-tradables sector, although some items are fairly obvious, such as local
services, construction, and low-quality local manufactured goods with a limited market appeal beyond the immediate
vicinity. Even modern industrial goods produced at high cost behind tariff barriers are non-tradable. Obviously, poor
transportation facilities lead to high marketing costs, meaning that many potentially tradable goods are non-tradable.
Lower transactions costs and lower tariff barriers are among the surest ways to improve competition and market
efficiency.

10
because the higher food prices choke off demand for non-food items, except from farmers with
significant surpluses to sell. The growing wages of workers in a rapidly expanding manufacturing
export sector also contribute to higher demand for the output from the non-tradables sector.

This Mellor model is a three-sector version of the standard Lewis model of the dual economy. In
Mellor’s version there are two ‘commercial’ sectors—industry and modern agriculture. The latter
has come to use modern technology and is market-driven, a departure from the traditional rural
economy envisioned by Lewis (1954). Relatively separate from the commercial sectors is the ‘non-
tradable’ sector, which is informal and mostly rural. The two commercial sectors are the ‘engines
of growth’ because of their potential for rapid productivity gains. Connecting them to the ‘non-
tradable’ sector, however, is the key to a high ‘elasticity of connection’ between overall economic
growth and rapid poverty reduction. This is the sector in which most of the poor make a living
(Timmer 1997, 2002). Unless demand from rising incomes in the commercial sectors spills over to this non-
tradables sector, the poor tend to be left out of the growth process.

The combined boom in agriculture, manufacturing, and non-tradables means the period from the
late 1970s to the mid-1990s is one of the most ‘pro-poor growth’ episodes in modern economic
history. This result is a surprise to many. The extensive economic restructuring that took place in
the 1980s was expected to create widespread unemployment and lead to lower wages for unskilled
labour. Instead, agricultural growth continued, labour-intensive exports surged, and poverty
continued to decline throughout the period of restructuring (Ravallion and Huppi 1991). 9

Figure 1 provides an illuminating view of the structural dimensions of this growth. AgGDPshr
declines steadily as the economy is growing fairly rapidly—about 3.6 per cent per year per capita
between 1970 and 2010. But AgEMPshr declines only slowly in the early stages of growth, from
1970 to 1990, and then declines quite rapidly as the rural economy is integrated into the dynamic
urban economy via better roads, communications, and an educated rural labour force in search of
economic opportunities. The rapidly narrowing AgGAPshr after 1990 reflects the falling poverty
rates shown in Table 3.

Corruption and increasing distortions in resource allocation from 1993 to 1998 followed the
interests of the Suharto family, especially the children (and grandchildren!). These interests
distorted trade policy and public-sector investments, and had visible effects on competitiveness,
which were partly masked by the inflow of FDI (Cole and Slade 1996). As the economy boomed,
deregulation lost steam, first in the Bulog commodities, then more broadly. The performance of
the overall economy started to suffer, although poverty levels in 1996, the last SUSENAS report
before the crisis, dropped to their lowest levels ever. Absolute poverty, measured in a comparable
fashion to the poverty statistics reported first in 1976, fell below 12 per cent.

The three decades of superb economic results were over. The Asian financial crisis hit in late 1997.
Investors started to lose confidence in the ability of the Suharto government to cope, especially
after the new cabinet was named in April 1998, packed with Suharto cronies and relatives. The
crisis caused a massive depreciation of the rupiah, which eventually led to chaos in the domestic
rice market (Schydlowsky 2000). Spiralling rice prices late in 1998 led to huge increases in poverty,
which is estimated to have reached over 30 per cent of the population by the peak in late 1998 or
early 1999. This was a dismal period for economic growth and poverty reduction.

9
It should be noted that the rice sector in particular was protected from the worst pressures of low prices in world
markets, and the substantial devaluations in 1983 and 1986 provided enhanced profitability for the rest of the tradable
sector, with spillover effects on incomes in the non-tradable sector.

11
3.4 The ‘democratic era’ (1999–)

After Suharto’s surprise resignation in late 1998, Indonesia successfully elected a democratic
legislature in 1999, which in turn selected a new president, Abdurrahman Wahid, a popular Islamic
scholar and cleric who headed Nahdlatul Ulama (NU), the country’s largest Islamic organization.
Wahid attempted to reform the military’s role in government and was ultimately impeached by
parliament for that effort and accompanying corruption scandals. He was replaced by his vice-
president, Megawati, the daughter of Sukarno. She served for three years, but was defeated in the
election by her former coordinating minister for politics and security, Susilo Bambang Yudhoyono.

Representative democracy brought forth a new political economy of economic policy. Populist
voices ostensibly spoke on behalf of the poor. With two decades of experience, it remains to be
determined whether Indonesia’s pro-poor growth experience under a highly centralized and
politically dominant regime put down sustainable, even irreversible, roots, or whether the very
foundations of the strategy will come undone under political challenge.

It is already clear that the transition from the autocratic rule of Suharto, an era of economic policy
designed and administered by an insulated group of skilled technocrats, to a politically responsive
system has been difficult for both economic growth and its connection to the poor. There are few
public institutions in place to protect economic policy from polemicists.

As part of a broad donor effort to help Indonesian development, after the fall of Suharto the
World Bank led an effort to help the country cope with corruption in the national government.
The main vehicle was to promote decentralization of political power (World Bank 2001). Domestic
reform groups supported this agenda and responsibility for schools and most local services
devolved to kabupaten (‘county’) levels in 2002 in a ‘big bang’ political decentralization. Not
surprisingly, given the speed, this transfer was made without adequate funding, policy guidelines,
or training of local officials (Alm et al. 2001).

Inevitably, perhaps, corruption at the local level has become rampant. Local ‘trade’ policies are
being used to enforce commodity taxes and trade barriers, especially for rice. Partly because of the
resulting ‘compartmentalism’ in the economy, and the higher transactions costs for most economic
activities caused by these activities, donor interest and activity have focused on improved local
governance. The stakes are high. Ongoing research by the Social Monitoring and Evaluation
Research Unit (SMERU), the most influential local research institute focusing on poverty issues,
indicates that measures of the ‘quality’ of local governance were closely associated with the rate of
poverty reduction between 1999 and 2002 (Sumarto et al. 2004).

The country is clearly paying the price for decades of forgone institution building, especially in the
arenas of property rights and rule of law. The new democratic governments have found it difficult
to make rapid progress in these arenas, although more has been accomplished than many observers
credit (MacIntyre 2003; Ramage 2004). The government of Susilo Bambang Yudhoyono (known
universally as SBY) (2004–14) and the current government of Joko Widodo (again, universally
known as Jokowi) (2014–) have made significant strides in strengthening transparency in
government and addressing corruption.

Their records on economic growth and poverty reduction are respectable but not outstanding
relative to regional and historical performance. The data on per capita GDP in Table 4, for
example, show Indonesia’s record of economic growth per capita by political era since 1880. Per
capita GDP in constant 2011 US dollars fell by 4.38 per cent per year from 1996 to 1999 during
the Asian financial crisis, and then recovered by 3.31 per cent per year from 1999 to 2005. Still,
from its peak in 1996, before the crisis, until the previous peak in per capita incomes was restored

12
in 2005, incomes per capita grew just 0.6 per cent per year. It took a long time for democratic
governments to figure out how to manage the economy.
Table 4: Economic growth in Indonesia, 1880–2016

Year RGDPNApc Average annual Notes on historical era

(2011 US dollars) increase/decrease
(%/year)
1880 1,147 0.36 Dutch colonial
exploitation
1905 1,255 1.37 Dutch ‘ethical policy’
1925 1,648 –0.62 Depression, war, fight for
independence
1950 1,410 1.14 Sukarno era
1966 1,690 4.47 Suharto era
1996 6,269 –4.38 Asian financial crisis
1999 5,481 3.31 Fall of Suharto; recovery
to previous peak
2005 6,663 4.59 Democracy in command
2016 10,911

Note: The average annual rate of economic growth from 1996, the peak year before the crisis, and its full
recovery nine years later, in 2005, is just 0.6 per cent per year. Maddison Data Project note: ‘In addition to the
𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 series, we provide a measure of growth of GDP per capita that relies on a single cross-country price
comparison, for 2011. This series is also expressed in 2011 US dollars (and 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 = 𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅 in 2011), but
its defining feature is that it tracks the growth rate of GDP per capita as given in country National Accounts (or
their historical reconstructions). Following PWT [Penn World Tables], we refer to this measure of real GDP per
capita as 𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅. This series is primarily useful for comparing growth rates of GDP per capita over time.’
Source: Author’s illustration based on data from the Maddison Data Project (Bolt et al. 2018).

By 2005, however, the new SBY government was able to restore investor confidence, both foreign
and domestic, and the economy started to grow rapidly again. Between 2005 and 2016, per capita
GDP grew by 4.59 per cent per year, slightly faster than during the 30 years of the Suharto regime
(see Table 4). But the return to rapid economic growth was not mirrored by changes in poverty.
Between 2010 and 2016, the poverty rate only declined from 13.3 per cent to 10.9 per cent. Most
other welfare indicators track these statistics on the poverty rate. This is respectable progress, but
leaves Indonesia lagging other countries in its region and income category. Economic growth is
no longer ‘pro-poor’.

4 Political economy and governance

The standard story to explain the political economy of the Suharto regime’s emphasis on
agriculture (and pro-poor growth) relies on conflict between traditional political forces—
communist-inspired peasants and workers faced opposition from an authoritarian military
(Simpson 2008). Buying off the peasants was cheaper than repression. Rural development was
seen as the least-cost approach to political stability. Large-scale ethnic (Chinese) businesses bought
protection from Suharto and his military allies and received lucrative import and operating licences
in return. When these highly protected businesses, and their closely associated banks, collapsed in
the Asian financial crisis, the entire regime came unravelled. The vacuum of political institutions,
deliberately created by Suharto to remove any challenge to his authority, exposed the country to
years of political chaos and weak leadership. This standard story suggested there was only modest
hope that the presidential election in September 2004, when Megawati ran against SBY, would
return the country to strong leadership and rapid, pro-poor growth.

13
As with most stories based on conventional wisdom, there are substantial elements of truth in this
one. But it misses what distinguishes the Suharto regime from otherwise similar military
dictatorships around the world: its focus on development and the effort to improve the welfare of
the poor by connecting the rural economy to rapid economic growth. Fear of radical peasants
wielding scythes simply does not explain this passion, or the massive budgetary resources devoted
to it. Oil revenues helped, to be sure, but the basic strategy was already laid down before the OPEC
price shock. A much more nuanced story is needed, one that includes the complexities of the
structure of political power and the role of leadership. The remainder of this paper attempts to
provide some of that nuance by tracing the political economy of poverty through a governance
lens to the problems caused by the massive decentralization that was the first priority of the new
democratic government.

4.1 The political economy of pro-poor growth

Political scientists speculate on the nature of the political coalition assembled by Suharto to
maintain and strengthen his hold on power. This coalition was held together by the distribution of
economic resources, often in the form of lucrative access to such easily marketable commodities
as oil or timber. Import licences for rice, wheat, sugar, and soybeans were equally lucrative. Bulog
controlled these closely in the interests of the Suharto regime. Whether the pro-poor policies, and
results, of the regime were tied to keeping these interest groups satisfied, even at the expense of
faster economic growth in the short run, is the subject of active debate, especially because Bulog,
despite being ‘privatized’ in 2003, established close ties with the husband of President Megawati
and successfully lobbied for renewal of Bulog’s monopoly control over trade in most agricultural
commodities. The ability of Bulog to stall the deregulation process in the early 1990s is seen by
some observers (including this one) as an early signal that the entire growth process was running
off the rails into corrupt and distortionary cronyism. From this perspective, the collapse of the
formal sector during the Asian financial crisis was not such a surprise, as it had become increasingly
dominated by these interests (Cole and Slade 1998; Stern 2003, 2004).

The most debated political economy aspect of the New Order government was the near-
schizophrenia between macro and sectoral policies. What is so puzzling is why macroeconomic
policy was left largely in the hands of very talented, but highly apolitical, technocrats. Persuasive
arguments are made that they provided access to the donor community, which has been a strong,
almost lavish, supporter of Indonesia since the late 1960s (Simpson 2008). But another argument
is simply that the technocrats delivered the economic growth the country so desperately needed.
In a comparison of the political economy of growth in the Philippines and Indonesia, Thorbecke
(1995) came to the following conclusion:

In the final analysis, the most fundamental difference between the development
environments in the two countries relates to the macroeconomic policy
management and the role of the technocrats. In Indonesia, the latter followed a
consistent, far-sighted, credible, and enlightened macroeconomic policy that was
outward-oriented and provided a framework within which both economic growth
and poverty alleviation could occur. In particular, the key policy instrument that
the technocrats relied on was an appropriate exchange rate. Even at the height of
the oil boom, when Indonesia was swimming in petro-dollars and was generating
large balance-of-payments surpluses, and when the natural inclination would have
been to let the rupiah appreciate, instead the technocrats devalued the currency to
protect the traditional tradable sectors—foremost among them agriculture—and
recycled large parts of these windfall profits back into agriculture. This policy
contributed substantially to the phenomenal poverty alleviation process in the rural

14
 areas that characterized Indonesia in the seventies and eighties. (Thorbecke 1995:
34–35)

The technocrats had no political base of their own. They depended entirely on their patron,
Suharto, to implement their plans and policies. The president was an active participant in every
major macroeconomic decision, especially the timing and magnitude of changes in the exchange
rate. Professor Widjojo Nitisastro, the acknowledged dean of the technocrats, receives high praise
for his ability to manage the economic team and to consolidate their advice when he explained
pressing economic issues to the president. His close colleague, Radius Prawiro, makes the
following observation:

All the people in the economic team were intelligent and dynamic men with no
shortage of self-esteem. It was Widjojo, however, who quickly became the
unofficial yet acknowledged head of the group. Widjojo was an expert in both
demographics and economics. He was a brilliant strategist and a person of vision
with a long-term focus. He was an excellent listener and without resorting to
domination or relying on the authority of conferred title, Widjojo was able to guide
a team comprising many of Indonesia’s most capable and influential economists.
He brought out the best in each member. He kept meetings moving without
imposing rigid procedures or formality. Widjojo was a natural leader. Throughout
his career, he deliberately kept a low profile and yet if there is anyone who deserves
the title of ‘architect’ of Indonesia’s economic development, it is Widjojo. (Prawiro
1998: 83)

Despite the control of the economic team over macroeconomic policy, with the president’s equally
clear support and blessing, Suharto used trade policy to protect special interests in his circle and
even beyond, sometimes with no more apparent rationale than a nationalist interest to develop a
modern industrial capacity. The role of good economic governance and political commitment to
poverty reduction is a key lesson from this experience, but the paradox is why the autocratic
Suharto regime provided both ingredients for so long, and why the new democratic governments
have not.

Part of Suharto’s commitment to the rural economy seems to have come from the highly visible
politics, and power, of food security. The drive for higher agricultural productivity—a key
ingredient in pro-poor growth in the circumstances Indonesia faced during this period—was
fuelled at least in part by the desire for households, and the country, to have more reliable supplies
of rice than were available, at least historically, from world markets. When the world rice market
quite literally disappeared for several months during the world food crisis in 1973, Indonesia’s
dependence on imported rice to stabilize domestic prices highlighted its vulnerability to external
markets beyond its control—the opposite of food security in the minds of most Indonesians—
and showed how important it was to increase rice production (Timmer 2000). A ratcheting up of
policy attention to agriculture and budget support for rural infrastructure followed the traumatic
loss of control of rice prices in 1972–73.

From this perspective, the political economy of pro-poor growth since the tumultuous events
surrounding the Asian financial crisis seems quite perverse, especially policies with respect to the
rice economy, which are intended to generate that growth. There is, however, a rationale that
explains the new political economy. Behind the Suharto regime’s commitment to pro-poor growth
were two important constituencies: one that backed economic growth itself; and the other that
expressed concern for the poor. The growth coalition was made up of the modernizing elements
of the military, the business elite not already comfortably protected by anti-growth protectionist
measures, and most of the rural sector, which was near starvation in the mid-1960s.

15
The voices for the poor included many of this same coalition, but for somewhat different reasons.
The military was concerned about rural unrest. It did not have the coercive resources to suppress
it by force alone. The Jakarta political elite, led by President Suharto, increasingly staked its
credibility on political stability. Both the urban and rural poor could pose a threat to that, as the
1974 Malari riots demonstrated. 10 Increasingly, the donor community came to stress the
importance of poverty reduction. The World Bank made a major commitment in the late-1980s to
the analytical work that surfaced in its 1990 report on poverty (World Bank 1990).

The fortuitous intersection of the growth and poverty coalitions thus offered the Suharto regime
a political opportunity to do well by doing good. In the context of powerful opportunities to
stimulate rapid growth in rural areas through high-payoff investments in rehabilitating irrigation
systems and rural infrastructure and the importation of new rice technologies, a cumulative process
started that built both rapid growth and poverty reduction into the basic dynamics of the
Indonesian economy. But the process started in the agricultural sector.

This cumulative process appeared to have ended in the early 1980s, as prices for agricultural
commodities collapsed in world markets, oil prices declined, and the whole growth process seemed
threatened. Fortuitously, again, but under the determined guidance of the technocrats, and with
the full support of the president, the economy was restructured to make it more open to foreign
trade and investment, just as Japan and Korea came looking for opportunities to invest in labour-
intensive manufacturing facilities. Only with the economic and political collapse in 1998 did this
source of pro-poor growth disappear (and with it the patron of the technocrats). In its place, a
new populist-driven strategy of poverty alleviation has emerged.

The political appeal of the new democratic strategy for dealing with poverty—direct fiscal and
food transfers to the poor—is obvious. In principle, these transfers have immediate and visible
impact on the recipients, and the political ‘pitch’ for the programmes makes it sound as though
the government is actively committed to poverty reduction. Although democracy has probably
increased the size and influence of the political coalition concerned about poverty, it has greatly
undermined the coalition supporting economic growth as the main mechanism for dealing with it.

In the political rhetoric since 2003, poverty reduction is no longer linked to economic growth.
Instead, the agency distributing subsidized rice to the poor, Bulog, has built an ‘anti-poverty’
political coalition similar to the one that long supported food stamps in the US Congress. Support
in the USA came from conservative rural legislators eager to have additional markets for the food
that is produced in surplus by their farm constituents and from urban liberals who have in their
constituency many poor people who use food stamps as a major source of income. 11

Similarly, Bulog is the agency that procures rice domestically with budget support from parliament,
and distributes this rice to the poor at low prices that are, again, subsidized by the budget. Bulog
has mobilized political support from two constituencies concerned with poverty: first, for its rice
procurement programme, on the grounds that it helps rice farmers; and second, for its
implementation of the ‘special market operations’ (OPK) programme that delivers subsidized rice
to the poor. As Stephen Mink (2004) of the World Bank observed, no parliamentarians have been
willing to take on both dimensions of the rice programme simultaneously. As a result, the huge

10
One of the major grievances of the student rioters was the loss of control over rice prices in the previous year, and
the continuing high rice prices.
11
Political polarization in the USA now threatens this longstanding coalition between rural farm interests and interests
determined to help the urban poor.

16
budget subsidies that accrue to Bulog to run these programmes, and the corruption that
accompanies them, go unchallenged. 12

Rebuilding the economic growth coalition in support of the poor is proving difficult. It depends
on the underlying conditions of economic governance—political stability, rule of law, control of
corruption, and so on—that moved in the wrong direction for several years after democratic
governments were elected. Probably the best that can be done in the short run is to minimize
policy damage to the interests of the poor while trying to improve the effectiveness of the
programmes transferring resources directly to the poor. 13

4.2 Trends in governance at the national level

A focus on the quality of governance at the national level is an essential component of any effort
to understand, in political economy terms, the historical and emerging patterns of support for pro-
poor growth. The poor themselves seldom have an active voice in the strategies that influence the
growth process, and they rely on government officials to act on their behalf. In addition to its
‘voice’ on behalf of the poor, governance is now seen as a major factor in the growth process itself.
First, such dimensions as government effectiveness and regulatory quality influence growth
directly. Second, dimensions such as rule of law and control of corruption influence growth
indirectly and in the longer run.

Understanding the quantitative significance of this influence is difficult because the components
of governance are so hard to measure. Still, measures of voice and accountability improved
between 1996 and 2002, and this improvement signalled the switch to reasonably representative
democratic government. There have been steep declines, however, in political stability, regulatory
quality, and control of corruption (Kaufmann et al. 2003). This deterioration has immediate
ramifications for the pace of economic growth, as it affects investment in a direct manner. FDI
was negative between 1998 and 2004. A concern is that such poor governance affects the extent
to which economic growth actually reaches the poor. That is, governance itself might be a decisive
factor influencing both dimensions of pro-poor growth performance. 14

There is accumulating evidence of this connection at the regional level. Indonesia conducted a
massive decentralization of many governmental functions in 2001. Research has attempted to
judge the quality of local governance and its impact on poverty reduction between 1999 and 2002.

4.3 The role of governance at the regional level

The experiment with decentralization of government services to the kabupaten level presents
challenges and opportunities. The challenges are obvious enough, and have stimulated important
efforts by the government of Indonesia and its major donors to provide legal guidelines and

12
The most complete historical analysis of the Indonesian experience with direct food deliveries to the poor (originally
OPK, but then called RASTRA, and now about to disappear in favour of cash transfers for most recipients) is offered
by Timmer et al. (2018).
13
Early in 2018, the Jokowi government announced significant changes in how the ‘rice for the poor’ program would
be administered. Most urban poor now receive pre-paid debit cards that can be used to buy rice (and other essentials)
in local markets. Direct deliveries are still used in many areas of the Outer Islands where food markets are not as well
developed.
14
It should be noted that the analytical effort to measure the components of governance at the national level was
sponsored by the World Bank but never officially adopted. Its directors, representing the member countries directly,
objected to the often critical characterization of their own countries.

17
financial incentives to make the new structure work. The opportunities are equally obvious. Design
and implementation of many basic services in health, education, and business regulation are now
closer to the people and far more transparent. Whether they become more responsive is the key
challenge to democracy.

One opportunity for researchers that stems from the decentralization is the introduction of a great
deal of variance into local government activities, which had been precluded by the strong
centralization of power during the Suharto regime. With variance comes an opportunity to analyse
what works and what does not. Obviously, it is too early to see many systematic effects, but
researchers at SMERU have examined initial data to determine if rough measures of the quality of
local governance matter for poverty reduction (Sumarto et al. 2004).

Perhaps the most interesting result comes from a survey of 87 cities and kabupatens for which a
research team was able to construct an index of bureaucratic culture at this level. There were four
levels: ‘disruptive, less conducive, conducive, and very conducive’ in fostering the local business
environment. None of the districts had a disruptive climate, and 12 were ‘less conducive’, 61 were
‘conducive’, and 14 were ‘very conducive’ (Sumarto et al. 2004: 28).

Using kabupaten-level reports from the SUSENAS reports for 1999 and 2002, the researchers then
compared the mean level of poverty reduction for each of these categories of local bureaucratic
culture. The overall mean level of poverty reduction for all 87 cities and districts was 7.8 per cent
between the two years, but it was only 3.4 per cent in the ‘less conducive’ environments, 7.0 per
cent in the ‘conducive’ environments, and 15.1 per cent in the ‘very conducive’ environments.
Local ‘good governance’, at least as measured by this index of bureaucratic culture with respect to
business climate, is associated with an immediate impact on poverty reduction.

5 Looking forward to 2045: optimism or pessimism?

Much of this paper has been spent looking backward, trying to understand the historical, and
positive, reversal in the fortunes of the poor in Indonesia after 1967, and the challenges the poor
have faced in the transition from the authoritarian Suharto regime to real democracy. The issue
here is to look forward: can the poor place their hope in economic growth guided by democratic
governments, or will they need to seek redress through other venues?

First, a brief restatement of the results of the economic and political history of the past 50 years is
in order. The declining role of agriculture in the economy and employment is the clearest lesson,
especially because Indonesia has used investments in the rural economy and policies to increase
financial incentives for the sector as a way to connect poor households to rapid economic growth.
A significantly reduced agricultural sector, in relative terms vis-à-vis labour force and GDP
contributions, removes a number of policy instruments to help the rural poor (see Figure 1 for a
graphical depiction of the declining share of agriculture in both employment and GDP).

These historical lessons are interesting, of course, but the most important reason for spending the
time and energy to understand the often paradoxical record is to offer useful insights into the
future. In particular, how do democratic governments move forward on an agenda of pro-poor
growth? This concluding section attempts to answer that question.

Press reports out of Indonesia were not encouraging after the Asian financial crisis in 1997/98 and
Suharto’s decision to resign in mid-1998. The country had been through a historic drought. Only
imports of six million metric tons of rice in 1997/98—20 per cent of consumption—kept per

18
capita supplies at their trend level. The Asian financial crisis swept Indonesia in its fullest fury. The
dominant political regime, which had spent 30 years building a support structure to ensure its
continued survival, collapsed. When that structure collapsed, it seemed that there was nothing.

But in fact the Suharto government had left a legacy of political involvement on the part of the
entire population. Every five years, elections were held for members of parliament, who
subsequently elected the president. No serious doubts existed about the ultimate outcome of these
elections, but they were taken very seriously by the government. Jakarta routinely voted against the
government, as did Aceh in Northern Sumatra. High-level officials in the government were tasked
to ask ‘why?’ What was the government doing wrong? The end result was a population that was
used to voting, and counting on their votes being heard. When Suharto stepped down in mid-
1998, general elections were called to elect a new parliament. It was empowered to name a new
president.

Still, the question remains: how does a society cope with three such massive shocks in the span of
just a year? If 1966, in the famous words of President Sukarno, was ‘the year of living dangerously’,
what was 1998? The quiet answer, in retrospect, is ‘the year of transformation to a freer and fairer
society’. This is not the stuff of headlines. The headline grabbers are the Bali and Marriott
bombings, judicial ineptitude and outright corruption, human rights abuses by the military, and
continuing games played by Jakarta’s political elite (McBeth 2003).

Little noticed is the steady progress that is being made. If occasional jabs at foreign companies by
local courts still roil investors, the clear competence of the new constitutional court is unreported.
The petty (and not so petty) corruption of newly empowered local governments makes news and
is troublesome. But the open competition to create good business environments at the local level
will probably have more long-term impact (Ramage 2004).

Amid the gloom and doom, the historical resilience and successes of Indonesian society need to
be appreciated and valued. From a long-run perspective, the Indonesian experience with pro-poor
growth provides genuine hope that desperately poor societies can escape from the worst
manifestations of their poverty in a generation, provided appropriate policies are followed. Much
of East and Southeast Asia provide evidence for the reality of this hope. This is a direct challenge
to Myrdal’s pessimism.

This is also a critically important message for the Indonesia of the future, unsure as it is over what
path to follow under democratic governments. In its broadest outlines, that path is clear. The
three-tiered strategy of growth-oriented macroeconomic policy, linked to product and factor
markets through progressively lower transactions costs, which in turn are linked to poor
households whose capabilities are being increased by public investments in human capital, is a
general model accessible to all countries, including the future Indonesia (Timmer 2004a).

Investments in infrastructure and human capital are at the heart of the public-sector dimensions
of this model, and the Jokowi government clearly recognizes this (The Economist 2018). The pace
of investments in infrastructure to lower transactions costs and in human capital to improve the
capabilities of the poor still depends on the country’s ability to generate public revenues and
incentives for private participation (World Bank 2017). A fairer and more effective form of public
taxation is essential if the new government is to make these investments. The design of these pro-
poor mechanisms of public finance is urgent, because foreign resources, from donors and foreign
consumers of petroleum, are becoming less available, although FDI could increase substantially
with the right policies in place. There is no escaping the hard fact that these resources dictate not
only the pace of economic growth but also the extent to which that growth reaches the poor. With

19
the right pro-poor growth model in place, continued foreign assistance and private investment
could have a very high payoff in both dimensions.

Indonesia’s success from 1967 to 1997 did not just happen as the accident of market forces. It was
planned, nurtured, and largely financed by an activist government intent on reducing poverty as
rapidly as possible. Although Myrdal was pessimistic in the late 1960s that such a ‘developmental
state’ could emerge in Indonesia, he likely would have agreed that the authoritarian Suharto
government did what was necessary for economic success and rapid poverty reduction.
Democratic governments with similar commitment can do the same in the future.

References

Afiff, S., and C.P. Timmer (1971). ‘Rice Policy in Indonesia’. Food Research Institute Studies in
Agricultural Economics, Trade, and Development, 10(2): 131–59.
Alm, J., R.H. Aten, and R. Bahl (2001). ‘Can Indonesia Decentralize Successfully? Plans, Problems,
and Prospects’. Bulletin of Indonesian Economic Studies, 37(1): 83–102.
Boeke, J.H. (1946). The Evolution of the Netherlands Indies Economy. New York: Institute of Pacific
Relations.
Bolt, J., R. Inklaar, H. de Jong, and J.L. van Zanden (2018). ‘Rebasing “Maddison”: New Income
Comparisons and the Shape of Long-Run Economic Development’. GGDC Research
Memorandum 174. Groningen: Groningen Growth and Development Centre, University of
Groningen.
Booth, A. (2016). Economic Change in Modern Indonesia: Colonial and Post-Colonial Comparisons.
Cambridge: Cambridge University Press.
Cole, D.C., and B.F. Slade (1996). Building a Modern Financial System: The Indonesian Experience.
Cambridge University Press: Melbourne.
Cole, D.C., and B.F. Slade (1998). ‘Why has Indonesia’s Financial Crisis been so Bad?’ Bulletin of
Indonesian Economic Studies, 34(2): 61–66.
Economist, The (2018). ‘The hard-hat president’. 5 May: 39.
Gardner, B.L. (2002). American Agriculture in the Twentieth Century: How it Flourished and What it Cost.
Cambridge, MA: Harvard University Press.
Hill, H. (1996). The Indonesian Economy since 1966. Cambridge: Cambridge University Press.
Hill, H. (2000). The Indonesian Economy since 1966, 2nd edition. Cambridge: Cambridge University
Press.
Kaufmann, D., A. Kraay, and M. Mastruzzi (2003). ‘Governance Matters III: Governance
Indicators for 1996–2002’. Working Paper. Washington, DC: World Bank.
Lewis, W.A. (1954). ‘Economic Development with Unlimited Supplies of Labor’. The Manchester
School, 22: 3–42.
MacIntyre, A. (2003). ‘Indonesia as a Poorly Performing State?’ Canberra: Australian National
University.
McBeth, J. (2003). ‘Leadership: The Betrayal of Indonesia’. Far Eastern Economic Review. Cover story,
26 June.

20
Mellor, J.W. (2000). ‘Agricultural Growth, Rural Employment, and Poverty Reduction: Non-
Tradables, Public Expenditure, and Balanced Growth’. Paper prepared for the World Bank
Rural Week 2000, March.
Mellor, J.W. (2017). Agricultural Development and Economic Transformation: Promoting Growth with Poverty
Reduction. Basingstoke: Palgrave Macmillan.
Mink, S. (2004). ‘Comments on Presentation by Jorge Garcia-Garcia on the Costs and Benefits of
Bulog’s Commodity Interventions from 1993 to 1998’. Washington, DC: World Bank.
Myrdal, G. (1968). Asian Drama: An Inquiry into the Poverty of Countries. New Delhi: Kalyani
Publishers.
Papanek, G. (2004). ‘The Poor During Economic Decline, Rapid Growth and Crisis: The Case of
Indonesia’. Paper prepared for the USAID Project on Pro-Poor Growth conducted by DAI
and BIDE.
Prawiro, R. (1998). Indonesia’s Struggle for Economic Development: Pragmatism in Action. Kuala Lumpur:
Oxford University Press.
Ramage, D. (2004). ‘The Political Dynamics of Indonesia’s Elections’. Presentation. Asia Society
and US Indonesian Association (USINDO), 29 April.
Ravallion, M., and M. Huppi (1991). ‘Measuring Changes in Poverty: A Methodological Case Study
of Indonesia during an Adjustment Period’. World Bank Economic Review, 5(1): 57–82.
Schydlowsky, D.M. (2000). ‘Misperceived Corporate Exchange Risk and Hyperinflation in
Indonesia 1998–99’. Washington, DC: American University and Boston Institute of
Developing Economies.
Simpson, B.R. (2008). Economists with Guns: Authoritarian Development and U.S.–Indonesian Relations,
1960–1968. Stanford, CA: Stanford University Press.
Stern, J.J. (2003). ‘The Rise and Fall of the Indonesian Economy’. CID Working Paper 100.
Cambridge, MA: Harvard University.
Stern, J.J. (2004). ‘The Impact of the Crisis: Decline and Recovery’. CID Working Paper 103.
Cambridge, MA: Harvard University.
Sumarto, S., A. Suryhadi, and A. Arifianto (2004). ‘Governance and Poverty Reduction: Evidence
from Newly Decentralized Indonesia’. Working Paper. Jakarta: SMERU Research Institute.
Thorbecke, E. (1995). ‘The Political Economy of Development: Indonesia and the Philippines’.
The Frank H. Golay Memorial Lecture. Ithaca, NY: Southeast Asia Program, Cornell
University.
Timmer, C.P. (1975). ‘The Political Economy of Rice in Asia: Indonesia’. Food Research Institute
Studies, 14(3): 197–231.
Timmer, C.P. (1994). ‘Dutch Disease and Agriculture in Indonesia: The Policy Approach’. HIID
Development Discussion Paper 490. Cambridge, MA: Harvard Institute for International
Development.
Timmer, C.P. (1996). ‘Economic Growth and Poverty Alleviation in Indonesia’. In R.A. Goldberg
(ed.), Research in Domestic and International Agribusiness Management, Volume 12. Greenwich, CT:
JAI Press.
Timmer, C.P. (1997). ‘How Well do the Poor Connect to the Growth Process?’ CAER II
Discussion Paper. Cambridge, MA: Harvard Institute for International Development.

21
Timmer, C.P. (2000). ‘The Macro Dimensions of Food Security: Economic Growth, Equitable
Distribution, and Food Price Stability’. Food Policy, 25: 283–95.
Timmer, C.P. (2002). ‘Agriculture and Economic Growth’. In B. Gardner and G. Rausser (eds),
The Handbook of Agricultural Economics, Vol. II. Amsterdam: North-Holland.
Timmer, C.P. (2004a). ‘The Road to Pro-Poor Growth: The Indonesian Experience in Regional
Perspective’. Bulletin of Indonesian Economic Studies, 40(2): 173–203.
Timmer, C.P. (2004b). ‘Indonesia Country Study’. World Bank Project on Operationalizing Pro-
Poor Growth. Washington, DC: World Bank.
Timmer, C.P. (2015). ‘Managing Structural Transformation: A Political Economy Approach’. 18th
Annual WIDER Lecture. Helsinki: WIDER.
Timmer, C.P. (2017). ‘The Structural Transformation in Japan, Indonesia and Malaysia from 1880
to 2010: Part 1—Data and Trends’. Available at www.ehm.my/publications/articles/the-
structural-transformation-in-japan-indonesia-and-malaysia-from-1880-to-2010-part-1-data-
and-trends.
Timmer, C.P., and S. Akkus (2008). ‘The Structural Transformation as a Pathway out of Poverty:
Analytics, Empirics, and Politics’. Working Paper 150. Washington, DC: Center for Global
Development.
Timmer, C.P., H. Hastuti, and S. Sumarto. (2018). ‘Evolution and Implementation of the Rastra
Program in Indonesia’. In H. Alderman, U. Gentilini, and R. Yemtsov (eds) The 1.5 Billion
People Question: Food, Vouchers, or Cash Transfers? Washington, DC: World Bank.
van Leeuwen, B., and P. Foldvari (2016). ‘The Development of Inequality and Poverty in
Indonesia, 1932–2008’. Bulletin of Indonesian Economic Studies, 52(3): 379–402.
van der Eng, P. (1993a). Agricultural Growth in Indonesia since 1880: Productivity Change and the Impact
of Government Policy. Groningen: Rijksuniversiteit Groningen.
van der Eng, P. (1993b). ‘Food Consumption and the Standard of Living in Indonesia, 1880–1990’.
Economics Division Working Papers Southeast Asia 93/1. Canberra: Research School of
Pacific and Asian Studies, Australian National University.
van der Eng, P. (2000). ‘Food for Thought: Trends in Indonesia’s Food Supply, 1880–1995’. Journal
of Interdisciplinary History, 30(4): 591–616.
van der Eng, P. (2002). ‘Indonesia’s Growth Performance in the Twentieth Century’. In A.
Maddison, D.S. Prasada Rao, and W.F. Shepherd (eds), The Asian Economies in the Twentieth
Century. Aldershot: Edward Elgar.
Warr, P.G. (1984). ‘Exchange Rate Protection in Indonesia’. Bulletin of Indonesian Economic Studies,
20(2): 53–89.
World Bank. (1968). Economic Development in Indonesia: Volume I, Main Report. Jakarta: World Bank
(then known as the International Bank for Reconstruction and Development, or IBRD).
World Bank. (1990). Indonesia: Strategy for a Sustained Reduction in Poverty. World Bank Country Study.
Washington, DC: World Bank.
World Bank. (2001). Indonesia: Poverty Reduction in Indonesia: Constructing a New Strategy. World Bank
Country Study. Washington, DC: World Bank.
World Bank. (2003a). Indonesia: Beyond Macro-Economic Stability. World Bank Brief for the
Consultative Group on Indonesia. Washington, DC: World Bank.

22
World Bank. (2003b). Indonesia: Country Assistance Strategy, FY 2004–2007. Jakarta: World Bank.
(Advance copy.)
World Bank. (2004). ‘Concept Paper on Operationalizing Pro-Poor Growth’. Research Project.
Washington, DC: World Bank.
World Bank. (2017). Indonesia Economic Quarterly: Closing the Gap. October. Washington, DC and
Jakarta: World Bank.

23
 Bulletin of Indonesian Economic Studies

ISSN: 0007-4918 (Print) 1472-7234 (Online) Journal homepage: https://www.tandfonline.com/loi/cbie20

Designing Anti-Poverty Programs in Emerging

Economies in the 21st Century: Lessons from
Indonesia for the World

Benjamin A. Olken

To cite this article: Benjamin A. Olken (2019) Designing Anti-Poverty Programs in Emerging
Economies in the 21st Century: Lessons from Indonesia for the World, Bulletin of Indonesian
Economic Studies, 55:3, 319-339, DOI: 10.1080/00074918.2019.1690411

To link to this article: https://doi.org/10.1080/00074918.2019.1690411

Published online: 18 Nov 2019.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=cbie20
 Bulletin of Indonesian Economic Studies, Vol. 55, No. 3, 2019: 319–339

Indonesia in Comparative Perspective

DESIGNING ANTI-POVERTY PROGRAMS IN

EMERGING ECONOMIES IN THE 21ST CENTURY:
LESSONS FROM INDONESIA FOR THE WORLD

Benjamin A. Olken*
Massachusetts Institute of Technology

Governments of developing countries around the world have dramatically expanded

social protection programs for the poor in recent decades. In doing so, they face a
host of challenges in the targeting, design and implementation of these programs. In
this paper, I describe the results from more than a decade of collaboration with the
Indonesian government to understand how best to tackle these challenges, drawing
primarily on evidence from randomised controlled trials. I highlight results that
show the advantages of both community-based targeting and self-targeting, the
importance of tangible information about beneficiaries’ rights in minimising leak-
age, and the remarkable impacts of conditional cash transfers in the medium term.
I also describe several recent studies that use randomisation at scale to generate
policy-relevant evidence.

INTRODUCTION
Governments of developing countries around the world have dramatically
expanded social protection programs for the poor in recent decades: between 2000
and 2017, the number of developing countries with at least some type of social safety
net program expanded from 72 to 149 countries (World Bank 2017). As a result,
most countries now have some type of social protection program. Conditional
cash transfers (CCTs) are a particularly popular type of program, which typically
give cash to poor households that meet some conditions, usually related to child
health and education. More than 60 countries have some type of CCT program, up
from 27 in 2008, and more than 90 countries have some type of unconditional cash
transfer (UCT) program (Honorati, Gentilini and Yemtsov 2015).
Of course, helping poor citizens is not a new idea in developing countries. But
there has been a global shift over the past two or three decades, away from sub-
sidising basic commodities, such as food or energy, and towards more targeted
programs, where aid is given only to households deemed eligible. One reason for
this shift is that subsidies, while simple and transparent, may not actually have a

* Contact email: bolken@mit.edu.

ISSN 0007-4918 print/ISSN 1472-7234 online/18/000319-339 © 2019 ANU Indonesia Project

http://dx.doi.org/10.1080/00074918.2019.1690411
320 Benjamin A. Olken

large impact on poverty. For example, if a country subsidises fuel, but consump-
tion of fuel is proportional to income, then most of the benefits of the subsidy will
end up going to the middle or upper classes, rather than to the poor. While it is
possible to find amenities that are consumed mainly by the poor, generally the
middle and upper classes consume more of most commodities. Thus, subsidising
basic commodities is generally less efficient than targeted transfer programs for
reducing poverty. One reason this shift is happening now, rather than, say, 30 or
40 years ago, is that while implementing targeted transfer programs is logistically
more complicated than subsidising commodities, recent developments in informa-
tion technology mean that these implementation challenges can more easily be
addressed. For example, targeted programs require keeping track of a country’s
population, figuring out who is poor and making sure that people do not double-
collect—i.e. receive the transfer more times than they are entitled to.
The evolving situation in Indonesia mirrors these global trends. Over the past
two decades, Indonesia has implemented a number of social protection programs
targeted directly towards poor households, including rice subsidy programs,
Rastra/the BPNT, and their predecessors, Raskin and the OPK; direct cash trans-
fer programs, the BLT and BLSM; and CCT programs, the PKH and BSM, among
others.1 The introduction of these programs has coincided with a scaling back
of blanket subsidies, particularly of energy, in Indonesia. The trend away from
across-the-board energy subsidies towards targeted transfer programs improves
the ability of the government to ensure assistance reaches those who need it the
most. This shift also reduces the distortions associated with social protection pro-
grams, since cash grants are less distortionary than subsidies, which change the
marginal price of a good.
Implementing targeted transfers, however, creates new policy challenges. If
we want to give assistance only to poor households, how do we identify which
households are poor and should be eligible? What type of assistance should we
provide: cash, in-kind transfers or something else? Concerning implementation
and governance, how do we ensure that eligible beneficiaries actually receive the
transfers they are entitled to? And finally, when looking at the bigger picture, does
any of this matter? Are these programs effective in reducing poverty and improv-
ing well-being?
In this paper, I will focus on identifying those challenges, sharing findings from
some of the work I have done in Indonesia, with a large number of collaborators,
in order to shed light on different aspects of these questions. In particular, I will
draw out some of the lessons that we have learned from decades of work with the
Indonesian government to bring evidence to bear on these issues, for the future of
anti-poverty programs both in Indonesia and in other emerging economies facing
similar challenges.
The answers to these questions will be very different in a developing country
context, such as in the context of Indonesia, than in a developed country context.

1. Rastra (Rice for the Prosperous Population program); BPNT (e-voucher, non-cash com-
ponent of Rastra); Raskin (Rice for the Poor program); OPK (Special Market Operation
program); BLT (Direct Cash Assistance program); BLSM (Temporary Direct Community
Assistance program); PKH (Hopeful Families Program); BSM (Poor Students Assistance
program). See glossary for the Indonesian names of these programs.
Designing Anti-poverty Programs in Emerging Economies in the 21st Century 321

For example, for the question of targeting—in developed countries, the govern-
ment typically can observe almost everyone’s income, so it can target transfers to
people with low income. Granted, targeting by income may not be a perfect solu-
tion (because it essentially taxes earnings, thus potentially discouraging work; see
Moffitt 2002; Krueger and Meyer 2002). Nevertheless, targeting by income is typi-
cally not a viable solution in a developing country because governments often do
not observe income, particularly for the poor, as most of the poor are out of the tax
net (Jensen 2019). Instead, other approaches need to be designed so that they are
customised to the realities of the developing country context. But these alternative
approaches involve trade-offs, and it is an empirical question as to which methods
work best. The answer may also depend on how to define what ‘best’ means in a
particular context.
In terms of what kind of assistance to provide—cash, in-kind transfers or
something else—the answer to this question will also differ between contexts.
For example, there is less banking infrastructure in many developing countries,
which may make the distribution of cash more difficult. Supply issues may also be
important to consider. If a government provides subsidies in remote areas, but the
subsidised good is not readily available in those areas, then subsidies may affect
local prices (Cunha, De Giorgi and Jayachandran 2019).
Challenges of program implementation, governance and leakages may also be
particularly severe in many developing countries (Olken 2006). Transparency and
access to information may be challenges as well. For example, developing countries
may experience relatively low levels of literacy, particularly among poor beneficiar-
ies, who may not understand their rights under a given program.
Finally, we also need to understand the impact of these types of social pro-
tection programs on poverty and well-being. While a lot of research has been
conducted to understand the impact of these programs in the short run, under-
standing their impact over the longer term is more difficult and an important
area for new research.
For the past decade in Indonesia, my colleagues and I have tried to address
these questions in a developing country context. Throughout this process, we
have collaborated with many government colleagues and agencies in Indonesia,
including the National Development Planning Agency (Bappenas), Statistics
Indonesia (BPS), the Ministry of Social Affairs (Kemensos) and the National Team
for the Acceleration of Poverty Reduction (TNP2K) under the vice president’s
office. Through this unique partnership, we have worked together to conduct a
number of randomised policy evaluations—the gold standard of evidence. By
randomising which locations get which treatment, we can truly understand the
impacts of particular policies. This article describes our findings, the implications
for current and future anti-poverty programs in Indonesia and the implications
for other countries facing similar challenges.

TARGETING: HOW DO YOU IDENTIFY WHO IS POOR?

Targeting Methods
There are three main approaches to thinking about poverty targeting in a develop-
ing country context. The first approach is a proxy means test (PMT). In a traditional
means test, if a person’s income is below a given level, they are eligible for the
322 Benjamin A. Olken

program. When a government cannot observe income, it can instead use a PMT,
which predicts a person’s income based on variables that can be observed, creat-
ing proxies for income. The government can then determine eligibility based on
this predicted income. In other words, a PMT is based on proxies of income rather
than actual income.
To conduct a PMT, government enumerators usually go from door to door and
conduct a census of all potentially poor households’ observable assets, such as
house-building materials, vehicles, televisions, etc. Government analysts then run
a regression in a different dataset, collected for research purposes, that estimates
the relationship between the assets they can observe and the income or consump-
tion level of a household, which is observed in research data but not observed
for the entire population. Once they have that regression, they can return to their
census assets data and predict a household’s income. As such, a PMT typically
involves a large data collection effort. In Indonesia, for example, a PMT census
has occurred about once every three years since 2005 through the Social Protection
Registration (PPLS) survey. Depending on the year, the government enumerators
will visit between 15 and 20 million households to collect data on different asset
characteristics.
Moreover, the proxies created for income may be imperfect. For example,
people have different tastes and consumption patterns. Given the same level of
income, some people may prefer to spend money on a house while other people
may prefer to save money in a bank account. A PMT will estimate that people who
prefer to spend money on a house are richer than those who prefer to save money
in a bank, because bank savings are not observed in a PMT. A PMT may also not
reflect income shocks. For example, if someone is laid off from their job, they may
still own the same house and assets, but their actual income may be quite low.
None of this is to say that a PMT is not valid. Rather, these observations highlight
that the proxies are imperfect. At a technical level, one can measure the predictive
power of a PMT through the R2, which captures how well the variables in the PMT
predict actual per capita consumption. In examples we have produced, these R2
values are usually between 0.4 and 0.6, confirming that the PMT asset variables
have substantial predictive power, but that there is still substantial residual vari-
ation in per capita consumption that is not captured.
A second approach is community-based targeting, which is based on the idea
that communities may have better information than the government about which
households are the poorest. However, if the government asks a community for
this information, and links money to citizens identified as poor, elite capture may
be a concern. For example, a commonly cited concern is that local elites may try
to manipulate the process by putting their nephews or cousins on the eligibility
list instead of the people that are the most deserving. How the community defines
poverty may also be different from how the central government defines it; I will
return to this issue in more detail later.
A third approach is ‘self-targeting’, where poor households are asked to apply
for a program. The PMT that I described above entailed automatic enrolment—
government enumerators went from door to door to screen households and placed
households below a certain poverty threshold on the eligibility list automatically,
without the households needing to take any action. In a self-targeting approach,
by contrast, the government first asks anyone who is interested in the program to
Designing Anti-poverty Programs in Emerging Economies in the 21st Century 323

apply. If people apply, they are then screened with a PMT in order to determine
whether they are eligible.
There are potential advantages and disadvantages to a self-targeting approach.
In automatic enrolment, the government tries to identify all potentially poor house-
holds, but may miss a few, particularly if poor households live on the margins of
villages. With a self-targeting approach, the poor can make themselves known,
potentially reducing so-called ‘exclusion error’. This can be particularly useful
if the on-demand application process is open continually; households who were
better off, but then experienced a negative economic shock, could potentially apply
under an on-demand system. Under self-targeting, the upper or middle classes
can also screen themselves out: if someone is in a relatively well-off household,
they may not bother to apply for the program because they do not think they will
pass the PMT screening test. This type of self-selection can help the government
reduce so-called ‘inclusion error’ in the PMT formula. As discussed, the household
that chooses to keep its money in a bank rather than spend it on building a house
would perhaps pass a PMT test, but the household may not bother to apply for
the program if it does not understand the screening formula.
However, a disadvantage of self-targeting is that some poor households may
not apply. They may feel intimidated by the application process, or they may face
time constraints. For example, to apply for the program, a poor household may
need to take time off work to visit an application office, which may mean they do
not earn any money that day and may have to skip meals in order to apply.

Comparing Targeting Methods Using Randomised Trials

It is not obvious ex ante which targeting method is most effective, as they all involve
trade-offs. Hence, we conducted two different randomised evaluations to under-
stand, in an empirical manner, which of these targeting methods was the most
effective. In the first study, done with Alatas et al. (2012), we compared a PMT
approach with a community-based targeting approach and a hybrid method for
a one-time cash transfer of about $3. In the second evaluation, done with Alatas
et al. (2016), we conducted a larger experiment in the context of the expansion of
the PKH. We compared an automatic enrolment PMT with a self-targeting-based
PMT, which required households to apply before receiving the same means test.

Comparing PMT with Community Targeting

In the first evaluation, we randomly assigned 640 hamlets, or sub-villages, to one
of three targeting methods: a PMT, community targeting or a hybrid method (note
that this experiment included both rural desa and urban kelurahan; I will henceforth
refer to both of them as villages for simplicity). In the PMT, we used 49 indicators
created by BPS, using its normal PMT method. In the community method, we
facilitated community meetings in which communities ranked households from
richest to poorest. Communities first discussed what their view of poverty was, and
then community members took turns ranking everyone in the community from
richest to poorest, placing name cards on a string. This produced a complete rank
ordering of the relative perceived poverty levels of everyone in the neighbourhood.
The households ranked the poorest received the transfer. The third method was a
hybrid, where we first conducted the community-based approach and then applied
the PMT in order to screen out incompatibility. In conjunction, we also conducted
324 Benjamin A. Olken

FIGURE 1 Targeting Error under Each Targeting Method (%)

40

35

30

25

20

15

10

0
PMT method Community method Hybrid method

Note: Results show targeting error, defined as either inclusion or exclusion error, and error bars repre-
sent 95% confidence intervals from Alatas et al. (2012).

an independent survey to determine a household’s true poverty level in order to

measure which of these targeting methods was most effective at identifying poor
households.
To assess outcomes, we looked at both targeting error and community satis-
faction. Targeting error was defined as either giving the transfer to a non-poor
household or failing to give the transfer to a poor household. We measured whether
a household was actually poor by conducting a consumption survey before the
targeting process and by matching the results of that consumption survey to the
targeting results. We measured community satisfaction in a variety of ways in an
end-line survey and also established a comment box where people could leave
feedback about the targeting method used.
The results in figure 1 indicate that the PMT method had less targeting error
than the other two methods by about 3 percentage points (10%). If we look at the
data more closely, the targeting methods mostly appear to agree on which house-
holds are the rich and which are the very poor. Where the PMT does a slightly
better job of sorting households is with people very close to the threshold. Thus,
the overall impact on poverty from using one method versus another actually looks
very similar. In cases where the methods disagree, the differences in income are
very small. In other words, even though the PMT technically does a somewhat
better job of reducing targeting error, our calculations suggest that the PMT would
not reduce poverty substantially more than the other targeting methods if it were
scaled up.
We also asked households which targeting method they preferred and how well
they thought each method worked. Did they think the method was reasonable?
Designing Anti-poverty Programs in Emerging Economies in the 21st Century 325

FIGURE 2 Impact of Three Different Targeting Methods on Community Satisfaction

2.0
PMT method Community method Hybrid method
1.8

1.6

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0
No. of households that No. of households that No. of complaints in
should be added to the should be removed the suggestion box
beneficiary list from the beneficiary list

Note: Results show measures of community satisfaction, and error bars represent 95% confidence
intervals from Alatas et al. (2012).

Would they prefer to use this method again if there were another targeting project
in their village? We also showed participants the final list of eligible beneficiaries and
asked if there was anyone they thought should be added to or removed from the list.
The results in figure 2 indicate that community targeting led to much higher
community satisfaction and better-selected households that self-identified as poor.
Under the community targeting method, the community wanted to make relatively
fewer changes to the beneficiary list and also submitted fewer complaints, about
a third lower than with the PMT.
The reason for these differences was not that the community performed worse
or did not have local information. Rather, it appears that the community had more
information than the central government and had chosen to weight this information
differently. In other words, the community’s definition of poverty differed slightly
from the government’s definition, which was based on per capita consumption.
For example, widows self-identified as poorer than other people with the same
per capita consumption level, perhaps reflecting a lower earning ability in their
households. Communities also agreed and included widow-headed households
on their targeting lists at a higher rate.
Not surprisingly, if the community has a different notion of what poverty is,
and it targets based on its own definition, then it will also think it does a better job
of targeting. Ultimately, the trade-offs between these two approaches come down
to whether the government wants to target strictly based on per capita consump-
tion, which is the typical measure used by the government to measure poverty, or
326 Benjamin A. Olken

whether it wants to recognise that communities have slightly different perceptions

of poverty, and include those as well.
A second potential reason for the higher level of satisfaction with the commu-
nity-based approach may be that it is a much more transparent method than a PMT.
Because a PMT is based on complicated formulas, it is inherently less transparent
than a community-based approach. For example, a household may have a televi-
sion but also have a dirt floor and a thatched roof, while another household may
have no television and a cement floor and a tin roof. Under a PMT test, a house-
hold likely has no idea why those variables are relevant and how they combine to
determine eligibility, and hence why one household receives the program while
the other does not. We also found no evidence of elite capture in the community-
based approach.
Taken together, these results imply that there are important trade-offs to con-
sider when choosing a targeting approach, and there can be benefits to adding a
community-based approach to existing targeting methods.

Comparing Self-targeting with PMT

In the second targeting evaluation, we compared an automatic targeting approach
with an on-demand application in the context of the expansion of the PKH. The
PKH, Indonesia’s CCT program, targets the very poorest people in Indonesia.
Targeting for this program has high stakes, as households receive about 11% of
their consumption, or about $900, over six years.
For the study, a facilitator visited each village and explained that the program
was beginning and was going to screen for poverty, without giving the exact screen-
ing formula. Under the automatic PMT method, the government conducted a
PMT from door to door and automatically enrolled households below the poverty
threshold. Under the on-demand application method, households were required
to apply for the program in advance. If people were interested in the program,
they had to apply at a specific time and nearby location, either at the subdistrict
(kecamatan) or the village (desa/kelurahan) office or in their hamlet (dusun). If they
applied, their poverty status was verified by a PMT, and for those households close
to the margin of eligibility, by a follow-up home visit as well.
The results in figure 3 indicate that requiring households to apply for benefits
led more poor and fewer non-poor households to receive benefits compared with
automatic screening. In other words, in on-demand-application villages a person
had a higher probability of receiving benefits if they were very poor and a lower
probability of receiving benefits if they were non-poor.
There are several different mechanisms driving these impacts. First, in the auto-
matic enrolment villages, it is possible that, despite the best efforts of government
enumerators, they may have missed some poor households during the PMT iden-
tification process. In comparison, in the on-demand approach, any poor household
could make itself known and apply, leading more poor households to receive the
program and to a reduction in exclusion error.
Second, wealthier, ineligible households appear to have chosen not to apply.
This can have significant budget implications, because most of the population
is ineligible (see the histograms in the background in figure 3). In general, since
there are errors in the PMT, a small fraction (about 2%–3%) of households over
the eligibility threshold end up on the eligibility list. While there is a very small
Designing Anti-poverty Programs in Emerging Economies in the 21st Century 327

FIGURE 3 Distribution of Beneficiaries under Self-targeting vs. Automatic Screening

0.8 Automatic screening (rhs) 0.20

Probability of receiving benefits

Self-targeting (rhs)
Poverty line
0.6 0.15
Log consumption
distribution (lhs)
Density

0.4 0.10

0.2 0.05

0 0
11 12 13 14 15 16
Log per capita consumption

Notes: Author’s calculations showing probability of receiving benefits by per capita consumption level
and treatment, using data from Alatas et al. (2016). The light-grey histogram shows the overall distribu-
tion of log per capita consumption.

probability that these households receive the benefit, since 90% of the population is
ineligible, the 2%–3% of ineligible households that pass the PMT screen by mistake
actually make up a relatively large proportion of total beneficiaries. In comparison,
under self-targeting, many of those ineligible households choose not to apply. In
particular, under self-targeting, 61% of eligible households apply, while only 10%
of ineligible households apply, indicating that households take their own income
status into account when deciding whether to apply.
These two effects combined meant that the beneficiaries who were selected
were 20% poorer on average in the villages with on-demand application than in
the villages with automatic enrolment. On-demand application, even though it
has some potential risks, seems practical and effectively reduces both inclusion
and exclusion errors.
One important issue is that different targeting methods may work better in dif-
ferent types of areas. To test this, both our targeting studies—comparing PMT with
both the community-based approach and the on-demand application approach—
took place in a mix of urban/peri-urban and rural environments. In most cases,
the results are broadly similar in these environments. Although the differences
between results from both PMT and the community approach and PMT and the
on-demand approach are somewhat less pronounced in urban areas, the differ-
ences between urban and rural areas are not statistically significant in most cases.

The (Lack of) Distortionary Effects from PMT Targeting

A third question we examined is whether knowledge of the variables used to
determine eligibility for benefits under the PMT affects households’ investment
decisions. In other words, does telling people about the assets that determine their
eligibility for benefits distort their investment decisions?
328 Benjamin A. Olken

To investigate this question, my colleagues and I have been working with BPS
to conduct a randomised evaluation, using a survey in 2015 for the national data-
base for social assistance (PBDT) (Banerjee, Hanna, Olken and Sumarto 2018). In
this survey, the government randomly varied the screening questions asked in
different provinces. To keep the number of questions in the survey constant, each
randomised question had one of two options. In half of the provinces, households
received (1) either a question on flat-screen television ownership or a question on
the number of rooms in the house (randomised by province), and (2) either a ques-
tion on the number of active mobile phone SIM card numbers in the house or on
whether the house had a modern toilet installed (also randomised by province).
We specifically chose these two key treatment questions—on flat-screen television
and SIM card ownership—because we had access to independent data sources on
actual asset ownership that did not rely on household self-reports.
We found that people appeared to be paying attention, but the distortionary
effects seemed limited. In provinces where the PBDT questionnaire asked about
television ownership, households were about 16% less likely to report owning a
flat-screen television in the next round of Susenas, carried out six months later.
However, by the next year, this effect had disappeared. More importantly, actual
television sales—as measured by an independent survey of television retailers—
did not appear to decline. We also saw no changes in any of the other assets we
tracked. Thus, while there is some evidence that asking those asset questions may
change how people strategically respond in a government survey, it does not seem
to actually change consumption behaviour.

Is Targeting Worth It?

The final topic I discuss in this section is the decision about whether to target at
all. There has been tremendous attention in the world recently on a universal basic
income (UBI). The general idea of a UBI is that instead of trying to figure out who
is eligible, the government simply gives the transfer to everybody. Such a scheme
is still redistributive: if a country taxes in proportion to consumption, the rich are
still paying substantially more than the poor. So, if everyone is receiving the same
transfer, the net effect of the scheme is to redistribute towards the poor.
The main downside of a universal transfer is that the beneficiaries receive less
per person. For a given budget, if the government targets only 10% of the popu-
lation, it can give each beneficiary 10 times more than if it gives the transfer to
100% of the population. In other words, when countries have a fixed budget, the
transfer per person mechanically decreases when more people are included. With
a universal transfer, the government does save money because it does not have
to run a poverty census, but it turns out that the cost of running a poverty census,
even though it is large in dollar terms, tends to be very small compared with the
budget of the transfer program itself.
Another benefit of a universal transfer is that the government saves the cost
of targeting. While intuitively this saving may seem large—after all, government
enumerators go from door to door to survey tens of millions of households—the
saving pales in comparison to the magnitude of the benefits delivered through
targeting. To be specific, the 2015 Indonesian poverty census cost the government
about Rp 900 billion. The government typically conducts a new census once every
three years; the annualised cost of targeting is therefore about Rp 300 billion. By
Designing Anti-poverty Programs in Emerging Economies in the 21st Century 329

FIGURE 4 Social Welfare vs. Inclusion Error Trade-off:

Simulations for Indonesia and Peru

−4.60e-07
−0.24
Indonesia (lhs)
CRRA utility: Indonesia

−4.80e-07 −0.25

CRRA utility: Peru

Peru (rhs)
−0.26
−5.00e-07
−0.27

−5.20e-07 −0.28

−0.29
−5.40e-07
0 0.2 0.4 0.6 0.8 1 = UBI
Inclusion error

Notes: Reproduced from Hanna and Olken (2018). For each cutoff value c, we calculate the per capita
benefit amount for included households and then calculate the constant relative risk aversion (CRRA)
utility with ρ = 3. If a household is not included in the program at a given value of c, it us assumed bi = 0.
Dashed lines indicate the point of maximum social welfare in each country. Inclusion error on the x-axis
shows the inclusion error from each potential cutoff value c. UBI = universal basic income.

contrast, in 2019, total expenditures on targeted social protection expenditures

(i.e. the value of all transfers allocated using the targeting census) amounted to
Rp 98 trillion (TNP2K 2019). The annualised cost of running the targeting census,
expressed as a fraction of the total amount of assistance being targeted, is therefore
about three-tenths of 1%. Given this, the benefits of not having to run the targeting
census tend to be swamped by other considerations.
To quantify the remaining trade-offs, Rema Hanna and I conducted a simulation
where we considered different targeting approaches and their related targeting
errors, holding the budget constant, and compared them with a universal program
(Hanna and Olken 2018). How well a targeted program performs depends in part
on the accuracy of the targeting. If the targeting is very accurate, then a targeted
transfer will perform relatively well at improving social welfare. If the targeting is
not very accurate, then a universal transfer might perform better.
Figure 4 shows the results for simulations we conducted using data from both
Indonesia and Peru. The results indicate that a program that is relatively narrowly
targeted (i.e. one that targets about 19% of the population, which is reasonably close
to the current Indonesian national targeting policy for many programs) achieves
a much higher level of social welfare than universal programs. This result holds
even when accounting for the cost savings from not having to run a poverty census,
which, as discussed above, turn out to be very small in comparison to the transfers
given out. In other words, targeted transfers, while imperfect, deliver a far greater
improvement in welfare than a UBI would, because they can transfer much larger
330 Benjamin A. Olken

amounts per beneficiary. The downside is that such a program does miss deserving
people: in our calculations, exclusion error from such a program would be about
30% in Indonesia and about 20% in Peru.
In sum, findings from these studies on different targeting methods suggest
that community feedback can improve targeting, in the sense that it can improve
the perceived legitimacy of targeting lists, and citizens’ satisfaction with them, by
making them conform better to local perceptions of poverty. Adding on-demand
application to a PMT can improve targeting, by reducing both inclusion and exclu-
sion errors. We do not see any distortionary effects in household consumption
or investment behaviour from the PMT surveys. And when countries have fixed
budgets, targeted transfers—at least based on our simulations—look much more
effective than universal transfers for helping the poor and improving social welfare.

WHAT TYPE OF ASSISTANCE SHOULD WE PROVIDE?

What type of assistance should we provide: cash, vouchers or in-kind transfers?
With each type of transfer, there is a trade-off between placing restrictions on how
transfers can be used versus giving people more choice. Transferring cash allows
the most choice for individuals. In-kind transfers, where the government gives
recipients the good itself, offer the most government control. Vouchers are an in-
between option, where the government typically allows people to choose where
to spend the transfer, but only for a restricted set of goods.
Each of these options has pros and cons. Cash is very flexible, but there are
political concerns about how people will spend the money, and concerns that prices
may rise if there is limited supply. In-kind transfers can improve the supply of the
subsidised good, which can be helpful in a limited-supply environment, but they
also raise concerns about quality and leakages. In addition, people may not need
the particular good they are given and may value it less than cash, which they can
use to buy what they need most. To the extent that the goal of the transfers (CCT
programs, for example) is to stimulate health and educational investments, other
questions arise. In particular, different approaches may be needed in remoter areas
with less-developed supply. I will discuss some of Indonesia’s pioneering work to
adapt ideas about CCTs to these kinds of environments.

Some Global Evidence

There is some existing global evidence addressing these questions. For example,
Cunha, De Giorgi and Jayachandran (2019) conducted a randomised trial in Mexico,
where the government randomised whether beneficiaries in particular communi-
ties received in-kind transfers or cash. They found that in isolated communities,
giving out cash caused prices for basic commodities to increase. The cash transfers
created a demand shock and because of limited supply, caused prices to rise. In
comparison, giving out in-kind assistance actually lowered prices in isolated com-
munities, acting as a positive supply shock. In Indonesia, one can imagine that
price effects would not be a large concern in areas with thick markets, such as
Java. However, in remoter areas, price effects could be very important to consider.
Overall, this evidence suggests that existing supply constraints should be consid-
ered when switching from in-kind transfers to cash transfers in isolated areas. We
are exploring these questions in ongoing work in Indonesia.
Designing Anti-poverty Programs in Emerging Economies in the 21st Century 331

A second study by Aker (2017) compared cash transfers with vouchers in a ran-
domised evaluation in an internally displaced person (IDP) camp in the Democratic
Republic of the Congo. The study found little difference in overall well-being under
either method, in part because people purchased similar goods. In urban Ecuador,
Hidrobo et al. (2014) conducted a randomised evaluation to compare cash with
in-kind transfers and with vouchers and found that all three types of transfers
improved the quantity and quality of food consumption. The in-kind food transfers
increased calories the most, while the vouchers increased dietary diversity the most.
One other finding is worth noting. Across these studies, and in many other
randomised evaluations on cash transfers, there is no evidence for the commonly
cited political concern that people spend cash transfers on ‘temptation goods’, such
as tobacco or alcohol.2 However, there may be other important concerns about
transferring cash. For example, the government may want beneficiaries to spend
the cash on food or a specific good, or supply may be limited.

Conditional Community Transfers in Indonesia

An important idea of conditional cash transfers is that by incentivising health and
education, they may stimulate demand for these services. But in areas that are
supply constrained, stimulating demand for these services (which could lead to
price increases if private providers are involved) may not necessarily increase
actual service uptake. For remote locations, another approach may make sense.
Indonesia again was a leader worldwide in thinking about these issues. In 2005,
Indonesia began planning the introduction of a CCT program in a working group
led by Bambang Widyanto, then at Bappenas. Given the concerns raised above,
Indonesia chose to pilot two different types of programs: a traditional, household-
based CCT program, and a new program aimed at more-rural locations and that
moved cash transfers to the community level rather than the individual level. The
idea behind the community-level grants was to directly increase supply in more-
rural locations—e.g. a village could use the funds to hire a midwife or to start a
remote-school classroom (kelas jauh)—rather than to stimulate demand through
household cash transfers. This program was launched in 2007 and called the
National Community Empowerment Program for a Healthy and Smart Generation
(PNPM Generasi Sehat dan Cerdas, also known as Generasi). The idea of incentives
from CCTs was incorporated into Generasi. Each village received points for the
various health and education services it obtained, such as for each child immunised,
each mother who received maternal care, each child enrolled in school, and so
on, following the same set of conditions used in the household CCT. Villages that
earned more points in each kecamatan would get a larger block grant the next year.
Moreover, remarkably, the government chose to evaluate both programs using
randomised controlled trials, at the kecamatan level. That is, the government first
drew up a list of locations where the programs could take place. As funds were
limited, not all kecamatan could receive the programs. The government therefore
decided to randomly select which kecamatan would receive the program—176
kecamatan for Generasi and about 438 kecamatan for the PKH—and which kecamatan

2. See Aker (2017); Banerjee et al. (2015); Banerjee et al. (2017); Blattman, Jamison and
Sheridan (2017); Evans and Popova (2016); and Haushofer and Shapiro (2016).
332 Benjamin A. Olken

would serve as control groups. Each of these evaluations separately represented

one of the largest randomised evaluations conducted anywhere in the world at the
time. For Generasi, the government chose to evaluate both the program with incen-
tives, as described above, and an identical program without incentives, in order to
test if the incentives mattered. I worked most closely on the Generasi evaluation,
along with Junko Onishi and Susan Wong from the World Bank, while Vivi Alatas
from the World Bank led the team on the PKH evaluation.
The CCT program found positive results after two years—consistent with
evidence from around the world (Alatas 2011). It is important to note that the
hypothesised supply-side concerns were important. Even though the PKH was
focused on ‘supply-side-ready’ locations, Triyana (2016) found that in the short
run, Indonesia’s CCT program (the PKH) increased the delivery fees charged by
midwives. However, this did not prevent the program from increasing medical
care overall. The number of midwives increased and so did the level of medical
care given.
What about Generasi? We found that the program was effective in increasing
the use of maternal and child health services, particularly weight checks for young
children, and resulted in reductions in malnutrition (Olken, Onishi and Wong
2014). The program also increased school enrolments. The incentives in Generasi
sped up how quickly the program results materialised. The program was most
effective in the areas with the lowest levels of service delivery at baseline, such
as rural locations in East Nusa Tenggara. Our calculations suggest that the cost-
effectiveness of the program was similar to cost-effectiveness of the PKH, at least
in the first two years.
These two evaluations suggest that ‘one size may not fit all’. That is, different
approaches may work best in different areas. Programs like Generasi may be par-
ticularly appropriate in areas with relatively low levels of service supply. In terms
of policy impact, both programs, the PKH and Generasi, were subsequently scaled
up. However, the scale-up of Generasi took this evidence in mind and focused
primarily on the more rural and remote locations, where the program was found
to be most effective.

How Do We Ensure that Assistance Reaches Eligible Families?

Like governments in many developing countries, the Indonesian government
wants to ensure that all of the benefits from social protection programs actually
reach the intended beneficiaries. However, challenges with leakages can reduce
program impacts. How do we reduce leakages and ensure that all program benefits
are reaching the targeted beneficiaries? In particular, can strengthening access to
information improve service delivery?
To study this question, my colleagues and I conducted a randomised evaluation
of Indonesia’s Raskin/Rastra program and studied whether providing people with
information on program benefits increased the amount of benefits that eligible
households received (Banerjee, Hanna, Kyle, Olken and Sumarto 2018). Raskin,
as it was called at the time, was a program where the poorest 30% of households
were entitled to about 15 kg of rice per month at about one-fifth of the market
price. However, there were substantial leakages in the implementation of this
program, both with ineligible households receiving subsidised rice and with the
actual delivery of the rice.
Designing Anti-poverty Programs in Emerging Economies in the 21st Century 333

At the time we started, the government was considering introducing identifica-

tion (ID) cards as tangible proof for beneficiaries of what their rights were under
the program. Providing this information to citizens could both inform them of their
eligibility and entitlements under the program and enable them to better bargain
for their rights with local officials. However, it was unclear whether introducing
the cards would actually work and be worth the potential economic and social
costs. So, the government invited us to work with it to conduct a randomised
policy evaluation to determine whether these ID cards would actually improve
the functioning of the program.
In addition to testing the impact of the ID cards overall, the government wanted
to know what variation of card would work best. We therefore evaluated several
variations, such as cards containing information on just the quantity of subsidised
rice that households were entitled to or cards containing both the quantity and
price of the subsidised rice that households were entitled to; we also evaluated
whether beneficiary lists should be posted publicly. We designed this evaluation
explicitly in collaboration with the government to help inform its choice of pro-
gram design.
Figure 5 indicates that the ID cards increased the subsidy that eligible house-
holds received by about 26%, relative to the subsidy received in comparison villages
where ID cards were not mailed out. Interestingly, ineligible households received
no less, indicating that there was an overall reduction in leakage from the program,
rather than just a transfer from ineligible households to eligible ones.
The content of the cards also mattered. Conditional on sending out the cards,
the impact of the subsidy was doubled in villages where the cards included both
price and quantity information, compared with villages that received ID cards with
only quantity information. Publicising beneficiary lists also increased the subsidy
received by eligible households by about 25%, compared with villages that received
ID cards without public information.
These results indicate that information matters, even for a program such as
Raskin, which in various forms has been around since 1999. Giving people tangible
information about their rights seems to be an important component in ensuring
they receive what they are entitled to. These positive results were some of many
factors that led the government to scale up this program. In 2013, the government
provided social protection ID cards to 15.5 million poor households, reaching 65.67
million people. This is an example of how research can have concrete implications
for policy, both in Indonesia and more generally.

WHAT IS THE IMPACT OF ASSISTANCE ON POVERTY AND WELL-

BEING? EVIDENCE FROM CONDITIONAL CASH TRANSFERS
Global Evidence
What is the impact of transfer programs on poverty and well-being? I will focus
here on CCTs, which, as described, are a common type of social protection program.
CCTs are typically long-term programs where transfers are given regularly to
poor households on condition that they make human capital investments in their
children. For example, for expectant parents, the conditions may include prenatal
care, delivery by a trained medical professional and so on. After the child is born,
the conditions may include early childhood health investments and enrolment in
334 Benjamin A. Olken

FIGURE 5 Impact of Raskin Cards on Subsidy Received

per Eligible Household (Rp/month)
40,000

35,000

30,000

25,000

20,000

15,000

10,000

5,000

0
Comparison group Cards

Notes: Results show the average amount of benefits received by eligible households in treatment and
control villages, and error bars represent 95% confidence intervals from Banerjee, Hanna, Kyle, Olken
and Sumarto (2018).

primary and secondary school. CCTs generally aim to improve the welfare of not
only the recipients but also their children, by improving the children’s human
capital accumulation.
CCTs began in the 1990s in Mexico, Bangladesh and Brazil and have since
spread worldwide. In 2014, over 60 countries had some type of CCT program. Some
rigorous research has been conducted, in part by J-PAL affiliates, on the impact
of CCT programs in the short run. This research has generally found that CCTs
are effective in improving incentivised outcomes (See Gertler 2004; Shultz 2004;
Baird et al. 2010; Baird et al. 2014). There have also been studies on whether con-
ditions are even necessary, as they are costly to enforce. For example, in Morocco,
Benhassine et al. (2015) compared a CCT with a labelled cash transfer (LCT), where
the programs were described similarly but conditions were enforced under the
CCT and not enforced under the LCT. They found that the outcomes of the CCT
and LCT programs were very similar; adding conditionality in this context made
almost no difference.
In comparison, in Malawi, Baird, McIntosh and Özler (2011) conducted a ran-
domised evaluation to compare unconditional cash transfers (UCTs) with CCTs,
and they found that there were trade-offs with conditionality. Like other research-
ers, they found that CCTs were effective at incentivising behaviour, but they also
found that UCTs were more effective for certain types of individuals. For exam-
ple, they found that UCTs were more effective in helping girls to delay marriage
and childbearing. They surmise that the CCTs successfully incentivised some
Designing Anti-poverty Programs in Emerging Economies in the 21st Century 335

households to meet the transfer conditions. However, there were other types of
households that were so far from being able to complete the conditioned activi-
ties that they did not respond to the incentive. Such households might have really
benefitted from the cash transfer. In those cases, enforcing conditionality stopped
cash transfers to those households, which had a countervailing negative effect on
that class of household. While these results confirm that CCTs improve indicators
of incentivisation, they also indicate that enforcing conditions may have negative
consequences, such as excluding benefits from households that are unable to meet
conditions for various reasons.

Medium-term Evidence from Indonesia

Most of the existing evidence on cash transfers looks at the short run, but what
happens in the longer run? As noted by Santiago Levy, a pioneer of the Progresa-
Oportunidades CCT program in Mexico, ‘clearly achieving good health is a
cumulative process, temporary investments in nutrition are of little help. The same
is true of education: children must be supported year after year’ (Levy 2006). This
helps to explain why CCT programs typically determine eligibility once and then
give transfers to households for many years, as the programs are intended to
support children as they grow up. However, it is difficult to study the long-term
impact of CCTs. In most cases, the government gave the program to the comparison
group after a relatively short period, making it difficult to learn about the long-
run impacts of these types of programs. For example, in the well-known case of
Progresa-Oportunidades in Mexico, the control group received the program just
18 months after the study began.
In Indonesia, Cahyadi et al. (2018) conducted a medium-run study of the impacts
of the PKH. As described above, in the original program design, 736 kecamatan were
randomised either to receive the PKH transfers or to serve as a comparison group.
The two-year results were analysed by Alatas (2011).
It turns out that as the government expanded the PKH after the initial evaluation,
it prioritised expanding the program to new provinces and districts in order to
ensure that it spread nationwide, rather than prioritising the original comparison
group. Thus, without any researcher intervention, by 2013, more than six years after
the program had started, all the original treatment locations from the first evalua-
tion were still receiving the PKH transfers, and 60% of the original control group
remained untreated. This created a unique opportunity to measure the impact of
the CCT program over the longer run.
We found substantial improvements in health and education behaviour in both
the short run and the long run. For example, the PKH improved maternal deliv-
ery behaviours after two years, and the effects increased in magnitude after six
years. While there was no impact on child immunisations after two years, the PKH
increased child immunisations after six years. Similarly, for school enrolment of
older students (aged 15–17), there were no effects after two years, but there were
positive effects after six years. This may be because the original program had not
been in place long enough to affect enrolments among older students. However,
our study suggests that after six years, the program was able to increase enrol-
ments among older students who had been exposed to the program year after year.
Perhaps most remarkably, after six years there was some evidence of cumula-
tive health impacts, particularly reductions in stunting. Stunting is a major policy
336 Benjamin A. Olken

challenge in Indonesia and can be reduced only through prolonged attention to

weight and nutrition over a child’s early life cycle. Our results indicate that the
PKH actually reduced stunting after six years. This may be because the children
evaluated in the long-term study benefitted from the PKH for a substantial part
of their early childhood, where they and their mothers were supported by the
program for six years through cash transfers conditional on better maternal health
and childhood investments.
Overall, what we found was not that the PKH is radically transforming the
economic well-being of households. Rather, the evidence suggests that sustained
investments in these households over time can lead to better human capital invest-
ments in their children. This in turn could help break the intergenerational cycle
of poverty and improve things in the longer run.
We also reassessed Generasi in the same way, almost 10 years after it had started.
As with the PKH, when Generasi was expanded, the government prioritised new
areas of the country (and, in particular, remoter areas, following the results of the
original evaluation). Consequently, the original control group received no special
priority in receiving the program, but the program continued in treatment areas,
creating an opportunity to evaluate its longer-term impacts.
We found that as Indonesia experienced economic growth, supply-side condi-
tions improved in both the treatment and control areas. Many of the original areas
that were initially supply deficient were no longer so 10 years later. As a result,
the program, which was most effective in areas with low levels of service delivery
at baseline, had substantially muted effects 10 years later (Olken and Sacks 2018).
The program was still effective in the areas with the lowest levels of service deliv-
ery, particularly in terms of improving child weight checks, immunisations and
vitamin A levels. However, in general, the changing health and education environ-
ment in Indonesia suggests that the optimal policy mix for a given place may need
to change too. The contrasting results—with increasing effects over time for the
PKH but decreasing effects over time for Generasi in most locations—suggest the
importance of continuing to gather evidence on program effectiveness over the
longer term, as the environments in which programs operate evolve.

CONCLUSIONS
Throug these examples, I have described an iterative process between the Indonesian
government and both domestic and international scholars in order to identify key
policy challenges and generate evidence to improve the design of social programs.
During more than a decade-long partnership, we have generated policy lessons
for improving the targeting of social protection programs, strengthening transpar-
ency and reducing leakages, moving from in-kind transfers to cash transfers, and
addressing policy challenges such as stunting. As the challenges have evolved,
we have worked closely with the government to answer new policy questions in
a rigorous way. As a result, Indonesia has become a leader in generating rigorous,
policy-relevant evidence for the rest of the world.
The results of the process described here are therefore useful far beyond
Indonesia. Indeed, part of the iterative process is that the topics we prioritised for
new studies are those that are not only immediately relevant to Indonesian policy-
makers but also helpful in contributing to global knowledge. The work in Indonesia
Designing Anti-poverty Programs in Emerging Economies in the 21st Century 337

has led to renewed academic interest in related targeting questions, which have
taken place in multiple countries in sub-Saharan Africa (Dupas et al. 2016; Dizon-
Ross, Dupas and Robinson 2017) and elsewhere. As other countries reform their
targeting systems, the lessons learned from Indonesia can help inform them as well.

ACKNOWLEDGEMENTS
The results described in this paper draw on my collaborations with many people,
including Vivi Alatas, Abhijit Banerjee, Nur Cahyadi, Gabriel Kreindler, Jordan
Kyle, Junko Onishi, Rizal Adi Prima, Ririn Purnamasari, Audrey Sacks, Elan
Satriawan, Sudarno Sumarto, Ekki Syamsulhakim, Julia Tobias, Matthew Wai-
Poi, Susan Wong and particularly Rema Hanna.
I thank our many partners in the Indonesian government, particularly Rudy
Prawiradinata, Pungky Sumadi, Maliki and Vivi Yulaswati at Bappenas, Bambang
Widianto at TNP2K, and many partners at BPS, the Ministry of Home Affairs and
the Ministry of Social Affairs—for their openness to using evidence to inform policy
decisions and for their support of the projects described here.
I thank the many staff members at J-PAL Southeast Asia for outstanding work
implementing these projects, including Talitha Chairunissa, Masyhur Hilmy, Amri
Ilmma, Nurzanty Khadijah, Chaerudin Kodir, Nadira Purdayinta, Héctor Salazar
Salame, Ritwiki Sarkar, Prani Sastiono, Freida Siregar, Jurist Tan, Poppy Widyasari
and particularly Lina Marliani; Gregorius Kelik Agus Endarso and Yulia Herawati
for their help with the Generasi project; and SurveyMeter and Mitra Samya for their
partnership on many of these projects. I also thank Yuen Ho, Evelina Smirnitskaya
and Claire Walsh for help preparing this paper.
Finally, I thank Sudarno Sumarto for hosting me as a young graduate student at
SMERU, studying the OPK/Raskin program in the summer of 2001; Lant Pritchett
for introducing me to some of the challenges of social assistance in Indonesia in
the fall of 2000; the Henry Luce Foundation for sending me to Indonesia in the first
place in 1997–98; and Scott Guggenheim for almost two decades of unwavering
support and encouragement.
Funding for the projects described here came from the Australian government
(AusAID and DFAT), the Royal Netherlands Embassy, 3ie and the World Bank. All
the views expressed in the paper are mine and do not necessarily reflect the views
of any of the institutions or individuals acknowledged here.

REFERENCES
Alatas, Vivi. 2011. Program Keluarga Harapan: Impact Evaluation of Indonesia’s Pilot Household Con-
ditional Cash Transfer Program (English). Washington, DC: World Bank. http://documents.
worldbank.org/curated/en/589171468266179965/Program-Keluarga-Harapan-impact-
evaluation-of-Indonesias-Pilot-Household-Conditional-Cash-Transfer-Program.
Aker, Jenny C. 2017. ‘Comparing Cash and Voucher Transfers in a Humanitarian Context:
Evidence from the Democratic Republic of Congo’. World Bank Economic Review 31 (1):
44–70.
Alatas, Vivi, Abhijit Banerjee, Rema Hanna, Benjamin A. Olken and Julia Tobias. 2012. ‘Tar-
geting the Poor: Evidence from a Field Experiment in Indonesia’. American Economic
Review 102 (4): 1206–40.
338 Benjamin A. Olken

Alatas, Vivi, Abhijit Banerjee, Rema Hanna, Benjamin A. Olken, Ririn Purnamasari and
Matthew Wai-Poi. 2016. ‘Self-targeting: Evidence from a Field Experiment in Indone-
sia’. Journal of Political Economy 124 (2): 371–427.
Baird, Sarah, Ephraim Chirwa, Craig McIntosh and Berk Özler. 2010. ‘The Short-term
Impacts of a Schooling Conditional Cash Transfer Program on the Sexual Behavior of
Young Women’. Health Economics 19 (S1): 55–68.
Baird, Sarah, Francisco Ferreira H.G., Berk Özler and Michael Woolcock. 2014. ‘Conditional,
Unconditional and Everything in Between: A Systematic Review of the Effects of Cash
Transfer Programmes on Schooling Outcomes’. Journal of Development Effectiveness 6 (1):
1–43.
Baird, Sarah, Craig McIntosh and Berk Özler. 2011. ‘Cash or Condition? Evidence from a
Cash Transfer Experiment’. Quarterly Journal of Economics 126 (4): 1709–53.
Banerjee, Abhijit, Esther Duflo, Nathanael Goldberg, Dean Karlan, Robert Osei, William
Parienté, Jeremy Shapiro, Bram Thuysbaert and Christopher Udry. 2015. ‘A Multifaceted
Program Causes Lasting Progress for the Very Poor: Evidence from Six Countries’. Sci-
ence 348 (6236): 1260799.
Banerjee, Abhijit, Rema Hanna, Gabriel Kreindler and Benjamin A. Olken. 2017. ‘Debunking
the Stereotype of the Lazy Welfare Recipient: Evidence from Cash Transfer Programs’.
World Bank Research Observer 32 (2): 155–84.
Banerjee, Abhijit, Rema Hanna, Jordan Kyle, Benjamin A. Olken and Sudarno Sumarto. 2018.
‘Tangible Information and Citizen Empowerment: Identification Cards and Food Subsidy
Programs in Indonesia’. Journal of Political Economy 126 (2): 451–91.
Banerjee, Abhijit, Rema Hanna, Benjamin A. Olken and Sudarno Sumarto’. 2018. ‘The (Lack
of) Distortionary Effects of Proxy-Means Tests: Results from a Nationwide Experiment
in Indonesia’. National Bureau of Economic Research (NBER) Working Paper No. 25362,
December.
Benhassine, Najy, Florencia Devoto, Esther Duflo, Pascaline Dupas and Victor Pouliquen.
2015. ‘Turning a Shove into a Nudge? A “Labeled Cash Transfer” for Education’. Ameri-
can Economic Journal: Economic Policy 7 (3): 86–125.
Blattman, Chris, Julian C. Jamison and Margaret Sheridan. 2017. ‘Reducing Crime and Vio-
lence: Experimental Evidence from Cognitive Behavioral Therapy in Liberia’. American
Economic Review 107 (4): 1165–206.
Cahyadi, Nur, Rema Hanna, Benjamin A. Olken, Rizal Adi Prima, Elan Satriawan and Ekki
Syamsulhakim. 2018. ‘Cumulative Impacts of Conditional Cash Transfer Programs:
Experimental Evidence from Indonesia’. National Bureau of Economic Research (NBER)
Working Paper No. 24670, November.
Cunha, Jesse M., Giacomo De Giorgi and Seema Jayachandran. 2019. ‘The Price Effects of
Cash versus In-kind Transfers’. Review of Economic Studies 86 (1): 240–81.
Dizon-Ross, Rebecca, Pascaline Dupas and Jonathan Robinson. 2017. ‘Governance and the
Effectiveness of Public Health Subsidies: Evidence from Ghana, Kenya and Uganda’. Jour-
nal of Public Economics 156: 150–69.
Dupas, Pascaline, Vivian Hoffmann, Michael Kremer and Alix Peterson Zwane. 2016. ‘Tar-
geting Health Subsidies through a Nonprice Mechanism: A Randomized Controlled
Trial in Kenya’. Science (353) 6302: 889–95.
Evans, David K. and Anna Popova. 2017. ‘Cash Transfers and Temptation Goods’. Economic
Development and Cultural Change 65 (2): 189–221.
Gertler, Paul. 2004. ‘Do Conditional Cash Transfers Improve Child Health? Evidence from
PROGRESA’s Control Randomized Experiment’. American Economic Review 94 (2): 336–41.
Hanna, Rema and Benjamin A. Olken. 2018. ‘Universal Basic Incomes versus Targeted Trans-
fers: Anti-Poverty Programs in Developing Countries’. Journal of Economic Perspectives 32
(4): 201–26.
Haushofer, Johannes and Jeremy Shapiro. 2016. ‘The Short-term Impact of Unconditional
Cash Transfers to the Poor: Experimental Evidence from Kenya’. Quarterly Journal of
Designing Anti-poverty Programs in Emerging Economies in the 21st Century 339

Economics 131 (4): 1973–2042. Hidrobo, Melissa, John Hoddinott, Amber Peterman, Amy
Margolies and Vanessa Moreira. 2014. ‘Cash, Food, or Vouchers? Evidence from a Rand-
omized Experiment in Northern Ecuador’. Journal of Development Economics 107: 144–56.
Honorati, Maddalena, Ugo Gentilini and Ruslan G. Yemtsov. 2015. The State of Social Safety
Nets (English). Washington, DC : World Bank. http://documents.worldbank.org/curated/
en/415491467994645020/The-state-of-social-safety-nets-2015.
Jensen, Anders. 2019. ‘Employment Structure and the Rise of the Modern Tax System’.
National Bureau of Economic Research (NBER) Working Paper no. 25502, January.
Krueger, Alan B. and Bruce D. Meyer. 2002. ‘Labor Supply Effects of Social Insurance’. Hand-
book of Public Economics 4: 2327–92.
Levy, Santiago. 2006. Progress against Poverty: Sustaining Mexico’s Progresa-Oportunidades
Program. Washington, DC: Brookings Institution Press.
Moffitt, Robert A. 2002. ‘Welfare Programs and Labor Supply’. Handbook of Public Econom-
ics 4: 2393–430.
Olken, Benjamin A. 2006. ‘Corruption and the Costs of Redistribution: Micro Evidence from
Indonesia’. Journal of Public Economics 90 (4–5): 853–70.
Olken, Benjamin A., Junko Onishi and Susan Wong. 2014. ‘Should Aid Reward Performance?
Evidence from a Field Experiment on Health and Education in Indonesia’. American
Economic Journal: Applied Economics 6 (4): 1–34.
Olken, Benjamin and Audrey Sacks. 2018. Indonesia - Long-term Impact Evaluation of Generasi
(English). Washington, DC: World Bank.
Shultz, T. Paul. 2004. ‘School Subsidies for the Poor: Evaluating the Mexican Progresa Pov-
erty Program’. Journal of Development Economics 74 (1): 199–250.
TNP2K (National Team for the Acceleration of Poverty Reduction). 2019. ‘Alokasi Anggaran
Program Perlindungan Social’ [Budget allocation for the social protection program].
PowerPoint presentation.
Triyana, Margaret. 2016. ‘Do Health Care Providers Respond to Demand-side Incentives?
Evidence from Indonesia’. American Economic Journal: Economic Policy 8 (4): 255–88.
World Bank. 2017. Closing the Gap: The State of Social Safety Nets 2017. Washington, DC: World
Bank. https://openknowledge.worldbank.org/handle/10986/26655.
758755
research-article2018
CPSXXX10.1177/0010414018758755Comparative Political StudiesKyle

Article
Comparative Political Studies
2018, Vol. 51(11) 1472­–1503
Local Corruption and © The Author(s) 2018
Article reuse guidelines:
Popular Support for sagepub.com/journals-permissions
DOI: 10.1177/0010414018758755
https://doi.org/10.1177/0010414018758755
Fuel Subsidy Reform in journals.sagepub.com/home/cps

Indonesia

Jordan Kyle1

Abstract
This article examines the role played by local governments in shaping
resistance to reforming fiscally and environmentally disastrous fuel subsidies.
Shifting from universal-access social programs, like fuel subsidies, to targeted
programs requires vesting authority with local politicians and bureaucrats,
whom the state relies on to identify poor households and to deliver benefits.
Where local governments are corrupt, citizens find promises to replace fuel
subsidies with targeted spending less credible and resistance to reform is
higher. Using household survey data from Indonesia, this article finds that
corruption in the implementation of targeted transfer programs increases
resistance to fuel subsidy reform among the poor citizens who consume
the least fuel and who stand to benefit the most from targeted programs.
Findings suggest that improving capacity within subnational governments to
deliver social programs is important in developing public support for reform.

Keywords
fuel subsidies, Indonesia, corruption and patronage, subnational politics,
social welfare programs

1International Food Policy Research Institute, Washington, DC, USA

Corresponding Author:
Jordan Kyle, International Food Policy Research Institute, 2033 K Street NW, Washington,
DC 20007, USA.
Email: jordanckyle@gmail.com
Kyle 1473

Countries as diverse as China, Egypt, Indonesia, India, Nigeria, Russia,

Thailand, and Venezuela, among others, devote large shares of government
resources toward lowering domestic prices on basic household commodities
like food and fuel. For poor households that spend large fractions of their
income on basic commodities, consumer subsidy programs often provide the
main form of social protection, by reducing the price of basic commodities
and by reducing households’ exposure to commodity price volatility
(Alderman, 2002). Keeping prices low can raise standards of living in the
short run, but consumer subsidy programs, particularly those that subsidize
the consumption of fossil fuels, are economically disastrous over time. In
2011, government expenditures on fossil fuel subsidies summed to nearly
$500 billion globally, and most of the subsidy benefits were captured by
richer, urban households (Clements et al., 2013). By comparison, replacing
subsidies on fuel consumption with targeted support for the poor would pro-
vide greater benefits to the poor without straining budgets (Arze del Granado,
Coady, & Gillingham, 2012; Clements et al., 2013).
Some countries have been more successful than others in making this tran-
sition. Attempts to raise fuel prices are linked to triggering the widespread
antigovernmental protests that contributed to the downfall of the Soeharto
regime in Indonesia, to setting off the Saffron Revolution in Myanmar, and to
precipitating the 2008 military coup in Mauritania. By contrast, Indonesia
raised fuel prices in 2013 and 2014 without major upheaval, and Iran, Jordan,
and Morocco all slashed fuel subsidies between 2008 and 2012.
This article explains variations in popular support for reforms based on the
varying credibility of the institutions entrusted with implementing reforms.
All public require some delegation, and citizens may have different beliefs
about which sets of institutions can more credibly and effectively implement
policies. Shifting from universal-access social protection programs—like
consumer fuel subsidies—to targeted social protection programs requires
vesting authority for policy implementation with local politicians and bureau-
crats, whom the state relies on to identify poor households and to physically
transfer benefits to them. This confers local elite with a great deal of discre-
tion, which they can use to divert public resources for private gain. When
local elites have used this discretion to faithfully implement targeted transfer
programs in the past, I argue, this can build confidence in reforms that
empower them; by contrast, when local elites have used discretionary power
for private gain, confidence is undermined.
I focus on the key role played by local elite in shaping popular support for
reform of consumer subsidies on automotive fuel (hereafter, “fuel subsidies”)
in Indonesia.1 While it is commonly hypothesized that scaling back fuel sub-
sidies with no form of compensation would be widely rejected, replacing
1474 Comparative Political Studies 51(11)

them with targeted programs could, in theory, offer benefits for the poor with
less fiscal strain and mitigate social unrest. The central argument of this arti-
cle is that where local elite have engaged in corruption in past implementa-
tion of targeted transfer programs, support for fuel subsidy reform is
undermined. While prevalent use of food and fuel subsidies are often attrib-
uted to the political influence of urban interests (e.g., Bates, 1981), this argu-
ment suggests that, where corruption is prevalent, even rural citizens who, in
theory, could benefit from targeted transfers, may oppose subsidy reforms.
I test this argument using household survey and administrative data from
194 randomly selected villages in Indonesia. Indonesia makes an interest-
ing case to examine the effects of local corruption as there is significant
variation: In around 40% of villages, I estimate quite low levels of corrup-
tion in the implementation of targeted transfer programs, while in others
corruption eliminates the benefits of the targeted transfer program within
the village altogether. By contrast, countries with universally weak local
capabilities may not even contemplate reforms which rely on local officials
for implementation.
Using this data, I estimate an important indicator of local corruption: the
gap between the official quota that a given village should receive from
Indonesia’s largest targeted transfer program and what households receive in
practice. To add validity to the measure and credibility to the argument that
citizens reason that corruption in past implementation of targeted transfer
programs will affect their ability to benefit from future targeted programs, I
show that the corruption measure used here strongly predicts whether poor
citizens receive targeted transfers during a later fuel subsidy reform.
Corruption is associated with, on average, 38% less support for fuel subsidy
reform among poor households. The finding is confirmed using matching
methods and in a number of robustness checks. I also test whether the argu-
ment carries to other contexts by examining public opinion on fuel subsidy
reform in Nigeria and find remarkably similar patterns.
The article proceeds as follows. The following section discusses consumer
subsidy programs generally and why they can be so difficult to reform. The
next sections present the main argument and the Indonesian context. This is
followed by a discussion of the data, the empirical strategy, and the main
results of the article. The final section concludes, discussing the implications
of the findings for the prospects for fuel subsidy reform.

Why Are Fuel Subsidies So Difficult to Repeal?

Consumer subsidies, including those on fuel, often begin as relatively modest
attempts to shield households from fluctuations in the price of basic goods
Kyle 1475

and services. In other words, they often begin as price stabilization mecha-
nisms rather than subsidy programs (Bril-Mascarenhas & Post, 2015). As
commodity prices rise, however, price stabilization mechanisms can trans-
form into large-scale consumer subsidies, particularly if governments ini-
tially design fixed price regimes (in which consumer price stays fixed and the
size of the subsidy fluctuates) rather than fixed subsidy regimes (in which
consumer price fluctuates to maintain a fixed subsidy size). Indeed, it is the
combination of fixed price regimes with rising oil prices which explains
much of the increase in the size of fuel subsidies over time: As market prices
rise, countries with fixed price regimes become subsidizers (Ross, Hazlett, &
Mahdavi, 2017).
Fossil fuel subsidies are typically adopted in countries where their imple-
mentation is relatively cheap compared with other means of influencing
household consumption. For example, major oil producers are frequent sub-
sidizers of domestic consumption (Cheon, Urpelainen, & Lackner, 2013). In
these countries, the public often expects to benefit from their country’s oil
wealth, and

the state has a strong incentive to provide goods and services that make it
appear that everyone is benefitting—even if some are actually benefitting
more—either because they are made available to all citizens or because they
have a high degree of visibility. (Jones Luong & Weinthal, 2010, p. 60)

Fuel subsidies are thus an attractive political tool because their benefits are
highly visible yet their costs have low visibility, as, in oil exporters, the cost
of providing them is born primarily as an off-budget opportunity cost (of not
selling fuel at market prices).
Subsidies on fuel are particularly prevalent in countries whose oil wealth
is controlled by state-owned oil companies (Cheon, Lackner, & Urpelainen,
2015). States whose oil wealth is managed by state-owned entities may face
higher pressures to distribute oil wealth in highly visible ways (Jones Luong
& Weinthal, 2010), and the higher capacities of state-owned oil companies
compared with their state patrons make it more feasible to do so via subsidiz-
ing fuel compared to other forms of distributing wealth (Cheon et al., 2015;
Victor, 2009).While these can provide powerful political motivations for
adopting fuel subsidies, maintaining them can become prohibitively costly
over time, even in oil exporters. For example, fuel subsidies in Nigeria, a net
oil exporter, consumed 20% of government expenditures in 2012, due to high
commodity prices and low domestic refining capacity, which forces Nigeria
to import 85% of domestically consumed fuel products (Siddig, Aguiar,
Grethe, Minor, & Walmsley, 2014).
1476 Comparative Political Studies 51(11)

Once enacted, Bril-Mascarenhas and Post (2015) argue that consumer sub-
sidy programs become policy traps—”initially modest policies that grow rap-
idly and become more entrenched quickly” (p. 99). Even if citizens have poor
knowledge of the overall size of subsidies, they argue, attempts to raise prices
are highly visible. As commodity prices rise, both the urgency and the difficulty
of subsidy reform reaches a crescendo: the cost of providing subsidies increases
dramatically for governments, yet the costs that citizens would face in adjust-
ing to higher market prices are also more substantial. Politicians thus seek to
avoid blame for repealing subsidies, fearing political backlash. These fears are
not unfounded: over the last decade, attempts to raise gasoline prices have trig-
gered protests in at least 19 countries (Ross et al., 2017), and, more broadly,
large consumer price shocks can trigger urban riots (Bellemare, 2015; Smith,
2014).
When can consumer fuel subsidies be scaled back? Many advocate ini-
tiating reform efforts while the real price of oil is low, as citizens face
lower costs of adjustment to subsidy reduction when prices are low (e.g.,
Benes, Cheon, Urpelainen, & Yang, 2016; Ladislaw & Cuyler, 2015).
While some governments have reduced the overall size of subsidies in the
current low oil price environment, overturning fixed price regimes has
proven particularly challenging (Ross et al., 2017). In many cases, low
prices simply temporarily eliminate subsidies in fixed price regimes,
which return when commodity prices rise. Thus, it is difficult to assess
whether reforms undertaken in low price environments have been success-
ful until market prices rise (and consumer prices are allowed to rise as
well).
Beyond just undertaking reforms when prices are low, many suggest that
social unrest could be avoided if social programs that target the poor were
implemented alongside subsidy reform (e.g., Clements et al., 2013; The
World Bank, 2009).2 Given the prominent role that citizen backlash can play
in derailing reform, surprisingly little is known about citizens’ attitudes
toward fuel subsidies, their receptiveness to reform, nor how popular sup-
port for policy changes can be built over time.
This article’s goal is to explain why some citizens become more amenable
to subsidy reform than others, focusing on the attitudes of citizens in Indonesia
toward consumer subsidies on automotive fuel. Although the empirical focus
is narrow, the argument is relevant to questions about scaling back other
types of consumer subsidy programs, moving from universal access to tar-
geted forms of social protection, and compensating the poor for costs of
adjustment to economic reforms more broadly. In the following section, I
provide an argument for how local institutional context shapes citizens’ atti-
tudes to reform.
Kyle 1477

An Institutional Explanation for Attitudes

The central government must delegate a range of governing activities, from
tax collection to the distribution of social benefits, to local bureaucrats on a
daily basis. This delegation can be legally mandated when authority for ser-
vice delivery is devolved to local governments. It can also occur as a de facto
delegation when central governments need the assistance of local officials to
find beneficiaries and deliver benefits to them; relying on local actors for this
information gives local actors implicit control over program targeting and
implementation (Jaspars & Shoham, 1999).
The degree of this delegation varies by policy. Universal-access consumer
subsidy programs, like fuel subsidies, require little delegation from central to
local governments, as there is no need to discriminate between beneficiaries
and nonbeneficiaries. By contrast, delegation is much higher for targeted
social protection programs, with local institutions frequently responsible for
determining eligibility for targeted programs and for distributing benefits.
When governing authority is delegated—either on a de jure or de facto
basis—it opens up the potential for local actors to exploit their position,
diverting public resources for private gain. When local elites are “captured,”
local authority over service delivery can result in overproviding benefits to
local elites at the expense of the poor (Bardhan & Mookherjee, 2005). Local
corruption in the implementation of targeted transfer programs can be so
costly that it eliminates the benefits of redistributive policies altogether
(Olken, 2006). One prominent feature, therefore, of shifting from universal-
access subsidy programs to targeted social protection programs is increasing
the resources and authority channeled through local officials.
There is considerable evidence that support for public policies is shaped
by trust in the government (Hetherington, 2005; Rudolph & Evans, 2005),
which is significantly diminished when governments are corrupt (Anderson
& Tverdova, 2003; Hakhverdian & Mayne, 2012; Seligson, 2002). In particu-
lar, support for policies that involve an intertemporal bargain—or those that
impose a cost today for the promise of a future benefit—and the delegation of
new resources and responsibilities to public officials depends on trust in the
institutions that implement the policies (Jacobs & Matthews, 2017). The
uncertainty created by this delegation, Jacobs and Matthews (2017) argue, is
such a significant factor in citizens’ attitudes that policy opponents “will fre-
quently seek to undermine support not by denigrating the good in question
(e.g., schools, clean air, roads), but by calling into question the credibility of
government promises to deliver the promised benefits” (p. 196).
Replacing consumer fuel subsidies with targeted support for the poor
involves both an intertemporal bargain and delegation. Citizens have no
1478 Comparative Political Studies 51(11)

guarantee that politicians will follow through on promises to increase invest-

ments in targeted transfer programs once they have been authorized to reduce
support for fuel subsidies. Uncertainty about future policies and political
interactions is particularly significant in developing democracies (Lupu &
Riedl, 2013). Second, even if politicians do follow through on a promise to
increase investments in targeted transfer programs following a fuel price
increase, this act requires investing the new resources and authority required
to implement the transfer program to local officials. This delegation opens up
the possibility that the local officials responsible for delivering the policy will
fail to deliver its promised benefits.
However, citizens, even within the same country, can have vastly different
beliefs about whether or not local officials will deliver on policy promises. In
making these assessments, I argue that citizens rely on past experiences with
local policy implementation to form policy attitudes. This logic features
prominently in the “policy feedback” literature, which argues that, in addition
to social and economic considerations, citizens’ experiences with policy
implementation, whether positive or negative, shape their policy preferences
(Campbell, 2012). For example, negative experiences with caseworkers
responsible for verifying eligibility for social assistance programs or repre-
sentatives of the criminal justice system reduces citizens’ trust in the govern-
ment in general in the United States (Soss, 1999; Weaver & Lerman, 2010).
Moreover, given their prominent role in the day-to-day lives of citizens and
in implementing state policy at the local level, local politicians and bureau-
crats are often seen as the key link between state and society (Auerbach,
2016; Weitz-Shapiro, 2008).
I argue that corruption among local public officials diminishes local pub-
lic support for reforms that delegate more resources and authority to the
local level. When local officials have failed to faithfully implement targeted
transfer programs in the past, citizens have little reason to believe that they
will faithfully implement future targeted transfer programs of the type prom-
ised by national politicians as compensation for raising fuel prices. On the
other hand, support for reform could be increased over time by building the
local capacity to implement targeted transfer program and by reducing local
rent-seeking.
However, local corruption may not affect the attitudes of all citizens toward
shifting from universal to targeted programs equally. For those citizens who do
not expect to be targeted under future programs, corruption among village offi-
cials in the implementation of targeted transfer programs may shape their atti-
tudes for other policy areas, but should not necessarily affect whether or not
they favor subsidy reform. For these citizens, moving from universal-access
consumer subsidy programs to targeted transfer programs means facing higher
Kyle 1479

prices without compensation for loss. Consequently, the views of these citizens
are more likely shaped by socioeconomic characteristics and vested interests in
subsidies than by considerations about empowering local officials.
Instead, corruption should primarily affect the attitudes of those citizens
who could be beneficiaries of future targeted transfer programs, for whom the
calculus about how well local officials will implement future programs is
directly relevant to the tradeoff between universal and targeted programs.3
Thus, I test the argument separately on citizens who are eligible and ineligi-
ble for targeted transfer programs and expect that corruption primarily drives
attitudes among the eligible. Understanding why poorer households (who
benefit less from the subsidy) may resist reform is particularly important, as
these households form the natural constituency which could be harnessed in
favor of reform. I turn to the case of Indonesia to test these ideas.

Subsidies and Targeted Transfer Programs in

Indonesia
Administrative—rather than market-based—pricing for fuel has existed in
Indonesia since at least the 1960s. At the time, government intervention in the
pricing of basic consumer goods was common in Indonesia and elsewhere in
Southeast Asia, particularly to stabilize (rather than subsidize) prices
(Timmer, 1993). However, the Asian Financial Crisis and the years of recov-
ery that followed dramatically changed the policy environment around fuel
subsidies in at least two key ways.
First, the sharp devaluation in Indonesian rupiah caused the costs of fuel
subsidies to escalate tenfold, from 0.3% of GDP in 1996 to 2.9% in 1998
(Clements, Jung, & Gupta, 2007, p. 222). Second, the devastating effects of
the Asian Financial Crisis on poverty rates in Indonesia (Poppele, Sumarto,
& Pritchett, 2000) highlighted to policymakers the need for a social safety net
that could target poor households (Haggard & Birdsall, 2000). The govern-
ment of Indonesia has made several attempts since the Asian Financial Crisis
to both reduce the fiscal burden of fuel subsidies and to provide more targeted
social safety net programs for the poor—efforts that have been deeply linked
to each other. Although reforms have proceeded in fits and starts and have not
all been successful, the extent to which policymakers across different politi-
cal parties and regimes have coalesced around these twin goals is remarkable.
Indeed, every regime that has held power since the fall of Soeharto has initi-
ated attempts to reform fuel subsidies.
One of the first significant moves to lessen the fiscal strain of the subsidy
program occurred in the immediate aftermath of the Asian Financial Crisis.
To survive the crisis, the government sought a loan by the International
1480 Comparative Political Studies 51(11)

Monetary Fund (IMF), and elimination of fuel subsidies was a key line item
in the set of commitments made by the government in accepting the loan.
Consequently, the government implemented a 71% increase in the price of
fuel in May 1998, setting off mass demonstrations and social violence across
the country and ultimately leading to Soeharto’s resignation later that month.4
The idea that fuel price hikes could trigger social unrest and topple govern-
ments loomed large across subsequent reform efforts.
Meanwhile, concerns about corruption among local officials hampered
efforts to introduce targeted social safety net programs to offset fuel price
hikes. Local politicians, who were needed both as sources of information
about which households to target and as local implementors of programs,
could leverage this power in order to establish electoral advantages by using
transfers to sway voters or to fund campaigns. The World Bank temporarily
suspended disbursements for social safety net loans to avoid charges that they
were indirectly supporting incumbents in the upcoming local elections
(Haggard & Birdsall, 2000). More generally, Indonesia’s “big bang” decen-
tralization of 1999 raised concerns that empowering local officials would
enable widespread rent-seeking by local officials (Hadiz, 2010).
A prominent fuel subsidy reform efforts in 2005 illustrates both how the
government tried to lessen fuel subsidies’ fiscal burden and how local corrup-
tion complicated these efforts. The reform raised fuel prices yet coupled these
increases with targeted cash transfers for the poor to help the poor cope with
the economic costs of adjustment to higher fuel prices. Where there was sig-
nificant local leakage in the cash transfer program, there was also diminished
social capital and increased crime (Cameron & Shah, 2014), reflecting both
that local politicians were able to divert resources from targeted transfer pro-
grams for personal gain and that doing so changed community relations and
attitudes. Yet, the reform established the idea that governments could raise
fuel prices without losing office, as long as compensatory programs were
established as well. This reform type—linking price hikes with targeted
transfer programs—became a model for later fuel subsidy reforms in
Indonesia and was followed in fuel price hikes in 2008 and 2013 and a failed
reform attempt in 2012 (Beaton, Lontoh, & Wai-Poi, 2017).
Despite efforts to replace consumer subsidies with targeted programs for
the poor, fuel subsidies remained entrenched. In particular, the administrative
pricing mechanism—which has no automatic link to market prices—had not
been amended.5 Thus, rising commodity prices could quickly outpace one-off
price adjustments. The fiscal burden of maintaining the subsidies ballooned
over time, as Indonesia faced declining oil production (becoming a net-con-
sumer in 2004), rising international oil prices, and increasing domestic con-
sumption. By 2013, when the survey used in this article was fielded, subsidies
Kyle 1481

on fuel and electricity accounted for around 25% of government expendi-

tures, an amount which exceeded total spending on education, health care,
and social protection combined (International Institute for Sustainable
Development [IISD], 2012, p. 5). Globally, Indonesia ranked behind only
Saudi Arabia and Iran in subsidy amount (Davis, 2014).
Since 2014, under the leadership of President Joko Widodo (“Jokowi”),
however, reform efforts have ramped up. During his presidential campaign,
Jokowi emphasized the significant financial strain imposed by fuel subsidies
and the potential to increase spending on targeted, pro-poor programs, as well
as on infrastructure and other public investments, if spending on fuel subsi-
dies were reduced (Setiawan, 2014; Suryowati, 2014). Upon assuming office,
Jokowi acted on his campaign promises. He issued identification cards to
nearly 18 million of Indonesia’s poorest households, intended to give them
access to health, education, and welfare programs and eliminated the subsidy
on gasoline as a line item in the national budget (Bayu, 2016). Jokowi’s status
as the first political “outsider” to hold the presidency—not affiliated with
prior regimes or with Indonesia’s oligarchy6—may have enhanced his credi-
bility as a reformer and his ability to push through reforms with limited popu-
lar backlash. Reform was also eased by low oil prices: reconciling market and
retail prices at the time required lowering rather than raising the retail price
of gasoline.
Despite this progress, efforts to reform the pricing mechanism have not
taken hold, meaning that the government still sets retail prices for gasoline on
an ad hoc basis. Indeed, as market prices ticked upward again more recently,
the government failed to adjust retail prices, and Indonesia returned to a net-
subsidizing position (IISD, 2015). This early backsliding on subsidy reform
suggests that as long as the government maintains control of pricing, the
temptation to delay price increases to avoid political backlash is strong.
Indeed, there is no reason to expect that citizens would not punish the govern-
ment for price increases as the government maintains agency over prices.
Thus, the importance of understanding how the state can build public support
for fuel subsidy reform remains highly significant, even as lower oil prices
have eased the fiscal burden.

Data and Empirical Strategy

Data Collection
In the following section, I present results from a survey—fielded in March-
May 2013—of 1,940 randomly selected households across 194 randomly
selected villages in six districts in Indonesia.7 The districts are spread across
1482 Comparative Political Studies 51(11)

Indonesia, including on- and off-Java. Full information on sampling can be

found in the Online Appendix. Compared to national averages, the sample is
slightly skewed toward rural areas, and contains fewer households in the
lower consumption deciles. To correct for sampling imbalances, I use entropy
balancing to reweight the survey data to match demographic information
from the population (Hainmueller, 2012; Hainmueller & Xu, 2013).8
Summary statistics for raw and weighted data can be found in the Online
Appendix (OA1).

Research Design
Policy attitudes.  To measure policy attitudes toward public spending on fuel
subsidies, I use two survey questions. In each case, the question prompts the
respondent to think about how they would allocate government resources if
given the opportunity. The first questions asks the respondent to select three
among a list of eight government programs on which she would increase
government spending, while the second question asks the respondent to select
three among the same eight government programs on which she would
decrease government spending.9 The programs were selected to represent the
primary ways that the government in Indonesia intervenes in the economy to
promote the welfare of poor citizens, and included cash transfers for the poor,
subsidized rice for the poor, health fee waivers for the poor, hiring teachers,
improving roads in rural areas, community-driven development funds, reduc-
ing the price of automotive fuel, and reducing the price of LPG (cooking
fuel). The options enable respondents to choose from a variety of types of
public spending: targeted transfer programs, local public goods, and con-
sumer subsidies.
The survey questions are intended to elicit a ranking of respondents’ pref-
erences toward various forms of public spending given a budget constraint. It
is worth noting that preferences over spending levels are not necessarily the
same as policy preferences. An individual may, for example, have strong
preferences in support of education but think that the existing education sys-
tem is ineffective and unworthy of additional resources. Rather than repre-
senting pure policy preferences, then, preferences for more or less public
spending on a particular policy domain represent the difference between an
individual’s preferred level of spending and status quo spending levels, given
beliefs about institutions and policy implementation (Wlezien, 1995).
This type of broad spending question offers several advantages over asking
directly about support for fuel subsidy reform in this context. First, a specific
fuel subsidy reform package had been proposed, met with widespread pro-
tests, and ultimately overturned by parliament during the 6 months prior to
Kyle 1483

survey implementation. In this environment, asking directly about support for

fuel subsidy reform risked being confounded with support for the president
(who proposed the policy), support for opposition parties (who overturned it),
and/or support for the specific reform package being proposed rather than
attitudes toward subsidy reform more broadly.10 Second, the more indirect
measure of support reduces concerns about response bias. Given that the
Indonesian government has been trying to reform fuel subsidies for many
years and communicating the costs of the policy to citizens, citizens may be
reluctant to express direct support for the subsidy in a survey. For these rea-
sons, I favor using a more general question that puts public spending on fuel
subsidies in context with public spending on other policy areas. I validate the
measure by testing the argument on a direct question about fuel subsidy reform
implemented by Afrobarometer in Nigeria, where I find both roughly similar
overall levels of support for reform as well as support for the argument.
I measure resistance to fuel subsidy reform using a three-category out-
come variable, where “1” indicates that the respondent selected to increase
spending, “0” that the respondent selected neither to increase nor to decrease
spending, and “−1” that the respondent selected to decrease spending on gas-
oline subsidies. I also show results using binary indicators for whether the
respondent selected to increase or decrease spending on gasoline subsidies. I
focus on support for gasoline subsidies rather than LPG subsidies as these
have been the focal point for reform (as well as the primary source of the fis-
cal and environmental burden) in Indonesia.
Figure 1 illustrates the level of support for fuel subsidy reform by con-
sumption decile. Overall, a minority of respondents support increasing gov-
ernment spending on gasoline subsidies (28%), and 39% of respondents
support decreasing government spending on gasoline subsidies. In fact,
except for the richest 20% of households, a greater number select to decrease
rather than increase spending on gasoline subsidies within all consumption
deciles. The poorest households exhibit the highest levels of support for
reform: Over 40% of households in the poorest 20% support decreasing
resources devoted to gasoline subsidies. By comparison, there is more overall
support for targeted transfer programs, with 53% of all households wanting
to increase resources devoted to cash transfers. Taken together, this suggests
that gasoline subsidies are only moderately popular as compared to targeted
transfer programs, and less so among the poor than the nonpoor. A substantial
portion of citizens support scaling back government spending on gasoline
subsidies.11

Estimating corruption.  A large literature explores the relationship between cor-

ruption and political attitudes by uncovering partial correlations between per-
ceptions of corruption and attitudes (e.g., Anderson & Tverdova, 2003;
1484 Comparative Political Studies 51(11)

Figure 1.  Support for fuel subsidy reform by consumption decile.

Clausen, Kraay, & Nyiri, 2011; Seligson, 2002). It is possible that the same
underlying characteristics that cause individuals to report high levels of cor-
ruption also cause them to have low confidence in public policies, making it
difficult to isolate the effect of corruption. Alternatively, respondents may be
reticent to report corruption in survey responses when doing so implies some
degree of personal wrongdoing, or simply because such questions can be
sensitive, and the resulting bias in corruption estimates can be large (Kraay &
Murrell, 2016).
Rather than relying on perceptions of corruption, I estimate corruption
among local officials directly using a “gap measurement” method, which
estimates corruption by identifying discrepancies between different data
sources.12 Specifically, I look at corruption in the implementation of
Indonesia’s largest targeted transfer program, Raskin. The program, in the-
ory, provides 15 kg of rice to 17.5 million low-income households at a copay
price that is about one fifth of the market price. Although the program is sup-
ported by the central government, the day-to-day logistics for the delivery of
subsidized rice to beneficiaries is handled by local governments. Local gov-
ernments are responsible for picking up the allotment of rice for their entire
village—on average 5,550 kg of rice each month to be distributed to 375
households—from a central warehouse (Banerjee, Hanna, Kyle, Olken, &
Sumarto, in press-a). At the time of the survey, the average village Raskin
quota was thus worth approximately Rp. 27 million—$2,80013—monthly, a
Kyle 1485

substantial sum in a context where average per capita income is only around
$3,300 annually.
Local officials have substantial leeway in how and whether this sum
reaches households, as it is difficult for the central government to monitor
whether households receive subsidized rice and the price that they pay.
Household purchases reveal that a substantial portion of this quota never
reaches households at all (Olken, 2006; The World Bank, 2012a), with house-
holds in the survey sample receiving only one third of the intended subsidy
(Banerjee, Hanna, Kyle, Olken, & Sumarto, in press-b).
To estimate how much rice goes missing, I use administrative data on the
village’s monthly rice quota14 and compare this to self-reported purchase data
from household surveys. For the household purchase data, I utilize data from
two different survey waves conducted in the same villages, covering a total
of 19 households per village. In each survey wave, households were asked
about the prior 3 months of Raskin purchases, including whether they pur-
chased Raskin, the amount they purchased, and the copay price.15 To arrive at
the village-level estimate, I weight the households that are eligible to receive
targeted transfer programs and those that are ineligible to do so based on their
proportions in the village population. I then compare this figure to the official
Raskin quota for the village: the difference between the value of the official
quota and the value of the total household rice purchases estimated for the
village is the missing Raskin subsidy for the village.
Using this method, I estimate that, on average, villages are missing 23%
of their intended subsidy. However, there is substantial heterogeneity: In 41%
of villages, less than 10% of the intended subsidy goes missing, while in
other villages the entire subsidy is missing (Figure 2). It is important to note
that this measure does not include mistargeting—when benefits are redi-
rected from eligible to ineligible households—but only rice that is never
reflected in household purchases.
While this measure does not directly capture all of the ways in which local
elites can divert public resources for private gain (e.g., diverting transfers
toward political supporters), Raskin is the largest transfer that is regularly
channeled through local officials and thus a significant opportunity for local
corruption. To add validity to the measure and credibility to the argument that
citizens reason that corruption in past implementation of targeted transfer
programs will affect their ability to benefit from future targeted programs, I
test whether the corruption measure used here predicts whether poor citizens
receive targeted transfers in Indonesia’s 2013 fuel subsidy reform. In mid-
2013, after the survey was fielded, the government implemented fuel price
increase alongside a compensatory targeted cash transfer. In a separate sur-
vey conducted in the same 194 villages in December 2013 to January 2014
1486 Comparative Political Studies 51(11)

Figure 2.  Distribution of missing subsidy.

(Banerjee et al., in press-a), we asked respondents whether they received the

cash transfer and whether they gave part of the money that they received to
village officials.
I find remarkably similar rates of “missing benefits” within the cash trans-
fer program as in Raskin, with an average of 23% of households reporting
having given part of the money to village officials. In one quarter of surveyed
villages, at least one surveyed eligible household did not receive the transfer
at all. In Table 1, I test whether past corruption in Raskin implementation
predicts the share of eligible households within the village that do not receive
the cash transfer (column 1) and whether households in the village report giv-
ing part of their cash transfer to village officials (column 2).16 In each case,
corruption in the Raskin program is associated with whether households
receive benefits from the targeted cash transfer program in the next year.
Association between past and future corruption across different programs
adds validity to the idea that corruption within the Raskin program is more
broadly representative of corruption among local officials. It also increases
confidence in the argument presented here that households use information
on corruption within current programs as an indication of how fuel subsidy
reforms will be implemented in their locality.

Controls.  The analyses include a series of control variables likely to affect atti-
tudes toward fuel subsidies. Vehicle ownership, transportation spending, agri-
culture, fuel-intensive job, and urban address individual incentives to maintain
Kyle 1487

Table 1.  Does Past Corruption Predict Future Corruption?

Eligible households do Gave cash to local

not receive cash officials

Variables (1) (2)

Share of Raskin subsidy missing 0.052* 1.475†
(0.021) (0.848)
District fixed effects Yes Yes
N 194 194
†p < .10. *p < .05. **p < .01. Standard errors in parentheses.

subsidies. Vehicle ownership is measured based on whether the household

owns a car, truck, motorbike, or motorboat. Transportation spending is mea-
sured as the share of the household’s expenditures that were spent on transpor-
tation in the past month. Agriculture indicates whether the household owns an
agricultural field, as fuel is an important input into agricultural production.
Individuals employed in more fuel-intensive sectors of the economy may have
higher incentives to maintain subsidies. I code fuel-intensive occupations as
those working in agriculture, mining, manufacturing, or transportation. Urban
indicates whether households live in an urban or rural area, as urban house-
holds are typically thought to be more supportive of fuel subsidies. On the
other hand, the per-unit subsidy for rural households is higher because of
higher transportation and storage costs in remote areas.
Education indicates the level of education attained by the respondent and
ranges from 0 (no schooling) to 10 (postgraduate degree). An individual’s
education level may shape the extent to which they understand the costs of
fuel subsidies and the potential social benefits of reform. Female indicates
whether the respondent is female, as women may have different priorities for
public goods and social spending. Finally, richer households are widely seen
to benefit more from fuel subsidies, so I include a variable measuring the
households’ (logged) per capita monthly consumption. Summary statistics
are reported in the Online Appendix (OA2).

Evaluation Strategy
To test the effects of local corruption on support for reform, I first use a
simple multivariate analysis, controlling for household and village
characteristics:

Attitudesij = α + β corruption j + δX j + γZ i + ωk + εij ,

1488 Comparative Political Studies 51(11)

where ij denotes individual i in village j, X are village controls, Z are

individual controls, ωk are district fixed effects, and ε is the error term. The
main goal is to estimate β in the equation above, which captures the impact
of corruption on attitudes. I estimate an ordered-logit model for the three-
category dependent variable and logit models for the binary dependent
variables.
The models will only provide valid estimates if the decisions by local
officials to engage in corruption are not endogenous to policy attitudes. One
concern is simply that corrupt villages are different from noncorrupt villages
on a host of dimensions, and it is possible that these differences, rather than
corruption itself, drive the correlations between corruption and policy atti-
tudes. Matching improves comparability between “treatment” and “control”
villages by preprocessing the data, removing villages from the sample that
are dramatically different and thus creating a matched sample from the origi-
nal data that contains covariates with similar values in both “treatment” and
“control” villages. However, matching cannot eliminate bias caused by unob-
served variables.
I use coarsened exact matching (CEM) to preprocess the data (Iacus, King, &
Porro, 2012). CEM does not require the use of a specific matching algorithm.
Instead, CEM requires “coarsening” the values of covariates into discrete cate-
gories. After coarsening, exact matching is used to sort the data into strata, and
only strata that contain at least one treatment and one control village are retained.
This restricts data to common support and ensures that balance between treat-
ment and control is improved for all covariates.17 Doing so ensures that only
inferences that are close to the data are made, which in turn means less model
dependence and reduced bias (Ho, Imai, King, & Stuart, 2007).
I include a number of covariates that likely drive decisions to skim from
targeted transfer programs to perform the CEM. Urban indicates whether the
individual lives in an urban area, and village remoteness is measured as the
distance to the subdistrict capital. Less urban and remote areas may be more
difficult for the central government to monitor, and citizens in these areas
may have less overall knowledge about program benefits, giving local offi-
cials within these areas more opportunities to divert resources for private
gain. Third, villages with higher levels of ethnic fragmentation and larger
populations may have more difficulty engaging in collective action to moni-
tor and punish the activities of corrupt officials. Fourth, villages with more
religious institutions—and thus more informal community leaders—may be
better positioned to monitor and punish corruption among local officials.
Finally, villages in Indonesia vary in whether citizens have the opportunity to
directly elect village leaders (and to punish corrupt officials at the polls). In
Kyle 1489

some villages, the village head is directly elected by local citizens, while in
others the village head is appointed by the (elected) district mayor.
While dichotomous variables (urban, direct elections) can be matched on
their exact values, continuous variables must be coarsened before exact match-
ing. I coarsen population into two categories: whether or not the village has a
population of over or under 1,200 households, which is the minimum popula-
tion required for new village formation. Religious institutions and village
remoteness are coarsened into three categories, and ethnic fragmentation is
coarsened into two categories.18 The CEM procedure identifies 25 matched
strata for 130 villages. For each village, the proportion of treated to control
observations within the strata is used to create a CEM weight, which is used in
the postmatching analysis. Because the matching is exact, there is no difference
between corrupt and noncorrupt villages on the covariates used for matching.
However, balance for continuous covariates is improved as well (OA4).

Results
Main Results
How does corruption affect attitudes toward fuel subsidies? I first discuss the
results for the models without matching. The first column in Table 2 presents
results for the three-category measure of resistance to fuel subsidy reform,
the continuous measure of corruption, and control variables. In this model,
the corruption variable is positive and significant for eligible households,
indicating that living in a village where local officials skim more from tar-
geted transfer programs indeed increases resistant to fuel subsidy reform.
Figure 3 illustrates the magnitude of the effect among eligible households.
When corruption levels are near zero, poor citizens are more than two and a
half times more likely to support rather than oppose fuel subsidy reform. As
the share of missing subsidy nears 100%, the predicted probability that poor
citizens support reform declines by 18 percentage points. Meanwhile, the
predicted probability that poor citizens oppose reform increases by 14 per-
centage points. Among control variables, per capita consumption signifi-
cantly impacts attitudes, as does living in an urban area.
Corruption matters on the extensive margin as well: A poor citizen living
in a village where officials skim from targeted transfer programs is 6 per-
centage points more likely to oppose fuel subsidy reform, a 38% increase
from the sample mean for eligible households in villages with no corruption
(calculated based on column 2). Column 3 includes the same variables as in
column 2 but is conducted on the matched sample only. Results are very
similar to the first two models: The presence of corruption in the
1490 Comparative Political Studies 51(11)

Table 2.  Local Corruption and Resistance to Fuel Subsidy Reform.

Panel A: Eligible households Panel B: Ineligible households

(1) (2) (3) (4) (5) (6)

Sample Full Full Matched Full Full Matched

Subsidy missing (%) 0.731* 0.125  
(0.360) (0.384)  
[0.107, [–0.274,  
1.132] 0.750]
Any subsidy missing 0.421* 0.560* 0.110 −0.311
  (0.192) (0.262) (0.215) (0.255)
  [0.032, [–0.022, [–0.112, [–0.578,
0.658] 0.878] 0.554] 0.357]
Vehicle ownership 0.266 0.274† 0.585* 0.727** 0.727** 0.902**
(0.167) (0.165) (0.251) (0.229) (0.229) (0.248)
Sh. trans. spending 0.878 0.948 −1.187 1.471 1.516 1.893
(1.008) (0.999) (1.678) (1.738) (1.751) (2.679)
Agriculture 0.162 0.180 0.001 −0.073 −0.079 −0.169
(0.162) (0.164) (0.229) (0.187) (0.188) (0.234)
Fuel-intensive job 0.076 0.069 0.115 −0.006 −0.006 −0.120
(0.153) (0.151) (0.213) (0.161) (0.161) (0.241)
Urban 0.714** 0.681** 0.009 0.561† 0.571† −0.079
(0.206) (0.210) (0.403) (0.321) (0.320) (0.472)
Education 0.083 0.088 0.348** −0.070 −0.007 −0.102
(0.075) (0.076) (0.103) (0.045) (0.045) (0.062)
Female 0.150 0.157 −0.052 0.007 0.010 0.201
(0.146) (0.145) (0.200) (0.145) (0.144) (0.214)
HH Cons. (log) 0.420* 0.428* 0.580* 0.197 0.195 0.265
(0.191) (0.187) (0.276) (0.158) (0.157) (0.183)
District fixed effects Yes Yes Yes Yes Yes Yes
n 1,129 1,129 755 715 715 470

Standard errors clustered by village in parentheses. 90% confidence intervals for

bootstrapping procedure reported in brackets.
†p < .10. *p < .05. **p < .01.

implementation of targeted transfer programs increases resistance to fuel

subsidy reform among eligible households. Further, results are consistent
using the binary outcome variables indicating whether respondents selected
to increase or decrease spending respectively on automotive fuel subsidies
(OA5).
Kyle 1491

Figure 3.  Impact of corruption on predicted probability of resistance to fuel

subsidy reform.
Figure 3 plots predicted probabilities from model in Table 2, column 1 with control variables
held at means.

Consistent with the argument, corruption does not affect attitudes among
households that are ineligible for targeted transfer programs, reported in col-
umns 4 to 6. Point estimates for corruption are near zero and consistently
insignificant across the models. Instead, resistance to fuel subsidy reform
among these richer households is shaped primarily by vehicle ownership and
living in an urban area.
Because corruption is estimated, this could introduce noise into the mod-
els. I use a bootstrapping procedure to address this issue: Within each village,
I redraw the data set 1,000 times (sampling with replacement) and calculate
corruption estimates for each draw of the data, reestimating the model for
each draw of the data. I report the 90% confidence intervals for the corruption
coefficients using the 5th and 95th percentiles of the stored coefficients and
report these in brackets in Table 2. While we can be 90% confident that cor-
ruption is linked with more support for fuel subsidies among eligible house-
holds in the unmatched sample (whether measured as a share or as an
indicator), confidence is slightly lower in the matched sample using the boot-
strapping procedure (approximately 85% confidence).19
I also estimate the models at the village level, as corruption measures are
taken at the village level. At the village level, a one standard deviation
increase in the share of subsidy that goes missing is associated with a five
percentage point decrease in the share of eligible households willing to
1492 Comparative Political Studies 51(11)

reduce spending on fuel subsidies. Although not all coefficients on corruption

are significant in the village-level models, all are in the same direction as the
household models (OA6).
I also conduct a variety of robustness checks. I repeat the matching exer-
cise using a threshold of at least 10% subsidy missing to define “treatment”
villages; using a higher threshold should boost confidence that there is a sub-
stantive difference between treatment and control villages in corruption lev-
els. I also reperform the CEM algorithm several times, shifting the values of
the cutpoints for each variable used for matching. Finally, to reduce concerns
about the effect of measurement error of the village population on estimating
corruption, I utilize data on village population from the 2011 Pendataan
Potensi Desa (PODES) to reestimate corruption.20 Results are robust to these
exercises, and corruption consistently predicts attitudes toward subsidy
reform (OA8-10).

Discussion
Although the relationship between local corruption and support for fuel sub-
sidy reform is quite robust empirically, several other pieces of evidence can
increase confidence that local corruption indeed motivates citizens’ attitudes
toward reform. Specifically, I extend the analysis in two directions. First, I
look across policy domains within Indonesia. To show that local corruption
indeed informs opinions on public expenditures based on whether expendi-
tures flow through local officials, I examine public support across a wide
range of policy areas in Indonesia, with varying degrees of delegation to the
local level. Broadly, we should expect to see that local corruption diminishes
support for policy areas that empower local officials and increases support
for policy areas that provide social benefits without relying on local officials.
Second, I show that the argument extends to other contexts using public opin-
ion data from Afrobarometer implemented in Nigeria in 2013. The Nigerian
case illustrates that not only does the argument extend to other contexts but
also that the relationship is robust to using a more direct measure of resis-
tance to fuel subsidy reform and perceptions of corruption.
If the argument presented here is correct, and corruption entrenches atti-
tudes toward fuel subsidies because they are less vulnerable to local manipu-
lation than targeted social safety net programs, then corruption should also
affect attitudes toward other forms of public spending in predictable ways.
Namely, corruption should diminish popular support for any form of public
spending that vests local governments with greater resources and authority
and enhance popular support for public spending on programs that vest
resources and authority with alternative government agencies. Thus, I
Kyle 1493

Figure 4.  Effect of corruption on support for social spending.

Figure 4 plots coefficients and 95% confidence intervals for effect of corruption across
eight separate ordered-logit models. In each model, the dependent variable is the three-
category outcome variable (“−1” = decrease spending on “x” policy, “0” = neither increase
nor decrease spending on “x” policy, “1” = increase spending on “x” policy). All models
include same controls used in Table 2 and district fixed effects. Models are conducted on the
matched sample only. Standard errors clustered by village.

examine the effect of corruption on support for each of the main forms of
social spending in Indonesia.
Figure 4 plots the coefficients and standard errors from eight separate
ordered-logit models conducted on the matched sample only. The dependent
variable in each model is the three-category outcome variable indicating
whether the respondent chose to increase, neither to increase nor decrease, or
to decrease spending on a given policy area. All models include the same
controls reported in Table 2 and district fixed effects (reported in OA7). As
predicted by the argument, corruption in the delivery of targeted transfer pro-
grams predicts attitudes across a broad range of social spending areas.
Corruption is associated with reduced support for targeted transfer programs,
including health fee waivers for the poor, subsidized rice for the poor, and
cash transfers for the poor. Meanwhile, corruption is correlated with more
support for the subsidies on automotive and cooking fuel (though it is not
statistically significant for cooking fuel). Corruption has no significant effect
on support for hiring teachers, building roads, or community development
funds, perhaps because none of these forms of public spending necessarily
empower village officials.
1494 Comparative Political Studies 51(11)

Does corruption explain attitudes toward subsidy reform in other contexts?

Until recently, the Nigerian government subsidized the private consumption
of automotive fuels to maintain a stable price at the pump.21 In 2012, subsidies
on petroleum products were estimated to account for around 20% of the total
public budget (Siddig et al., 2014). When the government attempted to end
fuel subsidies in 2012, a nationwide strike brought Nigeria to a standstill. The
rallying cry of protestors was, “Remove Corruption, Not Subsidies.” Protest
leaders argued that removing one of the few government benefits received by
ordinary citizens given that the government had not tackled corruption and
other forms of wasteful spending was unfair (Nossiter, 2012).
Afrobarometer (2013) inserted a fuel subsidies module into its question-
naire for Nigeria. The survey asked respondent to select which of the follow-
ing statements is closest to his or her view. Statement 1 reads, “The
government should maintain a fuel subsidy to keep fuel prices low and
improve living standards.” While Statement 2 reads, “The fuel subsidy is too
expensive, and the government should remove it.” Overall, 37% of respon-
dents agree with Statement 2. This level of overall support for fuel subsidy
retrenchment is remarkably similar to Indonesia, where 39% of respondents
agreed with fuel subsidy reform.
I measure corruption in local government using a question that asks
respondents to identify whether “none,” “some of,” “most of,” or “all of”
local government councilors are involved in corruption. For controls, I
attempted to follow the Indonesia models as closely as possible. Table 3
reports results: where the dependent variable indicates whether the respon-
dent agreed with Statement 2, neither agreed or disagreed with either state-
ment, or agreed with Statement 1. As in Indonesia, belief that local officials
are corrupt is strongly associated with resistance to fuel subsidy reform. This
adds confidence both that the argument travels across countries and also that
the findings are not driven by how the concepts are operationalized (e.g.,
perceived vs. estimated corruption).

Conclusions
The size and prevalence of consumer subsidies on fossil fuels has become an
area of increasing global concern. Yet scholars have only recently begun to
evaluate why countries enact these subsidies and why they are so difficult to
repeal. This article joins the emerging literature by investigating the influ-
ence of local corruption on attitudes toward subsidy reform. Local corrup-
tion influences attitudes toward subsidies because shifting government
resources away from social protection programs based on universal access—
such as automotive fuel subsidies—to social protection programs based on
Kyle 1495

Table 3.  Support for Fuel Subsidies in Nigeria.

Variables (1)
Corruption, local government 0.290**
(0.083)
Vehicle ownership −0.021
(0.151)
Urban 0.072
(0.118)
Education 0.019
(0.033)
Male 0.151
(0.110)
Poverty index −0.039
(0.078)
N 2,318
†p < .10. *p < .05. **p < .01. Standard errors clustered by region in parentheses.

targeting the poor requires vesting authority for policy implementation with
local politicians and bureaucrats. When these local elite use this discretion
to divert resources for private gain, citizens are warier of reform efforts that
would place more resources in their hands. This article investigates these
expectations empirically using household survey data from Indonesia.
Findings show that corruption in the implementation of targeted transfer
programs—estimated using a “gap” measurement method—is associated
with more resistance to reforming consumer automotive fuel subsidies.
Results remain consistent when matching is used and to a number of
robustness checks.
This finding has important implications for current policy discussions on
fuel subsidy reform. Compensating the poor for increases in fuel prices
through improved social policy is generally accepted as key for successful
reform. However, countries may utilize fuel subsidies as a redistributive tool
precisely because they lack the institutional capability to implement alterna-
tive forms of social policy (Victor, 2009). Building this capability, including
at the local levels, can be an essential component of gaining public support
for reform. Absent this capability, even the rural poor, who are thought to
benefit little from fuel subsidies, may resist reforms.22
As this type of capability can only be built over the long-term, it may be
more fruitful in the interim to design reform packages that rely on existing
patterns of institutional strengths and citizen trust, even if the reforms appear
to be less efficient. For example, a more coarsened approach to targeting (e.g.,
1496 Comparative Political Studies 51(11)

everyone within a village) would rely less on the information provided by

local officials and allow for less discretion. Similarly, if there is an institution
or agency that has high levels of citizen trust, entrusting it with reform imple-
mentation could diminish resistance to reforms. However, just because low
trust in local institutions is associated with reduced support for reform does
not necessarily mean that citizens would support reforms if compensatory
policies were implemented nationally rather than locally, as trust in national
governments is frequently lower than trust in local governments, including in
Indonesia (Olken, 2009). Exactly how we might expect trust in institutions to
shape support for reforms depends on the particular institutions that would be
authorized to implement reforms as well as on citizens’ views toward these
institutions.
While this article identifies a key source of opposition to reform, it also
points to several encouraging trends. First, these findings suggest that much
opposition to fuel subsidy reform comes not from a fundamental policy dis-
agreement or an unwillingness to curtail benefits—in fact, 39% of surveyed
citizens reported willingness to decrease government spending on automo-
tive fuel subsidies—but from a lack of confidence that public officials will
deliver on the programs they promise will replace subsidies. This suggests
that citizens’ views could be changed over time, if trust in public institutions
can be enhanced. Second, the history of fuel subsidy reforms in Indonesia
illustrates how this confidence can be built over time: In 1998, fuel subsidy
reforms contributed to the toppling of a regime, while just 16 years later, a
major reform effort in 2014 passed with little backlash. No doubt the varying
levels of popular support for the regime implementing the reform played sig-
nificant roles, but Indonesia’s growing administrative capabilities, including
in the ability to implement targeted transfer programs, likely also contributed
to these divergent outcomes. Many villages in Indonesia at the time of the
survey exhibit quite low levels of corruption, and, in these villages, poor citi-
zens do support reform.
In some ways, Indonesia may be a particularly hopeful case for the long-
term prospects of reform, given that every regime since the fall of Soeharto
has attempted fuel subsidy reform. This suggests that across the political
spectrum, there is consensus among political elites on the need for reform.
However, these dynamics are by no means unique to Indonesia, nor are the
fiscal pressures motivating reform unique to Indonesia’s status as a net
importer. Low trust in institutions plays a role in undermining support for
reforms in Nigeria, an oil exporter, as shown in the article. Even in Saudi
Arabia, fiscal pressure to reform fuel subsidies is high, and energy subsidy
reform is a key part of the government’s reform plan to adjust to lower global
oil prices and a significantly diminished government budget. The govern-
ment proposes a compensatory, targeted program for poor and middle-income
Kyle 1497

households to offset the effects of reforms, though both the reform and the
targeted program were still delayed as of mid-2017.23
More broadly, similar dynamics could play out for any type of economic
reform which entails transferring resources through institutions in which citi-
zens have varying levels of trust. In evaluating economic reforms with adjust-
ment costs, it is therefore equally important to consider how reforms,
including any compensatory social programs, are implemented in practice as
to consider the economic winners and losers from reforms. Even citizens who
seem to be economic “winners” from reforms may have a vested interest in
the status quo if they expect reforms to be implemented imperfectly.
This article also has several broader implications. The results show
that local policy implementation can significantly shape patterns of sup-
port for national policies. The way that individuals experience social poli-
cies varies significantly depending on how local politicians and
bureaucrats implement these policies, even in developed countries. This
is an important and understudied dimension of attitudes toward redistri-
bution and social policy. The results also draw attention to consumer sub-
sidy programs as significant components of welfare states in developing
countries. Future work could usefully examine across a wider range of
cases how countries shift over time from welfare states based on broad-
based consumer subsidy programs to systems which more narrowly target
the poor.

Supplemental Material
Supplementary material for this article is available online at the CPS website
http://journals.sagepub.com/doi/suppl/10.1177/0010414018758755.

Acknowledgments
I am very grateful to Abhijit Banerjee, Rema Hanna, Benjamin Olken, and Sudarno
Sumarto for supporting the inclusion of the survey questions used here. I thank
Allison Carnegie, Jeff Colgan, Rema Hanna, Macartan Humphreys, Yotam
Margalit, Maria Victoria Murillo, Benjamin Olken, Michael Ross, Sudarno
Sumarto, Johannes Urpelainen, and three anonymous reviewers for valuable
comments.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.

Funding
The author disclosed receipt of the following financial support for the research,
authorship, and/or publication of this article: The author acknowledges support
1498 Comparative Political Studies 51(11)

from the National Science Foundation Graduate Research Fellowship and from the
CGIAR Research Program on Policies, Institutions, and Markets (PIM), which is
led by the International Food Policy Research Institute (IFPRI) and funded by
CGIAR Fund donors. The article uses data from a project that was financially sup-
ported by AusAid through the Poverty Reduction Support Facility. All views
expressed are my own and do not reflect the views of any individuals or institutions
acknowledged here.

ORCID ID
Jordan Kyle https://orcid.org/0000-0003-0551-8047

Notes
  1. Consumer automotive fuel subsidies occur when governmental policies lower
the price on automotive fuels paid by end users. The size of the subsidy is the
gap between the domestic retail price and the economic price of the product,
which is determined by the wholesale price of the refined fuel product on the
international market; the costs of transportation, distribution, and storage; and
profit margins by retail outlets. Although many different types of fuel subsidies
exist—including producer subsidies and subsidies on different types of cooking
and automotive fuels—I use the term “fuel subsidies” in the context of Indonesia
to refer to consumer subsidies on automotive fuel for convenience.
  2. Compensating the losers from economic reforms is a commonly cited strategy
for reducing political backlash; for example, welfare protection may diminish
political backlash to opening to trade (Burgoon, 2009; Rodrik, 1998).
  3. Even among eligible households, not all households may be equally concerned
that corruption will reduce access to targeted transfer programs. Those that are
more densely connected to local patronage networks, for example, may be more
willing to support reform, even if local officials are corrupt because they could
have lower expectations that corruption will reduce their access to benefits com-
pared to those who are not connected to local patronage networks. Testing for
differences within villages in the effects of corruption based on connectedness to
clientelistic or patronage networks would be an interesting future research direc-
tion. I thank an anonymous review for making this point.
 4. Although reducing the subsidy was an International Monetary Fund (IMF)
requirement for the loan, it is not known why Soeharto implemented the price
hike so suddenly and by so much. The IMF, fearing social unrest, was advocating
a gradual approach at the time (Beaton, Lontoh, & Wai-Poi, 2017).
  5. In 2004, Indonesia’s Constitutional Court ruled that a formula-based system for
pricing fuel—which would have linked retail to market prices—was unconstitu-
tional because it violated Article 33 of the Indonesian Constitution, which speci-
fies that natural resources must be controlled by the State and used to benefit the
people (Beaton et al., 2017).
  6. Prior to Jokowi’s election, regimes in Indonesia post-Soeharto era were charac-
terized by “promiscuous powersharing,” overly broad ruling coalitions which
Kyle 1499

included Indonesia’s ruling oligarchy, limiting true political competition (Slater

& Simmons, 2013).
  7. The survey was fielded by SurveyMetre, an independent and widely respected
Indonesian survey company. Data from two other survey rounds conducted in the
same village—one in October and November 2012 and the other in December
2013 and January 2014 (Banerjee, Hanna, Kyle, Olken, & Sumarto, in press-a,
in press-b)—are used in some of the analysis. However, data on policy attitudes
were only collected in the March-May 2013 survey round.
  8. Demographic data from The World Bank (2012b).
  9. Exact question wording: “The government has a number of programs to help
the poor. Imagine that you could plan the government’s budget this year. If the
government could increase [had to decrease] the budget for only three of the
following programs, which programs would you select?” Question wording is
similar to that used by Rehm, Hacker, and Schlesinger (2012) in their cross-
domain analysis of social policy support in the United States. In prompting the
idea of “programs to help the poor,” it is worth noting that 96% of respondents
self-identify as being “less well off” compared with others.
10. It is possible that presidential approval affects responses to the broad spending
question as well, though less likely than with a direct question on the recent fuel
subsidy reform package. Inclusion of district fixed effects controls for variation
across districts in presidential approval.
11. Ideally, I would have randomized the ordering of the policies on the list in the
survey questions. It is possible that the ordering on the list affected the response
rates to certain policies. Although this could affect the overall rates of support
indicated by the survey, it should not affect the relationship between corruption
and attitudes.
12. Olken (2006) used this method to estimate leakage in the same program.
13. ([Market price for rice – copay price] × 5,550 kg) / (IDR / USD).
14. It is possible that some rice goes missing before it reaches the warehouse for
local officials to pick up. Yet, over 70% of the overall variance in missing rice
is between villages rather than between warehouses, and only 1% of Raskin dis-
tributors report receiving less than the full quota at the warehouse (Banerjee et al.,
in press-a).
15. Looking across multiple months and survey waves is important, as local officials
could skim from the program by taking a little off the top each month or by dis-
tributing rice in some months but not in others.
16. Past corruption also predicts the share of eligible, poor households that do not
receive the cash transfer within the village as well as the share of eligible, poor
households that do not receive the identification cards that are supposed to give
them access to the transfer. Available upon request.
17. Other matching algorithms cannot ensure that balance will be improved for all
covariates and can often worsen imbalance for some covariates while improving
it for others (Iacus, King, & Porro, 2012).
18. Cutpoints are reported in OA3.
1500 Comparative Political Studies 51(11)

19. In part, this is due to the coarsened exact matching procedure, which drops vil-
lages for which there is not an exact match on all matching covariates, resulting
in varying sample sizes for each draw of data.
20. Data from Badan Pusat Statistik (2011).
21. Nigeria removed fuel subsidies in 2016 by raising pump prices above newly low
market prices. However, as in Indonesia, they did not deregulate the pricing of
fuel. Reforms will thus be tested as market prices rise.
22. This could help explain why many countries in Southeast Asia maintain con-
sumer subsidy programs despite an overall rural bias in policymaking. On rural
bias in Southeast Asia, see Pierskalla (2016).
23. See Mahdi and Nereim (2017).

References
Afrobarometer. (2013). Nigeria (Round 5). Retrieved from http://afrobarometer.org/
countries/nigeria-0
Alderman, H. (2002). Subsidies as a social safety net: Effectiveness and challenges
(Social Protection Discussion Paper Series). Washington, DC: The World Bank.
Anderson, C., & Tverdova, Y. (2003). Corruption, political allegiances, and atti-
tudes towards governments in contemporary democracies. American Journal of
Political Science, 47, 91-109.
Arze del Granado, J., Coady, D., & Gillingham, R. (2012). The unequal benefits of fuel
subsidies: A review of evidence for developing countries. World Development,
40, 2234-2248.
Auerbach, A. (2016). Clients and communities. World Politics, 68, 111-148.
Badan Pusat Statistik [Central Bureau of Statistics]. (2011). Pendataan Potensi Desa
[Data on Village Potential]. Jakarta, Indonesia: Badan Pusat Statistik.
Banerjee, A., Hanna, R., Kyle, J., Olken, B., & Sumarto, S. (in press-a). The role of
competition in effective outsourcing: Subsidized food distribution in Indonesia.
Journal of Political Economy.
Banerjee, A., Hanna, R., Kyle, J., Olken, B., & Sumarto, S. (in press-b). Tangible
information and citizen empowerment: Identification cards and food subsidy pro-
grams in Indonesia. Journal of Political Economy.
Bardhan, P., & Mookherjee, D. (2005). Decentralizing antipoverty program delivery
in developing countries. Journal of Public Economics, 89, 675-704.
Bates, R. (1981). Markets and states in tropical Africa: The political basis of agricul-
tural policies. Berkeley: University of California Press.
Bayu, D. J. (2016, November 8). Penyaluran Kartu Indonesia Pintar Capai 70 Persen
[Distribution of Indonesia’s smart card reaches 70 percent]. Kompas. Retrieved
from http://nasional.kompas.com/read/2016/11/08/22563281/penyaluran.kartu.
indonesia.pintar.capai.70.persen
Beaton, C., Lontoh, L., & Wai-Poi, M. (2017). Indonesia: Pricing reforms, social
assistance, and the importance of perceptions. In G. Inchauste & D. Victor (Eds.),
The political economy of energy subsidy reform (Chapter 4). Washington, DC:
The World Bank.
Kyle 1501

Bellemare, M. (2015). Rising food prices, food price volatility, and social unrest.
American Journal of Agricultural Economics, 97, 1-21.
Benes, K., Cheon, A., Urpelainen, J., & Yang, J. (2016). Low oil prices: An opportu-
nity for fuel subsidy reform. International Growth Center. Retrieved from http://
www.theigc.org/blog/low-oil-prices-an-opportunity-for-fuel-subsidy-reform/
Bril-Mascarenhas, T., & Post, A. (2015). Policy traps: Consumer subsidies in
post-crisis Argentina. Studies in Comparative International Development, 50,
98-120.
Burgoon, B. (2009). Globalization and backlash: Polanyi’s revenge? Review of
International Political Economy, 16, 145-177.
Cameron, L., & Shah, M. (2014). Can mistargeting destroy social capital and stim-
ulate crime? Evidence from a cash transfer program in Indonesia. Economic
Development and Cultural Change, 62, 381-415.
Campbell, A. L. (2012). Policy makes mass politics. Annual Review of Political
Science, 15, 333-351.
Cheon, A., Lackner, M., & Urpelainen, J. (2015). Instruments of political con-
trol: National oil companies, oil prices, and petroleum subsidies. Comparative
Political Studies, 48, 370-402.
Cheon, A., Urpelainen, J., & Lackner, M. (2013). Why do governments subsidize
gasoline consumption? An empirical analysis of global gasoline prices, 2002-
2009. Energy Policy, 56, 382-390.
Clausen, B., Kraay, A., & Nyiri, Z. (2011). Corruption and confidence in public
institutions: Evidence from a global survey. World Bank Economic Review, 25,
212-249.
Clements, B., Coady, D., Fabrizio, S., Gupta, S., Alleyne, T., & Sdralevich, C. (Eds.).
(2013). Energy subsidy reform: Lessons and implications. Washington, DC:
International Monetary Fund.
Clements, B., Jung, H.-S., & Gupta, S. (2007). Real and distributive effects of petro-
leum price liberalization: The case of Indonesia. The Developing Economies, 45,
220-237.
Davis, L. (2014). The economic cost of global fuel subsidies. American Economic
Review: Papers & Proceedings, 104, 581-585.
Hadiz, V. (2010). Localising power in post-authoritarian Indonesia: A Southeast Asia
perspective. Palo Alto, CA: Stanford University Press.
Haggard, S., & Birdsall, N. (2000). The social fallout: Safety nets and recrafting
the social contract. In S. Haggard (Ed.), The political economy of the Asian
financial crisis (Chapter 5). Washington, DC: Peterson Institute for International
Economics.
Hainmueller, J. (2012). Entropy balancing for causal effects: A multivariate reweight-
ing model to produce balanced samples in observational studies. Political
Analysis, 20, 25-46.
Hainmueller, J., & Xu, Y. (2013). ebalance: A stata package for entropy balancing.
Journal of Statistical Software, 54(7), 1-18.
Hakhverdian, A., & Mayne, Q. (2012). Institutional trust, education, and corruption:
A micro-macro interactive approach. Journal of Politics, 74, 739-750.
1502 Comparative Political Studies 51(11)

Hetherington, M. (2005). Why trust matters: Declining political trust and the demise
of American liberalism. Princeton, NJ: Princeton University Press.
Ho, D., Imai, K., King, G., & Stuart, E. (2007). Matching as nonparametric prepro-
cessing for reducing model dependence in parametric causal inference. Political
Analysis, 15, 199-236.
Iacus, S., King, G., & Porro, G. (2012). Causal inference without balance checking:
Coarsened exact matching. Political Analysis, 20, 1-24.
International Institute for Sustainable Development. (2012). A citizen’s guide to
energy subsidies in Indonesia: 2012 update. Winnipeg, CA.
International Institute for Sustainable Development. (2015). Indonesia energy subsidy
briefing (November 2015). Winnipeg, CA.
Jacobs, A., & Matthews, J. S. (2017). Policy attitudes in institutional context: Rules,
uncertainty, and the mass politics of public investment. American Journal of
Political Science, 61, 194-207.
Jaspars, S., & Shoham, J. (1999). Targeting the vulnerable: A review of the necessity
and feasibility of targeting vulnerable households. Disasters, 23, 359-372.
Jones Luong, P., & Weinthal, E. (2010). Oil is not a curse: Ownership structures and
institutions in Soviet successor states. Cambridge: Cambridge University Press.
Kraay, A., & Murrell, P. (2016). Misunderestimating corruption. Review of Economics
and Statistics, 98, 455-466.
Ladislaw, S., & Cuyler, Z. (2015). Adjusting to low prices: Prospects for fossil-fuel
subsidy reform in oil-producing and exporting countries. Washington, DC:
Center for Strategic & International Studies.
Lupu, N., & Riedl, R. B. (2013). Political parties and uncertainty in developing
democracies. Comparative Political Studies, 46, 1339-1365.
Mahdi, W., & Nereim, V. (2017, July 6). Saudi Arabia delays energy-subsidy cuts.
Bloomberg. Retrieved from https://www.bloomberg.com/news/articles/2017-07-06/
saudis-are-said-to-delay-energy-subsidy-cuts-amid-economic-pain
Nossiter, A. (2012, January 9). Nigerians protest rise in oil prices. The New York
Times, A9.
Olken, B. (2006). Corruption and the costs of redistribution. Journal of Public
Economics, 90, 853-870.
Olken, B. (2009). Corruption perceptions vs. corruption reality. Journal of Public
Economics, 93, 950-964.
Pierskalla, J. (2016). The politics of urban bias: Rural threats and the dual dilemma of
political survival. Studies in Comparative International Development, 51, 286-307.
Poppele, J., Sumarto, S., & Pritchett, L. (2000). Social impacts of the Indonesian cri-
sis: New data and policy implications. Jakarta, Indonesia: The SMERU Research
Institute.
Rehm, P., Hacker, J., & Schlesinger, M. (2012). Insecure alliances: Risk, inequality,
and support for the welfare state. American Political Science Review, 106, 386-406.
Rodrik, D. (1998). Globalisation, social conflict, and economic growth. The World
Economy, 21, 143-158.
Ross, M., Hazlett, C., & Mahdavi, P. (2017). Global progress and backsliding on
gasoline taxes and subsidies. Nature Energy, 2, Article 16201.
Kyle 1503

Rudolph, T., & Evans, J. (2005). Political trust, ideology, and public support for gov-
ernment spending. American Journal of Political Science, 49, 660-671.
Seligson, M. (2002). The impact of corruption on regime legitimacy: A comparative
study of four Latin American countries. Journal of Politics, 64, 408-433.
Setiawan, S. R. D. (2014, January 27). JK: Rasio Ideal Subsidi BBM 10 Persen dari
APBN [Jokowi: 10 percent is the ideal ratio for fuel subsidies in state budget].
Kompas. Retrieved from http://ekonomi.kompas.com/read/2014/01/27/1503088/
JK.Rasio.Ideal.Subsidi.BBM.10.Persen.dari.APBN
Siddig, K., Aguiar, A., Grethe, H., Minor, P., & Walmsley, T. (2014). Impacts of
removing fuel import subsidies in Nigeria on poverty. Energy Policy, 69, 165-
178.
Slater, D., & Simmons, E. (2013). Coping by colluding: Political uncertainty and pro-
miscuous powersharing in Indonesia and Bolivia. Comparative Political Studies,
46, 1366-1393.
Smith, T. G. (2014). Feeding unrest: Disentangling the causal relationship between
food price shocks and sociopolitical conflict in urban Africa. Journal of Peace
Research, 51, 679-695.
Soss, J. (1999). Lessons of welfare: Policy design, political learning, and political
action. American Political Science Review, 93, 363-380.
Suryowati, E. (2014, May 21). Jokowi-JK Janji Subsidi BBM Turun Rp 60 Triliun
[Jokowi promises to reduce gas subsidy by Rp 60 trillion]. Kompas. Retrieved
from http://ekonomi.kompas.com/read/2014/05/21/1418596/Jokowi-JK.Janji.
Subsidi.BBM.Turun.Rp.60.Triliun
Timmer, P. (1993). Rural bias in the East and South-East Asian rice economy:
Indonesia in comparative perspective. Journal of Development Studies, 29, 149-
176.
Victor, D. (2009). The politics of fossil-fuel subsidies. Winnipeg, CA: International
Institute for Sustainable Development.
Weaver, V., & Lerman, A. (2010). Political consequences of the carceral state.
American Political Science Review, 104, 817-833.
Weitz-Shapiro, R. (2008). The local connection: Local government performance and
satisfaction with democracy in Argentina. Comparative Political Studies, 41,
285-308.
Wlezien, C. (1995). The public as thermostat: Dynamics of preferences for spending.
American Journal of Political Science, 39, 981-1000.
The World Bank. (2009). Tapping a hidden resource: Energy efficiency in the Middle
East and North Africa. Washington, DC: Author.
The World Bank. (2012a). Raskin subsidized rice delivery: Social assistance program
and public expenditure review. Jakarta, Indonesia: Author.
The World Bank. (2012b). Targeting poor and vulnerable households in Indonesia.
Jakarta, Indonesia: Author.

Author Biography
Jordan Kyle is a visiting research fellow at the International Food Policy Research
Institute in Washington, DC. Her research focuses on citizen-state relations, social
policy, and political attitudes.
 Bulletin of Indonesian Economic Studies

ISSN: 0007-4918 (Print) 1472-7234 (Online) Journal homepage: https://www.tandfonline.com/loi/cbie20

Tax Non-Compliance and Perceptions of

Corruption: Policy Implications for Developing
Countries

Arifin Rosid, Chris Evans & Binh Tran-Nam

To cite this article: Arifin Rosid, Chris Evans & Binh Tran-Nam (2018) Tax Non-Compliance and
Perceptions of Corruption: Policy Implications for Developing Countries, Bulletin of Indonesian
Economic Studies, 54:1, 25-60, DOI: 10.1080/00074918.2017.1364349

To link to this article: https://doi.org/10.1080/00074918.2017.1364349

Accepted author version posted online: 07

Aug 2017.
Published online: 02 Aug 2018.

Submit your article to this journal

Article views: 2304

View related articles

View Crossmark data

Citing articles: 2 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=cbie20
 Bulletin of Indonesian Economic Studies, Vol. 54, No. 1, 2018: 25–60

TAX NON-COMPLIANCE AND PERCEPTIONS

OF CORRUPTION: POLICY IMPLICATIONS
FOR DEVELOPING COUNTRIES

Arifin Rosid Chris Evans

Directorate General of Taxation, Ministry of University of Pretoria;
Finance of the Republic of Indonesia UNSW Sydney

Binh Tran-Nam
UNSW Sydney; RMIT University Vietnam

Tax non-compliance and perceptions of corruption are key challenges to state-building

in developing countries. Using a social psychology approach, we develop a theoretical
model in which different forms of perceived corruption can adversely influence the
way individual taxpayers behave. We then apply this model to Indonesia, placing our
empirical findings in the context of compliance risk management, identifying strategies
to improve tax compliance, and exploring how to implement these strategies effectively.
We shed light on the applicability of the traditional responsive regulatory approach
(used by revenue authorities to deal with intentionally non-compliant taxpayers),
which combines measures in attempting to achieve either voluntary or enforced
compliance. While the empirical evidence is based on the Indonesian experience, we
suggest that our model is sufficiently general and robust to be applicable to other
developing countries in the Asia-Pacific region.

Keywords: corruption, tax compliance, compliance risk management, norm-based intervention,

focused enforcement
JEL classification: D03, H20, H24, H26

INTRODUCTION
Revenue collection in developing countries is generally afflicted by widespread
evasion, coercion, and corruption (McKerchar and Evans 2009; Bird 2015).
Corruption, in particular, creates inefficient tax systems; erodes the legitimacy
of tax collection legitimacy; reduces the willingness of individuals and corporate
entities to pay their fair share of taxes; and, as a result, reduces tax revenue (Imam
and Jacobs 2007; Alm, Martinez-Vazquez, and McClellan 2016).1 Many authors

1. In the literature on corruption in general, there are two competing views on the impact of
corruption on economic efficiency: the ‘sand the wheels’ and ‘grease the wheels’ hypotheses
(see Nguyen et al. 2016). In the context of tax corruption, we are inclined to support the
‘sand the wheels’ hypothesis—that is, that tax corruption is harmful.

ISSN 0007-4918 print/ISSN 1472-7234 online/18/00025-36 © 2018 ANU Indonesia Project

http://dx.doi.org/10.1080/00074918.2017.1364349
26 Arifin Rosid, Chris Evans, and Binh Tran-Nam

recommend that developing countries looking to increase tax compliance must

first reduce corruption (McKerchar and Evans 2009; OECD 2012).
Due to corruption’s secretive nature, most of the indicators used to identify and
assess it are based on subjective measurements, typically known as ‘perceptions
of corruption’ (Campbell 2013). Perceptions of corruption can also have a negative
impact on the way taxpayers behave. That is, it may erode taxpayers’ willingness
to contribute their fair share of tax (Jahnke 2015). It may also cultivate a culture
of distrust among stakeholders towards related institutions (Melgar, Rossi, and
Smith 2010), strengthen already damaging public perceptions that cause taxpayers
to disengage from reciprocal relationships with the government (Fjeldstad and
Tungodden 2003), and reduce the moral cost of evading tax and encourage taxpayers
to behave opportunistically (Torgler et al. 2008). Perceptions of corruption can also
undermine compliance—taxpayers may assume that corruption will reduce any
government expenditure they receive (Alm, McClelland, and Schulze 1992), or they
may underreport their taxes in order to have more money with which to bribe tax
officials (Çule and Fulton 2009). Consequently, perceptions of corruption may be
more harmful than corruption itself (Melgar, Rossi, and Smith 2010).
Yet despite corruption being a serious problem, particularly in developing
countries, research into perceptions of corruption in the taxation context has received
relatively little attention, and empirical evidence of its impact on intentional tax
(non-)compliance is minimal. To what extent (and precisely how) does perceived
corruption affect tax compliance? Are there policies that a revenue authority—such
as Indonesia’s Directorate General of Taxation (DGT)—can introduce, or refine,
to improve compliance? These are the major questions we address in this article.
We focus on Indonesia for two main reasons. First, Indonesia is among the
more corrupt of the major countries in the Asia-Pacific region (Transparency
International 2016). Kuncoro (2004) found that the typical bribe paid by Indonesian
firms is 10%–15% of annual production costs, while survey results from the World
Economic Forum’s (2016) Global Competitiveness Report 2016–2017 suggest that
corruption is the biggest deterrent to those looking to do business in Indonesia.
Second, Indonesia has been classified as a major developing country with ongoing
compliance issues (OECD 2014b; DGT 2015a). Indonesia’s tax ratio is lower than
those of other low-income countries (IMF 2011), and, due largely to corruption and
poor governance, its percentage of individual income tax revenue to GDP is lower
than in neighbouring countries (that is, 1.3%, compared with 1.9% in Thailand, 2.1%
in the Philippines, and 2.7% in Malaysia) (Bird and Zolt 2005).
Indonesia recently ran a nationwide tax amnesty program for nine months, from
July 2016 to March 2017, in an attempt to generate more revenue and broaden the
tax base. Yet in the first phase of the program (July−September 2016, involving a
2% redemption rate), 82% (284,700 of 347,000) of participants were taxpayers who
had already lodged their annual tax returns and 89% ($6.64 billion of $7.46 billion)
of the tax declared was paid by individual taxpayers.2 This underlines the high
level of non-compliance by individual taxpayers in Indonesia.
While the empirical findings of this article are confined to Indonesia, the
theoretical model (developed in the next section) is sufficiently general and robust
to be applicable to other developing countries, especially those in the Asia-Pacific

2. Data from http://pajak.go.id/statistik-amnesti (accessed 27 Oct. 2016).

Tax Non-compliance and Perceptions of Corruption 27

FIGURE 1 Types of Perceived Corruption under Study

      Levels of public

officials involved

High Low

Grand Petty
General corruption corruption

Types of public General

sectors involved corruption

Tax
Grand Petty
tax corruption tax corruption

region. Different model settings may lead to different results, but the model does
not depend on any country-specific characteristics or assumptions.
There is no single definition of corruption (see, for example, Blackburn, Bose,
and Haque 2010). Most definitions, however, emphasise the governmental aspects
of corruption, by highlighting the abuse of public authority for personal gains―
commonly termed ‘government corruption’ (Shleifer and Vishny 1993, 599). A
popular way of classifying corruption is by referring to its scale (UNDP 2008). In
this sense, corruption is categorised by the sector in which it occurs or the amount
of money it involves. Doig and Theobald (1999), for example, classify two types
of corruption: grand and petty. Grand corruption represents the misuse of public
power by high-level public officials, such as ministers or senior staff, for personal
pecuniary gain. Petty corruption refers to the extortion of small payments by low-
level public officials in everyday interactions designed to ‘smooth’ transactions
(accordingly, such payments are often called ‘grease money’). We adopt this
classification in this article.
The Indonesian tax authority is far from immune to corruption (McLeod 2011).
To the best of our knowledge, no previous study has distinguished between general
corruption and tax corruption. In this article, we distinguish corruption in the DGT
from corruption in all other government sectors. We identify five different types
of perceived corruption: grand corruption; grand tax corruption; petty corruption;
petty tax corruption; and general corruption, which includes all of these types
(figure 1).
To maintain the specificity of the study, we then review the definition of tax
compliance behaviour. While there is no consensus of what precisely is meant
by compliant behaviour, we identify that the definitions of tax compliance can
be broadly categorised into two approaches: conceptual and operational. The
conceptual approach emphasises taxpayers’ willingness to comply without
28 Arifin Rosid, Chris Evans, and Binh Tran-Nam

enforcement (see, for example, James and Alley 1999, 32), whereas the operational
approach focuses on the administrative fulfilment of particular tax obligations (see,
for example, OECD 2014a). For practical purposes, the operational approach seems
to offer an advantage by evaluating whether taxpayers register in the system, file
timely tax returns, report accurately and fully, and pay their tax liabilities on time.
We also observe that two critical aspects need to be considered in examining
tax compliance behaviour: intention and outcome (see, for example, Bird 2015;
World Bank 2015). While the outcomes of compliance behaviour vary (Tran-Nam
2003; Langham, Paulsen, and Hartel 2012), intention can be considered an accurate
proxy for performed behaviour (Lewis 1982; OECD 2010). For this reason, we
set out two conceptual boundaries. First, we define tax compliance as ‘taxpayers’
willingness to accurately report income in accordance with the prevailing tax
law’ (OECD 2014a).3 Second, so far as tax reporting is concerned, we focus on
both intentional compliance and intentional non-compliance. Accordingly, tax
compliance behaviour in this article refers to the self-reporting of taxpayers’
compliance or non-compliance.
By identifying the impact that perceptions of corruption have on intentional non-
compliance, this study expands the increasingly popular theory of tax morale (see,
for example, Brink and Porcano 2016; Cummings et al. 2009; Halla 2012; Jahnke
2015).4 This study provides deep insights into a hitherto unexplored cultural
factor that may affect the tax morale of individuals. This is important because the
traditional economic model, which generally relies on traditional enforcement
methods through audits and penalties, is not capable of explaining the level of
tax compliance in society. That is, the evidence suggests that in many countries
the level of tax evasion should be higher than it is, given the low probability of
getting audited and the relatively small penalty if caught (Walsh 2012). Accordingly,
because deterrence theory often does not work well in practice (Osofsky 2014), an
extension to the standard model is needed to adequately explain the dynamics
of tax compliance (Ritsatos 2014). In response, we examine to what extent, and
how, perceptions of corruption influence intentional underreporting by personal
income taxpayers (PITs) in Indonesia. Public perception is one of the external
factors capable of shaping tax compliance (OECD 2014a).
While some empirical parts of this article draw heavily on a recent publication
of ours (Rosid, Evans, and Tran-Nam 2016), here we analyse these empirical
findings in the context of risk-based compliance management, in order to identify
strategies designed to improve tax compliance and to explore the necessary steps
to implement them effectively. The policy implications of this study are relevant
for developing economies such as Indonesia, particularly in relation to the United
Nations 2030 Agenda for Sustainable Development, to which increasing the
mobilisation of domestic revenue and improving domestic capacity for tax revenue
collection are central (United Nations 2015). Finally, in many cases decisions in tax
administration need to be made under uncertainty (Wenzel and Taylor 2003). This
article provides evidence-based results that may be beneficial, particularly to the
DGT, in reducing ex ante uncertainty.

3. This definition deliberately excludes the ‘registration’, ‘filing’, and ‘payment’ criteria.
4. Tax morale is frequently described as the intrinsic motivation to pay tax (Frey and Torgler
2007).
Tax Non-compliance and Perceptions of Corruption 29

METHODOLOGY
The primary concern in studying perceptions of corruption and tax compliance
behaviour is how to obtain reliable data. For this purpose, field research was
undertaken during January–July 2015. This study sought to measure perceived
levels of corruption and tax compliance behaviour at the individual level by asking
Indonesian PITs the extent of perceived corruption and tax underreporting. Such
questions invariably depend on context and are susceptible to social desirability
bias, or the possibility that taxpayers will not answer truthfully because of their
desire to accord with their peers. We used a number of strategies to minimise
respondent bias.
First, we included both qualitative and quantitative paradigms in a sequential
priority model. We used in-depth interviews with nine participants (three
taxpayers, three tax agents, and three tax officers) to clarify, modify, and develop
more robust observed variables in the design of the survey questionnaire from
theoretical perspectives. We then conducted an extensive survey of PITs who
were submitting personal annual tax returns (surat pemberitahuan tahunan) in 12
randomly selected tax offices across four Indonesian regions (Denpasar, Malang,
Surabaya, and Yogyakarta). Most of the offices we selected were in Java (75%).
The choice of locations was dictated by economic activity: despite its relatively
small size, Java accounts for almost 58% of Indonesia’s GDP (BPS 2013) and its 189
tax offices (57% of the total in Indonesia) administer 60% of Indonesia’s personal
taxpayers. However, our research budget allowed us to survey only 397 taxpayers
in 12 offices.5 Due to confidentiality, the options for collecting tax compliance data
are few. We therefore considered combined random sampling and convenience
sampling the most feasible options for this research.
Second, to minimise taxpayers’ suspicion of the data collection in the quantitative
phase, we involved a team of research intermediaries—trained research assistants
recruited from local universities—in the survey. The intermediaries played a crucial
role in assuring the taxpayers that the study was an academic project, as well as
indicating that the study had no link to the DGT. The survey comprised mixed
modes: face-to-face interviews and self-completion questionnaires. Employed
and self-employed PITs who attended the observed tax offices were selected at
random by the intermediaries and together treated as our accessible population.
Face-to-face interviews were used to recruit respondents, and then self-completion
questionnaires were used to capture respondents’ sensitive answers.
Third, two-stage pilot tests were undertaken before the survey took place, to
ensure that the questionnaire would collect the most accurate and reliable data
possible. Further, to reduce potential social desirability bias, we used an ‘everybody
does it’ technique (De Vaus 2014): the questionnaire included a preamble stating
that ‘not everyone necessarily reports all of his or her actual income to the tax
office’. We emphasised the anonymous nature of the survey by placing the section
that collected background information about respondents on the last page of the
questionnaire.

5. We estimated the study’s accessible population to be around 360,000, having taken the
number of registered PITs obliged to lodge annual tax returns (as of December 2012) and
multiplied it by the filing ratio of each tax office under study. Our structural equation
modelling stipulated that our minimum sample of taxpayers would be 384.
30 Arifin Rosid, Chris Evans, and Binh Tran-Nam

Finally, because intentional (non-)compliance is the focus of our study, we

adopted Ajzen’s (1991) theory of planned behaviour (TPB) as our primary
theoretical lens. By definition, perception is closely related to belief.6 The TPB deals
with three beliefs: behavioural, normative, and control. Once beliefs associated
with a certain behaviour have been developed, they provide the basis for the
attitudes, subjective norms, and perceived control which in turn form intention
and a given behaviour (Fishbein and Ajzen 2010).7 In this sense, a propensity to
evade taxes can be linked with taxpayers’ belief that the government system is
ineffective (Wearing and Headey 1997). High levels of perceived corruption may
weaken taxpayers’ willingness to contribute their fair share of tax (Torgler et al.
2008), and taxpayers may assume that corruption reduces the amount—or benefit—
of government expenditure (Alm, McClelland, and Schulze 1992). As a result,
taxpayers may underreport their tax in an attempt to disengage from reciprocal
relationships with the government (Fjeldstad and Tungodden 2003). In adopting
this framework, we were particularly interested to know whether different forms
of perceived corruption have different impacts on the tax reporting behaviour of
PITs in Indonesia.8

QUALITATIVE RESULTS
The interviews were open-ended, and interviewees were encouraged to provide
their own thoughts and opinions on the questions. The interview data were
subsequently transcribed verbatim. CDC EZ-Text 4.0 software was then used to
conduct a deductive and theoretical thematic analysis procedure described by
Braun and Clarke (2006).
On the basis of a review of the literature on the TPB and perceptions of
corruption, seven themes were established and coded (figure 2): taxpayers’ general
information (TGI); perceptions of corruption (POC); attitude towards behaviour
(ATB); subjective norms (SNO); perceived behavioural control (PBC); intention to
comply (ITC); and tax compliance behaviour (TCB). Each of the themes, except PBC

6. Perception can be defined as ‘a belief or opinion, often held by many people and based
on how things seem’ (https://dictionary.cambridge.org/dictionary/english/perception). As
is true when defining corruption, defining the perception of corruption depends on social
and cultural factors (Melgar, Rosi, and Smith 2010). However, in this context, perception
does not necessarily represent belief in the TPB model. Belief in this model refers to
‘subjective probabilities’ (Fishbein and Ajzen 2010, 221) instead of a sense of certainty about
the existence or truth of something. Smith (2001) offers a fuller discussion on the relation
between perception and belief.
7. This is in stark contrast to what the neoclassical economic model predicts: that most
rational taxpayers would become evaders, since they are unlikely to be caught and penalised
(Ritsatos 2014).
8. The causal relationships between elements in the TPB have been adopted to explain a
variety of psychological and social factors in several studies of tax compliance behaviour.
These include, for instance, the impact of moral obligation (Bobek and Hatfield 2003), the
influence of ethics as a representation of moral reasoning (Trivedi, Shehata, and Mestelman
2005), and the impact of the intention to comply on predicting compliance behaviour
(Langham, Paulsen, and Hartel 2012).
 FIGURE 2 Diagrammatic Interpretation of Interview Data

Level of tax knowledge Low (9) High (2)

1 General information
Opportunity not to comply Employed (1) Self-employed (7)

General corruption Low (-) High (18)

Grand corruption Low (-) High (10)

Petty corruption Low (-) High (9)

2 Perceptions of corruption Grand tax corruption Low (2) High (3)

Petty tax corruption Low (1) High (2)

Impact upon compliance Low (2) High (6)

Source of information Direct (8) Indirect (15)

Experiential attitude Positive (5) Negative (25)

3 Attitude towards behaviour
Instrumental attitude Beneficial (5) Harmful (10)

Injunctive norms Positive (-) Negative

4 Subjective norms
Descriptive norms Positive (-) Negative

5 Perceived behavioural control Low (2) High (25)

6 Intention to comply Positive (1) Negative (41)

Definition of compliance Formal (4) Material (4)

7 Tax compliance behaviour
Perceived level of compliance High (-) Low (27)

Source: Identified from the interview data.

Note: This figure shows a total of 272 assigned codes for 28 categories (pre-existing codes with no zero value) in a double-sided bar. The length of the bar proportionally
indicates the number of identified codes. For instance, as shown at the bottom of the chart, there was no code found indicating that the compliance level in Indonesia
was ‘high’. In contrast, 27 codes were found indicating that the level of tax compliance was ‘low’.
32 Arifin Rosid, Chris Evans, and Binh Tran-Nam

and ITC, was then categorised into relevant sub-themes, resulting in 15 sub-themes.
Each of the sub-themes was coded as a ‘child’ of its original code in the software
and each had a bipolar scale (for example, low−high, negative−positive, harmful−
beneficial). As a result, 34 pre-existing codes were established. The transcribed
data were read carefully to identify meaningful patterns of texts relevant to the
sub-themes and the pre-existing codes. The deductive thematic analysis resulted
in 28 categories (that is, pre-existing codes where meaningful text passages were
identified), which covered 82% of available pre-existing codes and left six pre-
existing codes unused. A total of 272 assigned codes were identified from the
interview data.
Three main findings emerged from the analysis of the interviews. First, Indonesian
PITs were mostly perceived by the participants as being highly unlikely to comply
with their tax obligations, indicated by the extent of negative intention to comply
(figure 2). The participants were of the opinion that complying with the tax law was
associated with ‘bad’ feelings and ‘disadvantages’, which represent the prevalence
of negative attitudes towards compliance. The negative value for the subjective
norms theme indicated that the perceived level of social pressure to comply with
the tax law was absent, which might implicitly encourage PITs to become non-
compliant. Our analysis of the interview data also revealed that PITs were generally
assumed to have considerable control over whether they intended to evade taxes.
Second, the interview data demonstrated an inverse relation between high levels
of perceived corruption and low levels of compliance. However, little was known
from the data about this correlation. The data might suggest that high levels of
perceived corruption are related to both situational and motivational factors. The
situational factor was indicated by the extent of negative psychological evaluations
of a high level of perceived corruption and complying with tax obligations.9
Corruption, either perceived or real, might lead to a negative psychological
evaluation of the completeness and accuracy of information declared in the
annual income tax returns of PITs. The motivational factor, meanwhile, could
be related to a lack of perceived social pressure among PITs to comply with the
tax law, particularly in providing complete and accurate information in annual
returns. Negative values for both injunctive norms (what others believe is right)
and descriptive norms (what others do) among the participants indicated that the
beliefs of certain groups of referents discourage PITs from complying with the
tax law and that beliefs that PITs’ important referents are not fully reporting their
income were evident.
Finally, as suggested by many scholars (Alm, McClelland, and Schulze 1992;
Bird, Martinez-Vazquez, and Torgler 2008; Fjeldstad and Tungodden 2003; Gangl
et al. 2015; Litina and Palivos 2016), it was possible to propose a hypothetical
model of relations between high levels of perceived corruption and low levels of
compliance. In this sense, this hypothetical model conceptualises that perceived

9. For instance, ‘Why should I pay tax if it is just financing corruption?’ was a frequently
expressed view from participants―this sentiment was expressed 18 times during the
interviews. Also, as several participants described, to compensate for a ‘non-deductible
payment’ (that is, a bribe) for tax calculation purposes, some taxpayers might have no choice
but to underreport their income. These taxpayers may otherwise have to pay double taxes,
weakening their financial circumstances.
Tax Non-compliance and Perceptions of Corruption 33

levels of corruption influence both attitudes and subjective norms of Indonesian

PITs. Then, the affected attitudes and subjective norms, combined with a high
level of perceived behavioural control for non-compliance, lead to the formation
of a negative intention to comply with the tax law. Subsequently, this negative
behavioural intention can be used to predict the low levels of compliance. In other
words, the relation between high levels of perceived corruption and low levels of
tax compliance is indirect.

SURVEY RESULTS
Demographic Data
The survey questionnaire comprised 72 questions. On the basis of a seven-point
rating scale, respondents were asked to indicate their beliefs, values, attitudes,
intentions, and behaviour regarding 10 constructs with 56 reflective indicators.
Respondents were then asked 11 questions about their demographic background.
The constructs and their indicators were as follows: (a) perceptions of general
corruption (PGC) = 5; (b) perceptions of grand corruption (GCO) = 5; (c) perceptions
of petty corruption (PCO) = 5; (d) perceptions of grand tax corruption (GTC) = 5; (e)
perceptions of petty tax corruption (PTC) = 5; (f) attitude towards tax underreporting
(ATB) = 9; (g) subjective norm towards tax underreporting (SNO) = 8; (h) perceived
behavioural control over tax underreporting (PBC) = 8; (i) intention to report actual
income (ITC) = 4; and (j) level of reported income (TCB) = 2. (Appendix table A1
contains details on all survey items.)
Respondents consisted of two groups: 196 self-employed PITs and 201 employed
PITs. The surveyed self-employed PITs were categorised as small and medium
enterprises with annual sales turnover of less than Rp 4.8 billion (roughly $360,000).
There was no income threshold applied for the employed PITs.
We looked at several demographic characteristics of respondents in order to
obtain insights that might guide our discussion of policy implications (table 1). By
age, respondents were concentrated in two year ranges: 24–34 (35%) and 35–44
(33%). A large proportion of respondents were male (64%). Most respondents were
married (90%), and the share of married self-employed PITs (93%) was higher than
that of employed PITs (86%).
Because the survey commenced in mid-March 2015 and ended in late June
2015, we performed a timing-bias test—by using a wave analysis, as suggested
by Armstrong and Overton (1977)—to check whether those who completed the
questionnaire in June 2015 were not significantly different, demographically, from
those who completed it in March 2015. We then carried out basic comparisons to
ensure the representativeness of the sample at four levels: national level (331 tax
offices), regional level (39 tax offices), tax office level (12 tax offices), and surveyed-
respondent level (201 employed PITs).10 We found no timing-response bias in the
survey data and concluded that respondents, particularly employed PITs, were

10. Considering that the survey collected data from similar portions of self-employed and
employed PITs in each of the surveyed tax offices, the variability of these portions should
not be evident. For this reason, the variability of types of annual return lodged by employed
PITs—who may use either 1770 S or 1770 SS tax returns—was used.
 TABLE 1 Respondents’ Demographic Characteristics (% of sample)

Self-employed Employed
taxpayers taxpayers Total
Characteristic (n = 196) (n = 201) (n = 397)

Age (years) <25 6.6 6.0 6.3

25–34 36.7 33.3 35.0
35–44 35.2 31.3 33.2
45–54 17.3 20.9 19.1
55–64 2.6 6.5 4.5
>65 1.5 2.0 1.8
Sex Male 68.4 58.7 63.5
Female 31.6 41.3 36.5
Marital status Single 6.1 9.5 7.8
Married 93.4 86.1 89.7
Separated/divorced 0.5 4.5 2.5
Race Javanese 67.3 65.7 66.5
Balinese 18.1 17.8 17.9
Others 6.6 4.5 5.5
Sumatranese 2.6 6.5 4.5
Sundanese 2.6 3.0 2.8
Sulawesinese 1.5 2.0 1.8
Kalimantanese 1.5 0.5 1.0
Education level High school 20.9 24.9 22.9
Diploma 11.2 9.5 10.3
Undergraduate 60.2 58.7 59.4
Postgraduate 7.7 7.0 7.3
Annual income <60 62.8 76.1 69.5
(Rp million) 60–99 20.9 13.4 17.1
100–149 8.2 5.5 6.8
150–199 5.6 3.5 4.5
>200 2.6 1.5 2.0
Tax handling Self-prepared 62.2 40.8 51.4
Tax office’s assistance 17.9 45.8 32.0
Friends’ assistance 16.3 12.4 14.4
Tax agent 3.6 1.0 2.3
Previous interaction No contact 63.8 83.6 73.8
with DGT Contact, no audit 32.1 16.4 24.2
Audit, no penalty 1.5 — 0.8
Audit with penalty 2.6 — 1.3

Source: Authors’ calculations using survey data.

Tax Non-compliance and Perceptions of Corruption 35

reasonably representative of the sample frame in terms of annual tax return type
and income group.11 In addition, our survey data provided in table 1 suggest that
the share of male taxpayers is 63.5%. This is comparable with that in Sakernas, the
Indonesian Labour Force Survey, which during 2007–14 recorded the share of male
taxpayers as 62.8% (ILO 2015).
Nine of the twelve surveyed tax offices were in Java, so the study sample was
dominated by Javanese respondents (almost 67%). Balinese respondents were in
the second position, with nearly 18%. Both types of taxpayer shared almost similar
portions of these two-dominant ethnic groups. Other ethnic groups included
Sumatranese, Sundanese, Sulawesinese, Kalimantanese, and ‘Others’ (such as
Tionghoa, or Chinese-Indonesians).
The study sample was dominated by respondents whose annual income was less
than Rp 60 million ($4,500) (70%), followed by those who earned Rp 60 million–Rp
99 million (17%). In the third and fourth position were those who had income group
Rp 100 million–Rp 149 million (7%) and Rp 150 million–Rp 199 million (5%). The
smallest income group was those who had more than Rp 200 million in annual
income (2% of the total respondents).
The vast majority of respondents had university degrees, with the largest such
group having undergraduate qualifications (60%), followed by those with diploma-
level (10%) and postgraduate qualifications (7%). Respondents with high-school
qualifications composed 23% of the total. More than half of respondents were ‘self-
preparers’ in completing their annual tax returns (51%), followed by those who
had received assistance from tax offices (32%). In contrast, only 2% of respondents
had used tax agents to deal in full with their annual income tax return. While most
self-employed PITs were self-preparers (62%), the largest share of employed PITs
had received assistance from tax offices (46%).
With respect to previous interactions with the DGT, by far the largest proportion
of respondents had never been contacted by tax offices (74%), followed by those
who had been contacted but never audited (24%). Those who had been audited
and penalised by tax offices constituted about the same small proportion as those
who had been audited but never been penalised (1%). Those who had been audited
were all self-employed PITs, and the proportion of those who had been penalised
(3%) was slightly higher than those who had not (2%).
In addition, we found that respondents’ primary sources of information about
corruption were press media (65%), television (63%), and the internet (40%).

Perceptions, Attitudes, Norms, Perceived Control, and Intentions

Table 2 summarises key survey outcomes relating to perceived levels of corruption
and to attitudes, norms, perceived control, and intentions, on the basis of responses
to a seven-point rating scale. The variable ‘perception of corruption’ measures
different forms of perceived corruption, using the conventional definition and
classification of corruption derived from Doig and Theobald (1999), Kaufmann,

11. The representativeness test of the sample frame should ideally be performed on various
demographic variables, such as age, gender, level of education, and race. Unfortunately,
owing to the nature and availability of data provided by the DGT, only these two variables
could be used for the test.
 TABLE 2  Survey Responses (based on a seven-point rating scale of interval measurement)

Self-employed Employed All respondents

(n = 196) (n = 201) (n = 397)

Construct Description Mean SD Mean SD Mean SD

Perception of general corruption Taxpayer’s perceived level of abuse of entrusted power by 6.21 0.84 6.10 0.93 6.15 0.89
(PGC) public official for private gain
Perception of grand corruption Taxpayer’s perceived level of misuse of public power by 6.49 0.64 6.48 0.70 6.48 0.67
(GCO) high-level public official for private gain which often
involves large sums of money
Perception of petty corruption Taxpayer’s perceived level of extortion of small payments by 6.41 0.66 6.27 0.75 6.34 0.71
(PCO) low-level public officials in daily interactions with the public
(‘grease money’)
Perception of grand tax corruption Taxpayer’s perceived level of misuse of public power by 5.57 1.14 5.49 1.07 5.52 1.11
(GTC) high-level tax official for personal pecuniary gains, which
in certain tax cases involves large illegal payments
Perception of petty tax corruption Taxpayer’s perceived level of misuse of public power by 5.01 1.03 5.02 1.08 5.01 1.06
(PTC) low-level tax official for personal pecuniary gains or the
extortion of small payments by operational staff in daily
interactions with taxpayers (‘grease money’)
Attitude towards tax Taxpayer’s psychological evaluation (i.e., some degree of 4.51 1.25 4.30 1.28 4.40 1.27
underreporting (ATB) favourableness) towards underreporting income in annual
tax return
Subjective norms & tax Taxpayer’s perceived social pressure with respect to report 3.89 1.22 3.72 1.30 3.80 1.26
underreporting (SNO) incomplete income in the annual tax return
Perceived behavioural control & Taxpayer’s perceived ability and opportunity to report 4.56 1.23 3.12 1.23 3.83 1.42
tax underreporting (PBC) incomplete income in the annual tax return
Intention to report actual income Taxpayer’s willingness to fully report actual income in annual 3.57 1.68 4.02 1.69 3.80 1.70
(ITC) tax return

Source: Calculated from the survey data. Note: The mean and standard deviation (SD) for each construct are calculated from the corresponding reflective indicators. The
number of indicators are as follows: PGC = 5; GCO = 5; PCO = 5; GTC = 5; PTC = 5; ATB = 9; SNO = 8; PBC = 8; ITC = 4; and TCB = 2. Thus, for instance, the PGC mean score
of 6.15 for all respondents was calculated from 1,985 observations (5 × 397).
Tax Non-compliance and Perceptions of Corruption 37

Kraay, and Mastruzzi (2007), Mishler and Rose (2008), and the UNDP (2008).
Following the work of Melgar, Rossi, and Smith (2010) and Torgler et al. (2008), we
then developed survey questions relating to perceptions of corruption. The elements
of TPB in the survey questionnaire (that is, attitudes towards underreporting,
subjective norms towards underreporting, perceived behavioural control over
underreporting, intention to report actual income, and income reporting behaviour)
were modified and adapted from Fishbein and Ajzen’s (2010) constructs on the
basis of a specific target, action, context, and timeframe—that is, on the reporting
of actual income on an annual income tax return for the fiscal year 2014.

Perceived Levels of Corruption

Respondents were generally of the opinion that the levels of corruption in Indonesia
were very high. Specifically, the highest perceived level of corruption was that of
grand corruption, with the mean score (M) for its indicators being 6.49 (out of 7.00)
for self-employed individuals and 6.48 for all respondents. In contrast, perceptions
of petty tax corruption had the lowest mean score, with an indicator of 5.01.

Attitudes towards Tax Underreporting

In support of the qualitative results, fully reporting actual income in annual tax
returns seems to be associated by respondents with a lack of positive experiential
attitudes and unpleasant instrumental attitudes. The survey data show that
respondents tend to be favourably disposed to underreporting income (M = 4.40).

Subjective Norm towards Tax Underreporting

Respondents generally perceived that their significant referents (persons likely
to influence their intentions and behaviours) think they should not fully report
their actual income. As a result, they appear to have somewhat positive subjective
norms towards underreporting their actual income. The survey data suggest that
the perceived level of social pressures not to report actual income was slightly
below the scale mean (M = 3.80).

Perceived Behavioural Control over Tax Underreporting

Respondents perceived that they have the ability and the opportunity to underreport
income. This is particularly the case with the self-employed respondents. The data
reveal that self-employed PITs appear to have a considerable control over whether
they want to intentionally engage in tax evasion (M = 4.56). In contrast, employed
PITs scored much lower (M = 3.12) than self-employed PITs. This significant
divergence may reflect the different nature of opportunity for non-compliance
between employed and self-employed PITs, as employed PITs typically have their
income subject to withholding tax.

Intention to Report Actual Income

The data suggest that respondents are generally unwilling to report their actual
income (M = 3.80). Self-employed PITs are less willing to report their actual income
(M = 3.57) than employed PITs (M = 4.02).

Levels of (Under-)Reported Income

The survey data indicate that only 18% (36 of 196) of self-employed PITs agreed or
strongly agreed that they had reported their actual income in the 2014 annual tax
38 Arifin Rosid, Chris Evans, and Binh Tran-Nam

TABLE 3 The Extent to Which Respondents Self-Reported

Their (Non-)Compliance Behaviour (TCB)

Question (a): I have fully reported my actual income in my annual

tax return for fiscal year 2014 / I have fully reported my income
other than salary, wage, or other tax-withheld income on my annual
tax return for 2014

Type of PIT SD D sD N sA A SA Total

Self-employed 4 15 42 54 45 27 9 196
Employed 5 18 13 27 13 84 41 201
Total 9 33 55 81 58 111 50 397

Question (b): As far as I can remember, the amount of income I

have reported in my annual tax return was roughly ... of my actual
incomea

No
Type of PIT income 0% 16% 33% 50% 67% 83% 100% Total

Self-employed 7 2 7 34 42 53 37 14 196
Employed 167 34 NA NA NA NA NA NA 201
Total 174 36 7 34 42 53 37 14 397

Note: PIT = personal income taxpayer; SD = strongly disagree; D = disagree; sD = somewhat disagree;
N = neutral; sA = somewhat agree; A = agree; SA = strongly agree; NA = not applicable.
a
For employed PITs this refers to any income other than salary, wage, or other tax-withheld income.

return (table 3). In contrast, 62% (125 of 201) of employed PITs agreed or strongly
agreed that they had reported their income other than salary, wages, or other tax-
withheld income in their return. A large portion of the surveyed employed PITs
had no additional income other than from employment (83%), implying that the
rest (17%, or 34 employed PITs) received additional income. Further, roughly 4%
of employed PITs (seven respondents) have reported that they did not earn income
in 2014 (table 3).
The data also indicate that 44% of self-employed PITs underreported between 50%
and 100% of their actual income in their return. Surprisingly, not one respondent
from the group of employed PITs who received additional income other than from
employment (34 PITs) reported their additional income.

QUANTITATIVE APPROACH AND RESULTS

We used the survey data to examine the hypothetical relationships among
variables under study derived during the qualitative phase, applying three basic
models to our five types of perceived corruption. In doing so, we took a structural
equation modelling (SEM) approach, using IBM SPSS Statistics and IBM SPSS
Tax Non-compliance and Perceptions of Corruption 39

Amos software. Commonly used to assess dependence relationships among latent

variables simultaneously (Hair et al. 2010), SEM is ‘a technique to specify, estimate,
and evaluate models of linear relationships among a set of observed variables in
terms of a generally smaller number of unobserved variables’ (Shah and Goldstein
2006, 149).
In analysing the survey data, we further specified two categories of model—
full and partial—because self-employed PITs who did not earn income (seven
respondents) and employed PITs who did not receive additional income other
than from employment (167 respondents) could not be included in the full-
model analysis.12 Accordingly, we used the whole sample (397 respondents) in
our analysis of partial models but only a partial sample (223 respondents) in our
analysis of full models.
We initially performed discriminant validity tests to determine whether a
construct differed from other latent constructs, examining correlation coefficients
among latent variables and the square root of average variance extracted. The results
(appendix table A2) suggest that the correlation coefficients among constructs were
all lower than the square root of average variance extracted. We concluded that all
constructs under study have acceptable discriminant validity.
As a result, we examined six models. Because the initial models did not fit the data,
we added several covariance of error terms. The outcomes of our re-estimation are
summarised in appendix table A3. As the table shows, perceptions of corruption
have an indirect effect on the intention to report actual income (ITC) and the level
of reported income (TCB). The details of these indirect effects are presented in
appendix table A4.
With reference to the fit model indices, we found that model B (both full and
partial models) appeared to best fit the data. Partially supported by the outcomes
of model C, model B revealed a tendency for perceptions of grand corruption
to influence taxpayers’ attitudes towards tax underreporting more strongly than
their subjective norms towards tax underreporting. That is, perceptions of grand
corruption and grand tax corruption made taxpayers’ attitudes towards tax
underreporting more negative, while perceptions of petty corruption tended to
affect taxpayers’ subjective norms related to tax underreporting.
The quantitative findings demonstrate the underlying mechanisms by which the
different forms of perceived corruption negatively influence the way individual
taxpayers behave, by way of three principal findings.13 First, perceptions of
corruption appear to have stronger influences on taxpayers’ attitudes towards tax
underreporting than their subjective norms towards tax underreporting. In this
regard, perceptions of grand corruption and grand tax corruption were found to
be influential in affecting taxpayers’ attitudes towards tax underreporting (with
path coefficients ranging from 0.317 to 0.363 and from 0.225 to 0.290, respectively),

12. Full models include the ‘level of reported income’ (TCB) variable. Partial models exclude
this variable.
13. These three findings draw heavily from Rosid, Evans, and Tran-Nam (2016). Mainly
due to space considerations, the detailed process of empirical investigation is not discussed
further in this article.
40 Arifin Rosid, Chris Evans, and Binh Tran-Nam

TABLE 4  Statistically Significant Differences among

Perceptions of Corruption by Demographic Variables

Predictors (demographic variables)

Income Employ-
Outcomes (perceived level of) Age Gender Education level ment status

General corruption (PGC) — — Ya** — —

Grand corruption (GCO) — — — — —
Petty corruption (PCO) — — — — Yc*
Grand tax corruption (GTC) — — — — —
Petty tax corruption (PTC) — — Yb* — —

Note: — = statistically significant difference not evident; * = significant at p < 0.05; ** = significant at p < 0.10.
Ya = statistically significant difference evident, with the higher the level of education, the lower the score of
PGC (Wilk’s Ʌ = 0.942, F[15, 1074] = 1.57, p = 0.07, partial ƞ2 = 0.02). Yb = statistically significant difference
evident, with the higher the level of education, the lower the score of PTC (Wilk’s Ʌ = 0.929, F[15, 1074] =
1.94, p = 0.02, partial ƞ2 = 0.02). Yc = statistically significant difference evident, with self-employed PITs
scored higher than employed PITs (Wilk’s Ʌ = 0.966, F[5, 391] = 1.94, p = 0.02, partial ƞ2 = 0.03).

while perceptions of petty corruption tend to affect taxpayers’ subjective norms

towards tax underreporting (with path coefficients ranging from 0.326 to 0.331).
Second, taxpayers’ attitudes towards tax underreporting and their subjective
norms towards tax underreporting, coupled with the perceived behavioural control
over tax underreporting, appear to undermine taxpayers’ intention to report
actual income. Attitudes towards tax underreporting seem to have the strongest
direct effect on taxpayers’ intention to report actual income (with path coefficients
ranging from −0.400 to −0.445)—followed by perceived behavioural control over
tax underreporting, with path coefficients ranging from −0.165 to −0.244. Subjective
norms towards tax underreporting appear to have the weakest direct effect on
taxpayers’ intention to report actual income, with path coefficients ranging from
−0.122 to −0.175. The findings also suggest that the intention to report actual income
is a good predictor of tax reporting behaviour, with path coefficients greater than
0.550.
Third, perceptions of corruption have an impact on taxpayers intentionally
underreporting their income tax. When perceived levels of grand corruption,
grand tax corruption, and general corruption increase by one standard deviation,
intentional tax underreporting increases by 0.073, 0.071, and 0.034 standard
deviations, respectively.
Our multivariate analysis of variance (MANOVA) showed that perceived levels
of corruption, as outcomes, were not related to age groups, gender, or income
levels but were related to education levels and employment status (table 4). In
particular, perceptions of general corruption were related to education levels,
with respondents who have higher education levels expressing lower levels of
perceived corruption (M for high school = 6.27, diploma = 6.21, undergraduate
= 6.14, and postgraduate = 5.83). Similarly, perceptions of petty tax corruption
were found to be only related to education levels, with respondents who have
higher education levels expressing lower levels of perceived corruption (M for
Tax Non-compliance and Perceptions of Corruption 41

TABLE 5  Statistically Significant Differences among the

TPB Constructs by Demographic Variables

Predictors (demographic variables)

Income Employ-
Outcomes (TPB constructs) Age Gender Education level ment status

Attitude towards tax

underreporting (ATB) — — — — —
Subjective norms towards tax
underreporting (SNO) — Ya* — — —
Perceived behavioural control
over tax underreporting
(PBC) — — — Yc* Yd*
Intention to report actual
income (ITC) — Yb* — — Ye*
Level of reported income
(TCB) — — — — Yd*

Note: — = statistically significant difference not evident; * = significant at p < 0.05. Ya = statistically
significant difference evident, with male PITs scored higher than female PITs (Wilk’s Ʌ = 0.959, F[8, 388]
= 2.06, p = 0.04, partial ƞ2 = 0.27). Yb = statistically significant difference evident, with male PITs scored
lower than female PITs (Wilk’s Ʌ = 0.972, F[4, 392] = 2.84, p = 0.02, partial ƞ2 = 0.03). Yc = statistically
significant difference evident, with the higher the level of income, the higher the score of PBC (Wilk’s
Ʌ = 0.887, F[32, 1421] = 1.47, p = 0.04, partial ƞ2 = 0.03). Yd = statistically significant difference evident,
with self-employed PITs scored higher than employed PITs (for perceived behavioural control over tax
underreporting: Wilk’s Ʌ = 0.584, F[8, 388] = 34.49, p = 0.00, partial ƞ2 = 0.42; for level of reported income:
Wilk’s Ʌ = 0.473, F[2, 220] = 122.49, p = 0.00, partial ƞ2 = 0.27). Ye = statistically significant difference
evident, with self-employed PITs scored lower than employed PITs (Wilk’s Ʌ = 0.958, F[4, 392] = 4.278,
p = 0.00, partial ƞ2 = 0.04).

high school = 5.18, diploma = 5.16, undergraduate = 4.98, and postgraduate = 4.48).
Furthermore, a statistically significant difference between employment status was
evident when considered jointly on the dependent variables of petty corruption,
with self-employed PITs (M = 6.41) scoring higher than employed PITs (M = 6.26).
Considering the small effect sizes, however, these might be negligible. We also
found that the TPB constructs under consideration, as outcomes, were not related
to age groups and education levels. Instead, they were associated with gender,
income level, and employment status (table 5).
A statistically significant difference between males and females was evident
when we considered gender jointly with the dependent variables of subjective
norms towards tax underreporting, with males (M = 3.88) scoring higher than
females (M = 3.65). Similarly, we found a statistically significant result when we
considered gender jointly on the dependent variables of the intention to report
actual income, with males (M = 3.70) scoring lower than females (M = 3.96). As
with education levels and employment status, however, the effect of gender on the
intention to report actual income might be negligible.
The extent of perceived behavioural control over tax underreporting seems to
be related to income levels. We found a statistically significant difference among
income levels when considered jointly on the dependent variables of PBC, with
42 Arifin Rosid, Chris Evans, and Binh Tran-Nam

respondents who have a higher level of income tending to have a higher perceived
level of behavioural control for tax underreporting (M for those with incomes under
Rp 60 million = 3.78, between Rp 60 million and Rp 99 million = 3.94, between Rp
100 million and Rp 149 million = 4.03, between Rp 150 million and Rp 199 million =
3.61, and over Rp 200 million = 4.26). Again, this might be negligible.
In contrast, the extent of perceived behavioural control over tax underreporting
appears to be substantially related to employment status. A statistically significant
difference between self-employed and employed PITs was evident when considered
jointly on the dependent variables of PBC, with self-employed PITs (M = 4.56)
scoring higher than employed PITs (M = 3.12).
We also found a statistically significant difference between self-employed and
employed PITs, albeit with a small effect size, when considered jointly on the
dependent variables of the intention to report actual income, with self-employed
PITs (M = 3.57) scoring lower than employed PITs (M = 4.02). Finally, a statistically
significant effect of employment status was evident when considered jointly on
the dependent variables of the level of reported income, with self-employed PITs
(M = 4.37) scoring higher than employed PITs (M = 1.62).14
In summary, we found that high levels of perceived corruption were pervasive,
regardless of taxpayers’ demographic background. While self-employed PITs
appeared to have greater perceived behavioural control over tax underreporting,
employed PITs had a lower level of reported income than self-employed PITs so
far as extra income was concerned.

POLICY IMPLICATIONS
Compliance Risk Management
The DGT has identified several issues relating to its low tax ratio and is particularly
concerned with the ongoing low level of tax compliance (DGT 2015b). It aims to
increase the tax ratio from 12.2% in 2015 to 15.0% in 2019 (DGT 2015b). In relation
to this aim, the DGT has recently outlined several strategic initiatives to improve
tax compliance in Indonesia that are scheduled to take place between 2015 and
2019. Among these is an initiative to implement compliance risk management
(CRM).
The DGT defines CRM as ‘a systematic process to select appropriate instruments
to improve taxpayer compliance and to deter non-compliance effectively, based on
knowledge of taxpayer behaviour and the DGT’s resource availability’ (DGT 2015b,
36). The DGT acknowledges the absence of a comprehensive risk-based approach
in dealing with compliance issues, and as a result it is unable to allocate, in an
efficient and effective way, its resources and appropriate treatments for various
types of taxpayer on the basis of their risk (DGT 2015b). It further notes that several

14. This result seems counterintuitive. Nevertheless, recent data on the Indonesian
government’s tax amnesty program might partly reflect this phenomenon. The internal
data provided by the DGT on 12 October 2016 suggest that, surprisingly, more than 53%
of the total tax payments collected from phase one of the 2016 tax amnesty program
(July−September 2016) was paid by employed PITs. As our findings suggest, one possible
explanation for this is that the employed PITs tend to conceal much of their additional
income other than income from employment.
Tax Non-compliance and Perceptions of Corruption 43

decision-making processes on crucial issues are based on subjective judgements

and that its understanding of taxpayer compliance behaviour is inadequate (DGT
2015b).15
In administering and collecting tax revenue a tax authority is often shaped
by the specific organisational, socio-economic, and political contexts in which it
operates (OECD 2004). Among many external factors that affect tax administration
in developing countries is the extent of corruption (Bird 2015). In this sphere, our
empirical findings support the notion that high levels of perceived corruption are
an important part of the external operating context.
Once the external context is established, the CRM process then involves at least
seven further stages. The four phases of CRM that are relevant to our empirical
results include (a) risk identification, (b) risk assessment and prioritisation, (c) the
analysis of compliance behaviour, and (d) the formulation of treatment strategies.
The three other steps (planning and implementation of strategies, compliance
outcome evaluation, and performance monitoring) are beyond the scope of this
article.

Implementing Compliance Risk Management

Tax administrations need to better understand taxpayer behaviour if they are to
develop more effective and stronger compliance risk treatments (Walsh 2012).
Given the DGT’s future implementation of the CRM and the high level of perceived
corruption as a major feature of its operating context, several of our findings are
significant.

Risk Identification
The first step in the implementation of CRM is risk identification. At the strategic
level, this phase is needed to assess specific types of non-compliance behaviour that
seem to have substantial implications for tax revenue if left untreated. In this regard,
our study suggests that the intentional underreporting of income, particularly by
self-employed PITs, is a compliance risk that the DGT must confront.
For an appropriate treatment, this risk needs to be identified at a particular level.
Two major indicators can be used to specify risk: macro-level and public opinion
indicators (OECD 2004). While macro-level indicators generally illustrate a relation
between an aspect of compliance and an external statistic or benchmark for tracking
a global trend, public opinion indicators link the incidence of non-compliance to
public awareness, perception, attitudes, and motivation, on the basis of public
opinion surveys or research results (OECD 2004).
In pursuit of a narrower definition, we provide evidence of the identified risk from
the perspective of the individual taxpayers, emphasising the socio-psychological
perspective. In this sense, our results suggest that an underreporting risk is evident,
with only 18% of surveyed self-employed PITs agreeing or strongly agreeing that
they had reported their actual income, and 44% of them suggesting that they had
underreported between 50% and 100% of their actual income. Our results also

15. In this sense, the DGT describes a situation where many of its tax officers are constantly
monitoring the same taxpayers without prioritising compliance, and terms this business-
as-usual phenomenon as berburu di kebun binatang, or ‘hunting at the zoo’ (DGT 2015b, 36).
44 Arifin Rosid, Chris Evans, and Binh Tran-Nam

indicate that the surveyed employed PITs tend to conceal much of any income
other than from their employment. According to Sakernas data, 13% of employed
PITs had additional jobs during 2009–13.16

Risk Assessment and Prioritisation

The second step involves risk assessment and prioritisation, in order to separate
major risks from minor ones on the basis of their consequences and likelihood.17
The OECD (2004) recommends that this step be based on objective evidence. In this
sense, our analysis provides two relevant objective pieces of evidence.
First, in terms of consequences, and as noted above, we found that only 7% of
self-employed PITs had reported all of their actual income, while 44% of them
had underreported between 50% and 100% of the actual income. In terms of level
of risk, therefore, we may reasonably consider the consequences to be either
‘high’ or ‘very high’. Second, in terms of likelihood, we use our path coefficient of
behavioural intention as an indicator of probability. Behavioural intention can be
considered the closest and most accurate proxy for performed behaviour (Fishbein
and Ajzen 2010). In our analysis, we find that the direct effect of intention to report
actual income on the level of reported income is approximately 0.56. In this sense,
we have established that Indonesian PITs do not appear to have a high level of
intention to report their actual income, with the indicators’ mean score of 3.81 out
of 7.00. Worryingly, self-employed PITs tend to have a lower level of intention
(3.56) than employed PITs (4.02). Thus, it seems safe to say that the likelihood of
underreporting risk is almost certain.18
On the basis of these findings, the risk rating relative to the consequences and the
likelihood of intentional underreporting by self-employed PITs can be prioritised as
either high or severe. This is an arbitrary risk rating when the consequences of the
risk range from high to very high and the likelihood of the risk is almost certain. It is
critical to understand that tax authorities will not be able to address all risks (OECD
2004). For this reason, because of a balanced approach (see below), some risks being
addressed may not deal with instant exposure to high revenue (OECD 2010).

Analysis of Compliance Behaviour

The third step involves compliance behaviour analysis. One benefit of implementing
compliance risk management is the ability to focus on the underlying drivers—
rather than the symptoms—of non-compliance and promote different methods
of mitigating various tax compliance risks, rather than adopting a ‘one size fits
all’ approach (OECD 2004). Understanding the factors that influence specific

16. Data from http://www.ilo.org/surveydata/index.php/catalog/history, accessed 5 May

2017. By way of comparison, our findings suggest that 17% of employed PITs earn income
other than from employment.
17. Consequence is evaluated in the terms of its impact on organisational objectives. It can
be measured qualitatively or quantitatively or by both methods (OECD 2004). Likelihood
is assessed in terms of its probability of occurring. It can also be measured qualitatively or
quantitatively or by both methods (OECD 2004). In other words, it is somewhat arbitrary.
18. A subjective definition of ‘almost certain’ would be ‘expected to occur in most
circumstances’, while an objective definition would be ‘likely to occur this year or at frequent
intervals’ (OECD 2004, 28).
Tax Non-compliance and Perceptions of Corruption 45

compliance behaviour is essential to selecting appropriate compliance strategies.

That is, on the one hand, by allowing the tax authority to identify the ‘causes’ of
non-compliance, a longer-term sustainable compliance outcome can be achieved.
On the other hand, failing to appropriately treat the causes of non-compliance not
only exacerbates the underlying systemic problem but may also cause community
outrage. Unfortunately, there is no easy way to understand the determinants of
either compliance or non-compliance (OECD 2004).
In response, using the OECD’s (2004) process for discovering the drivers of
particular behaviour,19 our findings arguably provide three important insights.
First, intentional underreporting of income is occurring among Indonesian PITs.
Second, in terms of taxpayer characteristics, our findings suggest that while self-
employed and employed PITs seem to have similar values in relation to attitudes
towards underreporting behaviour and perceived subjective norms (table 5), they
have a significant difference in terms of perceived behavioural control over tax
underreporting, with self-employed PITs scoring much higher (M = 4.56) than
employed PITs (M = 3.12).
This notable gap represents, in all probability, the different nature of opportunity for
non-compliance between employed and self-employed PITs. The Indonesian PITs not
only seem to have a relatively high score for their attitudes towards underreporting
behaviour (M = 4.40) but also appear to have lack of strong social pressures or
normative norms in terms of reporting full actual income (M = 3.80) (table 2).
Third, and most importantly, our study provides an objective answer for the
critical question ‘Why are they doing it?’. We find that in the Indonesian context high
levels of perceived corruption, among other factors, appear to influence taxpayers
to intentionally underreport their income. Our study empirically demonstrates
that perceptions of grand corruption, grand tax corruption, and general corruption
appear to be responsible for intentional underreporting, with standardised total
negative effects of 0.073, 0.071, and 0.034, respectively (appendix table A2).

Determining Appropriate Treatment Strategies

It has been recognised that while influencing taxpayer behaviour is not yet a science,
it is also not simply guesswork (OECD 2010). Treatment strategies can be put in
place only after the risk has been appropriately addressed and the possible causes
of non-compliant behaviour have been clearly established (OECD 2004). It has
also been argued that it is impossible to apply a single specific tax administration
strategy (OECD 2010). This is because tax authorities not only have to deal with
different motivational postures of taxpayers but also generally have finite resources
(Osofsky 2014). In this sense, with respect to the applications of responsive regulatory
framework,20 our findings suggest five practical implications for treatment strategies.

19. Three basic questions need to be addressed in understanding taxpayer compliance

behaviour (OECD 2004, 42): What is occurring (for example, the underreporting of income)?
Who is doing it (what are their characteristics)? And why are they doing it (for example,
because of perceived inequity)?
20. The fundamental principle of this approach is that policymakers should be responsive to
the culture, conduct, and context of those they seek to regulate. Thus, it requires a diversity
of regulatory strategies to be context appropriate (Wood et al. 2010).
46 Arifin Rosid, Chris Evans, and Binh Tran-Nam

First, a norms-based intervention using extensive publicity campaigns or

customised letters sent to targeted taxpayers may represent a sensible strategy for
the DGT (Onu and Oats 2015). The goal of an intervention is generally to alleviate
an identified social problem in a given segment of the taxpaying population. In
this sense, the most common intervention strategy is to use a persuasive message
to disseminate relevant information to the target audience. It can be used to reach
a wide population at relatively low cost (Fishbein and Ajzen 2010). For instance,
using a randomised controlled trial in promoting tax compliance in Guatemala,
Kettle et al. (2015) found that increasing the moral cost by referring to a social norms
condition had a positive impact on tax payment. Similar to a study by Hallsworth et
al. (2017) which concluded that messages referring to fiscal exchange significantly
increased payment rates, a recent study in Indonesia by Bursztyn et al. (2015),
which found that moral messages increased credit card repayments, indicates that
norm-based intervention could also improve tax reporting behaviour.
This strategy may seem easy and inexpensive, but it should be implemented
cautiously. In our findings we provide evidence that most taxpayers appear to
perceive that their peers do not pay their fair share of taxes. In particular, with
regard to reporting actual income, we found that both low descriptive (what others
do) and low injunctive (what others believe is right) norms co-occur.21 Thus, to
improve the level of compliance, it may be ineffective to target low-compliant
taxpayers using normative descriptive norms (such as stating ‘most taxpayers fully
report their income’ or ‘few people evade tax’ in letters sent to taxpayers) because
the messages would feel counterintuitive or in contrast to their own beliefs. On the
other hand, while injunctive norms appeared to be less effective than descriptive
norms (Hallsworth et al. 2017), it would also seem to be problematic to campaign
using the actual descriptive norm arising from our findings because it may be seen
by the targeted taxpayers as a ‘gold standard’. That is, there is a risk that campaigns
that communicate the actual descriptive norm would encourage those above the
average to adjust their behaviour downwards and therefore would reinforce the
existing low level of compliance. In addition, moral messages appear to be less
effective for promoting tax compliance by rich people than by poor people (Duch
and Solaz 2016). In response, we propose that it may be appropriate to disseminate
the normative injunctive message (such as ‘underreporting tax is wrong’ or ‘paying
tax is the right thing to do’) only among the targeted low-compliant taxpayers.
The second implication relates to the implementation of a focused enforcement
approach. In a non-compliant group, taxpayer behaviour can be influenced
by increasing the frequency of audit (OECD 2004, 2010). Although it often has
significant costs, enforced compliance is required where non-compliance exists
(OECD 2014a). Because the DGT’s enforcement resources are very limited (DGT
2015a), however, the audit rate is extremely low. For example, in 2014, the target
audit coverage ratio for individual taxpayers was only 0.1%.22 This highlights two

21. In this sense, while descriptive norms tend to guide people in uncertain situations,
injunctive norms are likely to help people avoid social disapproval (Onu and Oats 2015).
22. See the DGT’s official letter SE-15/PJ/2014 on 21 March 2014 regarding Audit Plan and
Strategies for 2014. The audit coverage ratio for PITs represents the number of audited PITs
against the number of PITs’ annual tax returns received by the DGT. The actual ratio is
Tax Non-compliance and Perceptions of Corruption 47

points. First, without specific treatment, taxpayers are unlikely to believe that
their risk of getting audited has increased. Second, enforcement resources need
to be used to deal only with targeted taxpayers. This is so because the response of
compliant taxpayers to audits is ambiguous and might have undesired outcomes
(Beer et al. 2016).
To make an enforcement approach among low-compliant taxpayers work, as
suggested by Osofsky (2014), we propose segmentation and rotation of special
audits among targeted taxpayers.23 By way of illustration, approximately 640,000
self-employed PITs lodged a 2014 annual tax return, and there are roughly 4,600 tax
auditors (DGT 2015a). Assume that, in a given year, one tax auditor is able to perform
one special audit (that is, an audit related to a non-compliance indicator) for one PIT.
If these 4,600 auditors were spread equally across all the self-employed PITs, then the
audit rate for each self-employed PIT would be 0.7%. Under a focused enforcement
approach, the 640,000 self-employed PITs would be split into different sectors or
segments. In this case, for example, the focus of a ‘special audit project’ could be
the trading sector (categorised as ‘G’ in the 2012 Business Sector Classification).24
Assume that this sector accounts for about 30% of the whole. Within this sector, 10%
are those in car and motorcycle sales, repairs, and maintenance (coded as ‘45’ in the
classification). These 19,200 self-employed PITs would be the targeted subsector. If,
for example, 2,300 tax auditors (50%) were allocated to this subsector, taxpayers in
this subsector would face a 12% chance of being audited. Consequently, taxpayers
in other subsectors would face lower audit rates. But, because they already had a
very small chance of being audited, this variation would not be evident for them—
possibly due to a phenomenon termed ‘probability neglect’ (Sunstein 2002). Thus,
taxpayers outside the enforcement project may be unresponsive to a very small
reduction in the likelihood of being audited (that is, from 0.70% to 0.36%). On the
other hand, we argue that a significant increase in the probability of being audited
in the targeted subsector would increase taxpayers intention to report actual income,
because they now have a lower level of perceived behavioural control over tax
underreporting, particularly due to ‘uncertainty aversion’25 (Reeson and Dunstall
2009) and ‘availability bias’26 (Sunstein 2002). Accordingly, we suggest that their
levels of reported income would increase.

probably much lower. In 2011, for example, our calculation suggests that the random-based
audit rate for PITs was less than 0.02%.
23. Although audits are crucial in enforcing compliance among non-compliant taxpayers,
coercive audits might have undesired outcomes (Lisi 2014) because tax compliance results
from a combination of effective enforcement and mutual trust between taxpayers and the
authorities. Once both trust and power are at moderate levels, destructive multiplier effects
will occur, resulting in a downward spiral of decreasing compliance (Kirchler, Hoelzl, and
Wahl 2008).
24. Alternatively, this categorisation could be based on types of profession, such as lawyer,
accountant, architect, artist, and other prominent jobs.
25. Focused enforcement creates greater perceived uncertainty among the population
(Osofsky 2014). Taxpayers may know there is some chance of getting audited but not know
the increased probability. Research suggests that people tend to avoid uncertainty.
26. Cognitive bias suggests that individuals tend to rely more heavily on information that
is more readily accessible when evaluating the likelihood, or frequency, of an event.
48 Arifin Rosid, Chris Evans, and Binh Tran-Nam

To make this focused enforcement approach effective in raising the levels of

reported income, two further activities need to be undertaken. First, because
our findings indicate that respondents relied on mass media (that is, press,
television, and the internet) as their primary source of information, wide-reaching
announcements of a particular enforcement project must be made by the DGT
before the start of tax lodgement season. Those targeted in the focused enforcement
project would then recognise that they are able to change their underreporting
behaviour accordingly. Second, there should be no official announcement about
the withdrawal or completion of the enforcement project. This quiet withdrawal
may allow the DGT to free-ride for some time on the taxpayers’ perceived increased
chance of being audited and may extend the compliance benefits of the projects.
The third practical implication that derives from our findings is that the norm-
based intervention and focused enforcement approaches must be combined
proportionately. A relatively high level of compliance would appear to be necessary
in order to create norms of compliance and to foster future compliance.
Some scholars suggest that norms appear to be subject to so-called tipping points
(Besley and Persson 2014). This means the intended norms become more widespread
once some threshold is achieved. Unfortunately, in the Indonesian context, our
findings suggest that this is not the case. This implies that, as an initial step, the
focused enforcement approach would appear to be crucial in increasing the level of
compliance. Taxpayers knowing that a tax authority punishes tax evaders provides
the greatest leverage (Kirchler 2007). This approach must be complemented by
a norm-based intervention approach. If the norms that regulate the taxpayers’
behaviour remain unchanged, non-compliance will probably increase once the
probability of audit is lowered.
The fourth practical implication is that the integrity of the DGT must be
established and maintained to create trust (OECD 2010). A lack of integrity destroys
trust; trust and corruption cannot coexist (OECD 2004). While the DGT (2013) has
to some extent made efforts to build integrity and tackle corruption, our findings
imply that perceptions of tax corruption among taxpayers persist. Our results also
provide evidence that media coverage is very influential on the perceived levels of
corruption. Thus, an effective way to disseminate any effort of building integrity and
tackling corruption must be sought, either through taxpayers’ personal experiences
or through extensive media coverage. Also, credible anti-corruption measures
currently undertaken by the DGT should be made more visible to taxpayers. A
study by Kasper, Kogler, and Kirchler (2015) suggests that, when exposed to media,
content that emphasises the trustworthy nature of tax authorities or a capability to
deter tax evasion may increase taxpayers’ trust of the tax authorities.
Finally, the DGT needs to establish strategic alliances and partnerships with other
relevant government representatives. The objectives of this treatment strategy are
twofold. First, it should be evident that the pervasiveness of general corruption in
Indonesia, while significantly influential in terms of the operating context, is beyond
the control and responsibility of the DGT. The DGT needs to advise the relevant
institutions, at the highest governmental and political levels, of the adverse impact of
perceived corruption on its revenue collection. Indeed, a conducive political reality
of a country is one of the crucial ingredients in making major changes to taxation
(Bird 2013). It implies that strong commitment of political leadership to eradicate
the current levels of corruption is critical to optimal tax collection. Second, our
Tax Non-compliance and Perceptions of Corruption 49

FIGURE 3 A Combined Approach to Addressing the Malign Impact of Perceptions of

Corruption upon Intentional Underreporting Behaviour: A Tax Authority Perspective

findings suggest that perceived behavioural control over underreporting tax among
self-employed PITs is relatively high. There is a need to extend the nature and scope
of the information reporting regime by third parties (such as banks or other related
financial institutions).27 This is particularly important in an environment where a
large portion of self-employed PITs in Indonesia are presumably involved in cash
businesses and 60% of non-agricultural employment in Indonesia is informal (OECD
2015). Thus, the DGT needs strong support from other entities in implementing a
more extensive third-party information reporting regime to increase compliance by
identifying, controlling, and catching those who evade.
In summary, these practical implications are illustrated in figure 3.

CONCLUSION
This article describes the empirical relation between perceptions of corruption
and intentional underreporting, in the context of PITs in Indonesia. We show that
high levels of perceived corruption appear to have an adverse impact on taxpayers’
attitudes and norms towards reporting actual income. These circumstances, coupled
with relatively high levels of perceived behavioural control over tax underreporting,
appear to undermine taxpayers’ intention to report actual income.

27. The Indonesian government has recently (on 8 May 2017) issued Government Regulation
in Lieu of Law 1/2017 on Financial Information Access for Taxation Purposes, which gives
the DGT the power to scrutinise banking data.
50 Arifin Rosid, Chris Evans, and Binh Tran-Nam

Underreported tax, among both self-employed PITs and employed PITs, appears
to be prevalent. From a CRM perspective, underreporting is an identified risk that
the DGT should prioritise. We elaborate several strategies to address this risk by
establishing high levels of perceived corruption as an important external operating
context. We then discuss how our findings can be related to risk assessment and
prioritisation, compliance behaviour analysis, and the formulation of treatment
strategies within the CRM framework.
To the extent that the DGT has limited resources, we propose a balanced
combination of customised norm-based intervention and focused enforcement as
two primary treatment strategies to address intentional underreporting behaviour
among Indonesian PITs. Our discussion also suggests that to deter, detect, and
address intentional underreporting by Indonesian PITs, strategic alliances and
partnerships with other relevant government representatives are crucial for the DGT.
Corruption and tax (non-)compliance are distinct and separate problems. But,
as Alm, Martinez-Vazquez, and McClellan (2016) emphasise, they can easily
become intertwined and reinforcing. This article shows that corruption—or, more
strictly, perceptions of certain forms of corruption—can have a malign impact on
tax compliance through a variety of forces and relationships. This is particularly
problematic in countries like Indonesia, where corruption is perceived to be
endemic and tax compliance outcomes are typically sub-optimal. The article has
then established that within the broader context of compliance risk management,
and in the specific instance of Indonesia, much can still be done to mitigate that
malign impact.
The greater insights into the impact of corruption on tax compliance behaviour
derived from this study will never be more than a small—albeit important—part
of our overall understanding of tax compliance behaviour. The question of why
taxpayers comply with their tax obligations is not easy to answer. Accordingly, from
a tax authority perspective, a range of other options should also be considered to
influence taxpayers’ behaviour. These include, inter alia, improving the identification
and registration of taxpayers (including the detection of non-registration and
false registrations), the verification of received information (including audit and
enforcement activities), and the provision of service excellence and assistance to
taxpayers.

REFERENCES
Ajzen, Icek. 1991. ‘The Theory of Planned Behaviour’. Organizational Behavior and Human
Decision Processes 50 (2): 179–211.
Alm, James, Jorge Martinez-Vazquez, and Chandler McClellan. 2016. ‘Corruption and Firm
Tax Evasion’, Journal of Economic Behavior & Organization 124 (April): 146−63.
Alm, James, Gary H. McClelland, and William D. Schulze. 1992. ‘Why Do People Pay Taxes?’,
Journal of Public Economics 48 (1): 21−38.
Armstrong, J. Scott, and Terry S. Overton. 1977. ‘Estimating Nonresponse Bias in Mail
Surveys. Journal of Marketing Research 14 (3): 396−402.
Beer, Sebastian, Matthias Kasper, Erich Kirchler, and Brian Erard. 2016. ‘Do Audits Deter
Future Noncompliance? Evidence on Self-Employed Taxpayers’. Presentation given at
the Internal Revenue Service–Tax Policy Center Research Conference, Urban Institute,
Washington, DC, 11 November.
Besley, Timothy, and Torsten Persson. 2014. ‘Why Do Developing Countries Tax So Little?’.
Journal of Economic Perspectives 28 (4): 99–120.
Tax Non-compliance and Perceptions of Corruption 51

Bird, Richard M. 2013. ‘Taxation and Development: What Have We Learned from Fifty Years
of Research?’. Public Finance and Management 13 (4): 266−88.
———. 2015. ‘Improving Tax Administration in Developing Countries’. Journal of Tax
Administration 1 (1): 23−45.
Bird, Richard M., Jorge Martinez-Vazquez, and Benno Torgler. 2008. ‘Tax Effort in
Developing Countries and High Income Countries: The Impact of Corruption, Voice
and Accountability’. Economic Analysis & Policy 38 (1): 55−71.
Bird, Richard M., and Eric M. Zolt. 2005. ‘The Limited Role of the Personal Income Tax in
Developing Countries’. Journal of Asian Economics 16 (6): 928−46.
Blackburn, Keith, Niloy Bose, and M. Emranul Haque. 2010. ‘Endogenous Corruption in
Economic Development’. Journal of Economic Studies 37 (1): 4−25.
Bobek, Donna D., and Richard C. Hatfield. 2003. ‘An Investigation of the Theory of Planned
Behavior and the Role of Moral Obligation in Tax Compliance’. Behavioral Research in
Accounting 15 (1): 13−38.
Braun, Virginia, and Victoria Clarke. 2006. ‘Using Thematic Analysis in Psychology’.
Qualitative Research in Psychology 3: 77−101.
Brink, William D., and Porcano, Thomas M. 2016. ‘The Impact of Culture and Economic
Structure on Tax Morale and Tax Evasion: A Country-Level Analysis Using SEM’. In
Advances in Taxation. Vol. 23, edited by John Hasseldine, 87−123. Bingley: Emerald.
Bursztyn, Leonardo, Stefano Fiorin, Daniel Gottlieb, and Martin Kanz. 2015. ‘Moral
Incentives: Experimental Evidence from Repayments of an Islamic Credit Card’, NBER
Working Paper 21611, National Bureau of Economic Research, Cambridge, MA.
Campbell, Stuart V. 2013. ‘Perception Is Not Reality: The FPCA, Brazil and the
Mismeasurement of Corruption’. Minnesota Journal of International Law 22 (1): 247−81.
Çule, Monika, and Murray Fulton. 2009. ‘Business Culture and Tax Evasion: Why Corruption
and the Unofficial Economy Can Persist’. Journal of Economic Behavior & Organization 72
(3): 811−22.
Cummings, Ronald G., Jorge Martinez-Vazquez, Michael McKee, and Benno Torgler. 2009.
‘Tax Morale Affects Tax Compliance: Evidence from Surveys and an Artefactual Field
Experiment’. Journal of Economic Behavior & Organization 70 (3): 447−57.
De Vaus, David. 2014. Surveys in Social Research. 6th ed. Sydney: Allen & Unwin.
DGT (Directorate General of Taxation). 2013. Annual Report 2012: Harmonization in Building
the Nation. Jakarta: DGT.
———. 2015a, Annual Report 2014: Upholding the Spirit of Revitalization for Integrated
Transformation. Jakarta: DGT.
———. 2015b. Strategic Plan of the Directorate General of Taxation Years 2015–2019, Decree of
Director General of Taxation Number 95/PJ/2015. Jakarta: DGT.
Doig, Alan, and Robin Theobald. 1999. ‘Introduction: Why Corruption?’. Commonwealth &
Comparative Politics 37 (3): 1−12.
Duch, Raymond M., and Hector Solaz. 2016. ‘Why We Cheat: Evidence from Tax Compliance
Experiments’. Working paper, Centre for Experimental Social Sciences, Nuffield College,
University of Oxford, 12 November.
Fishbein, Martin, and Icek Ajzen. 2010. Predicting and Changing Behavior: The Reasoned Action
Approach. New York: Psychology Press.
Fjeldstad, Odd-Helge, and Bertil Tungodden. 2003. ‘Fiscal Corruption: A Vice or a Virtue?’.
World Development 31 (8): 1459−67.
Frey, Bruno S., and Benno Torgler. 2007. ‘Tax Morale and Conditional Cooperation’. Journal
of Comparative Economics 35 (1): 136−59.
Gangl, Katharina, Erich Kirchler, Christian Lorenz, and Benno Torgler. 2015. ‘Wealthy
Tax Non-filers in a Developing Nation: The Roles of Taxpayer Knowledge, Perceived
Corruption and Service Orientation in Pakistan’. CREMA Working Paper 2015-08, Center
for Research in Economics, Management and the Arts, Zurich.
Hair, Joseph F., William C. Black, Barry J. Babin, and Rolph E. Anderson. 2010. Multivariate
Data Analysis. 7th ed. Englewood, NJ: Pearson Prentice Hall.
52 Arifin Rosid, Chris Evans, and Binh Tran-Nam

Halla, Martin. 2012. ‘Tax Morale and Compliance Behavior: First Evidence on a Causal Link’.
BE Journal of Economic Analysis & Policy 12 (1): 1935−1682.
Hallsworth, Michael, John A. List, Robert D. Metcalfe, and Ivo Vlaev. 2017. ‘The Behavioralist
as Tax Collector: Using Natural Field Experiments to Enhance Tax Compliance’. Journal
of Public Economics 148:14−31.
ILO (International Labour Office). 2015. Labour and Social Trends in Indonesia 2014−2015:
Strengthening Competitiveness and Productivity through Decent Work. Jakarta: ILO.
Imam, Patrick A., and Davina F. Jacobs. 2007. ‘Effect of Corruption on Tax Revenues in the
Middle East’. IMF Working Paper WP/07/270, International Monetary Fund, Washington,
DC.
IMF (International Monetary Fund). 2011. ‘Revenue Mobilization in Developing Countries’.
Policy paper, IMF, Washington, DC.
Jahnke, Björn. 2015. ‘How Does Petty Corruption Affect Tax Morale in Sub-Saharan Africa?
An Empirical Analysis’. Hannover Economic Paper 564, Leibniz Universität, Hannover.
James, Simon, and Clinton Alley. 1999. ‘Tax Compliance, Self-Assessment, and Tax
Administration’. Journal of Finance and Management in Public Services 2 (1): 27−42.
Kasper, Matthias, Christoph Kogler, and Erich Kirchler. 2015. ‘Tax Policy and the News:
An Empirical Analysis of Taxpayers’ Perceptions of Tax-Related Media Coverage and
Its Impact on Tax Compliance’. Journal of Behavioral and Experimental Economics 54:58–63.
Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi. 2007. Measuring Corruption: Myths
and Realities. Washington, DC: World Bank.
Kettle, Stewart, Marco Hernandez, Simon Ruda, and Michael Sanders. 2015. ‘Behavioural
Interventions in Tax Compliance: Evidence from Guatemala’. Internal Revenue Service
(IRS) Research Bulletin. https://www.irs.gov/pub/irs-soi/15resconhemandez.pdf.
Kirchler, Erich. 2007. The Economic Psychology of Tax Behaviour. Cambridge: Cambridge
University Press.
Kirchler, Erich, Erik Hoelzl, and Ingrid Wahl. 2008. ‘Enforced versus Voluntary Tax
Compliance: The “Slippery Slope” Framework’. Journal of Economic Psychology 29:210–25.
Kuncoro, Ari. 2004. ‘Bribery in Indonesia: Some Evidence from Micro-Level Data’. Bulletin
of Indonesian Economic Studies 40 (3): 329–54.
Langham, Jo’Anne, Neil Paulsen, and Charmine E. J. Hartel. 2012. ‘Improving Tax Compliance
Strategies: Can the Theory of Planned Behavior Predict Business Compliance?’. EJournal
of Tax Research 10 (2): 364−402.
Lewis, Alan. 1982. The Psychology of Taxation, Oxford: Martin Robertson.
Lisi, Gaetano. 2014. ‘The Interaction between Trust and Power: Effects on Tax Compliance
and Macroeconomic Implications’. Journal of Behavioral and Experimental Economics 53:24−33.
Litina, Anastasia, and Theodore Palivos. 2016. ‘Corruption, Tax Evasion and Social Values’.
Journal of Economic Behavior & Organization 124:164−77.
McKerchar, Margaret, and Chris Evans. 2009. ‘Sustaining Growth in Developing Economies
through Improved Taxpayer Compliance: Challenges for Policy Makers and Revenue
Authorities’. eJournal of Tax Research 7 (2): 171−201.
McLeod, Ross H. 2011. ‘Survey of Recent Developments’. Bulletin of Indonesian Economic
Studies 47 (1): 7–34.
Melgar, Natalia, Máximo Rossi, and Tom W. Smith. 2010. ‘The Perception of Corruption’.
International Journal of Public Opinion Research 22 (1): 120−31.
Mishler, William, and Richard Rose. 2008. ‘Seeing Is Not Always Believing: Measuring
Corruption Perceptions and Experiences’. Paper presented at the Elections, Public
Opinion and Parties 2008 Annual Conference, University of Manchester, 12–14 September.
Nguyen, Ngoc Anh, Quang Hung Doan, Ngoc Minh Nguyen, and Binh Tran-Nam. 2016.
‘The Impact of Petty Corruption on Firm Innovation in Vietnam’. Crime, Law and Social
Change 65 (4): 377−94.
OECD (Organisation for Economic Co-operation and Development). 2004. ‘Compliance Risk
Management: Managing and Improving Tax Compliance’. CTPA guidance note, Centre
for Tax Policy and Administration, OECD, Paris.
Tax Non-compliance and Perceptions of Corruption 53

———. 2010. ‘Understanding and Influencing Taxpayers’ Compliance Behaviour’. CTPA

information note, Centre for Tax Policy and Administration, OECD, Paris.
———. 2012. Economic Policy Reforms: Going for Growth. Paris: OECD.
———. 2014a. Measures of Tax Compliance Outcomes: A Practical Guide. Paris: OECD.
———. 2014b. Revenue Statistics in Asian Countries: Trends in Indonesia and Malaysia, 1990–
2012. Paris: OECD.
———. 2015. OECD Economic Survey: Indonesia 2015. Paris: OECD.
Onu, Diana, and Lynne Oats. 2015. ‘The Role of Social Norms in Tax Compliance: Theoretical
Overview and Practical Implications’. Journal of Tax Administration 1 (1): 113−37.
Osofsky, Leigh. 2014. ‘Concentrated Enforcement’. Florida Tax Review 16 (6): 325−91.
Reeson, Andrew, and Simon Dunstall. 2009. Behavioural Economics and Complex Decision-
Making: Implications for the Australian Tax and Transfer System. CMIS Report 09/110.
Canberra: CSIRO.
Ritsatos, Titos. 2014. ‘Tax Evasion and Compliance: From the Neo Classical Paradigm to
Behavioural Economics, A Review’. Journal of Accounting & Organizational Change 10 (2):
244−62.
Rosid, Arifin, Chris Evans, and Binh Tran-Nam. 2016. ‘Do Perceptions of Corruption
Influence Personal Income Taxpayer Reporting Behaviour? Evidence from Indonesia’,
EJournal of Tax Research 14 (2): 387−425.
Shah, Rachna, and Susan Meyer Goldstein. 2006. ‘Use of Structural Equation Modeling in
Operations Management Research: Looking Back and Forward’. Journal of Operations
Management 24:148−69.
Shleifer, Andrei, and Robert W. Vishny. 1993. ‘Corruption’. Quarterly Journal of Economics
108 (3): 599−617.
Smith, A. D. 2001. ‘Perception and Belief’. Philosophy and Phenomenological Research 62 (2):
283−309.
Sunstein, Cass R. 2002. ‘Probability Neglect: Emotions, Worst Cases, and Law’. Yale Law
Journal 112 (1): 61−107.
Torgler, Benno, Ihsan C. Demir, Alison Macintyre, and Markus Schaffner. 2008. ‘Causes
and Consequences of Tax Morale: An Empirical Investigation’. Economic Analysis and
Policy 38 (2): 313−39.
Tran-Nam, Binh. 2003. ‘Tax Compliance Research: An Economic Perspective’. New Zealand
Journal of Taxation Law and Policy 9:455−68.
Transparency International. 2016. ‘Corruption Perceptions Index 2015’. Last modified 1
February. http://www.transparency.org/cpi2015.
Trivedi, Viswanath Umashanker, Mohamed Shehata, and Stuart Mestelman. 2005. ‘Attitudes,
Incentives, and Tax Compliance’. Canadian Tax Journal 53:29−61.
UNDP (United Nations Development Programme). 2008. A Users’ Guide to Measuring
Corruption. Oslo: UNDP Oslo Governance Centre.
United Nations. 2015. ‘Transforming Our World: The 2030 Agenda for Sustainable
Development’. General Assembly resolution A/RES/70/1, United Nations, New York.
Walsh, Keith. 2012. ‘Understanding Taxpayer Behaviour—New Opportunities for Tax
Administration’. Economic and Social Review 43:451−75.
Wearing, Alexander, and Bruce Headey. 1997. ‘The Would-Be Tax Evader: A Profile’.
Australian Tax Forum 13:3−17.
Wenzel, Michael, and Natalie Taylor. 2003. ‘Towards Evidence-Based Tax Administration’.
Australian Journal of Social Issues 38:409−30.
Wood, Charlotte, Mary Ivec, Jenny Job, and Valerie Braithwaite. 2010. ‘Applications
of Responsive Regulatory Theory in Australia and Overseas’. Occasional Paper 15,
Regulatory Institutional Network, The Australian National University, Canberra.
World Bank. 2015. World Development Report 2015: Mind, Society, and Behavior. Washington,
DC: World Bank.
World Economic Forum. 2016. The Global Competitiveness Report 2016–2017. Geneva: World
Economic Forum.
 TABLE A1  Survey Questions

Construct Questions Code Rating scale

A. Perception The current level of corruption is high. PGC1 Unipolar disagree–agree

of general Most people think that the level of corruption is high. PGC2 scale, scored 1–7
corruption The frequency of corruption is rare. PGC3R
(PGC) The current level of corruption is a serious problem. PGC4
When I think about the level of corruption, I am unhappy. PGC5
B. Perception The current level of corruption involving high-level public officials is high. GCO1 Unipolar disagree–agree
of grand Most people think that the level of corruption involving high-level public GCO2 scale, scored 1–7
corruption officials is high.
(GCO) The frequency of corruption involving high-level public officials is rare. GCO3R
The current level of corruption involving high-level public officials is a serious problem. GCO4
When I think about the level of corruption involving high-level public officials, I am GCO5
unhappy.
C. Perception The current level of corruption involving low-level public officials is high. PCO1 Unipolar disagree–agree
of petty Most people think that the level of corruption involving low-level public officials is PCO2 scale, scored 1–7
corruption high.
(PCO) The frequency of corruption involving low-level public officials is rare. PCO3R
The current level of corruption involving low-level public officials is a serious problem. PCO4
When I think about the levels of corruption involving low-level public officials, I am PCO5
unhappy.
D. Perception The current level of corruption involving high-level tax officials is high. GTC1 Unipolar disagree–agree
of grand tax Most people think that the level of corruption involving high-level tax officials is high. GTC2 scale, scored 1–7
corruption The frequency of corruption involving high-level tax officials is rare. GTC3R
(GTC) The current level of corruption involving high-level tax officials is a serious problem. GTC4
When I think about the levels of corruption involving high-level tax officials, I am GTC5
unhappy.
E. Perception The current level of corruption involving low-level tax officials is high. PTC1 Unipolar disagree–agree
of petty tax Most people think that the level of corruption involving low-level tax officials is high. PTC2 scale, scored 1–7
corruption The frequency of corruption involving low-level tax officials is rare. PTC3R
(PTC) The current level of corruption involving low-level tax officials is a serious problem. PTC4
When I think about the levels of corruption involving low-level tax officials, I am PTC5
unhappy.
F. Attitude I would feel guilty if I underreported my actual income in my annual tax return. ATB1 Unipolar disagree–agree
towards To feel guilty for underreporting income tax is good. ATB2 scale, scored 1–7
(under-) Underreporting my income tax makes me better off. ATB3
reporting There are a number of government services, infrastructures, facilities for which I am ATB4R
tax (ATB) very thankful.
Underreporting income is acceptable if any portion of the money collected is wasted by ATB5
government.
Paying as little tax as possible is important. ATB6
Government has spent the money collected from tax efficiently. ATB7R
Underreporting my income will not hurt the society as a whole. ATB8
I feel that I have made a positive contribution to my country by fully reporting all my ATB9R
income.
G. Subjective Most taxpayers expect me to report all my income on the annual tax return. SNO1R Unipolar disagree–agree
norms Generally, I would do what most taxpayers expect me to do. SNO2 scale, scored 1–7
towards Most people who are important to me (e.g., family, friends & business partners) SNO3R
(under-) expect me to report all my income on the annual tax return.
reporting Generally, I would do what people who are important to me would expect me to do. SNO4
tax (SNO) Most taxpayers underreport their actual income on their annual tax returns. SNO5
Generally, I would do what most other taxpayers do. SNO6
The people who are important to me (e.g., family, friends & business partners) SNO7
underreport their actual income on their annual tax returns.
Generally, I would do what people who are important to me would do. SNO8
H. Perceived I could underreport my actual income if I wanted to. PBC1 Unipolar disagree–agree
behavioural I have many opportunities to underreport my income in the annual tax return. PBC2 scale, scored 1–7
control over People I know do not fully report their income and they have never been detected. PBC3
(under) It is easy to exclude some of my income in completing my tax return. PBC4
reporting With my knowledge and experience, it is easy for me to underreport my actual income PBC5
tax (PBC) without worrying about getting caught.
People I know do not fully report their income and they have never been penalised. PBC6
All my income is subject to tax withholding tax by others. PBC7R
The DGT is unlikely to detect whether I underreport my income. PBC8

I. Intention Scenario A
to report Agus is a full-time employee. His monthly income is subject to withholding tax by his employer. In 2014, he received a casual
actual commission for his effort to sell a second-hand house. The commission was about Rp 7 million & was paid in cash by his friend.
income He understood that as the commission was paid in cash, it would be difficult for the DGT officials to detect this irregular income.
(ITC)
If I were Agus, the probability I would fully report my commission in my annual tax ITC1 Unipolar unlikely–likely
return is ... scale, scored 1–7

If I were Agus, the most likely percentage of the commission I would intend to report ITC2 Unipolar scale nothing–
on the annual tax return is .... all, scored 0%, 16%, 33%,
50%, 67%, 83%, 100%
Scenario B
Ani is a self-employed taxpayer. Categorised as small and medium enterprise (SME), she should pay 1% final income tax based
on her gross sales turnover. As her average monthly turnover is between Rp 100 million & Rp 200 million, her monthly tax
payment should range between Rp 1 million & Rp 2 million. She knows that as her annual gross sales turnover is less than Rp 4.8
billion, she is required to only do simple document recording instead of standard bookkeeping. Reflecting on the nature of her
business activities and the extent of simple recording, she understands that it would be difficult for the DGT officials to trace her
actual sales turnover.

If I were Ani, the probability I would fully report my actual sales turnover in my annual ITC3 Unipolar unlikely–likely
tax return is .... scale, scored 1–7

If I were Ani, the most likely percentage of sales turnover I would intend to report in ITC4 Unipolar scale nothing–all,
my annual tax return is .... scored 0%, 16%, 33%, 50%,
67%, 83%, 100%
J. Income I have fully reported my actual income in my annual tax return for fiscal year 2014. TCB1 Unipolar disagree-agree
reporting scale, scored 1–7
behaviour As far as I can remember, the amount of income I have reported in my annual tax TCB2 Unipolar scale nothing-all,
(TCB)* return was roughly ... of my actual income.** scored 0%, 16%, 33%, 50%,
67%, 83%, 100%

* = an ‘everybody does it’ technique (De Vaus 2014) was used to reduce potential social desirability bias in responding to questions in this section. To do so, the
phrase ‘not everyone necessarily reports all of his or her actual income to the tax office’ was added as a preamble in the actual questionnaire.
** = Q55 and Q56 were questions for self-employed PITs. In the actual survey form, this section was specified into two groups of respondents: those whose main
income was from employment (i.e., employed PITs) and those whose was not (i.e., self-employed PITs). Also, in the measurement scale, optional answers such as ‘I
did not receive any income other than from employment’ or ‘I did not receive any income’ were included.
58 Arifin Rosid, Chris Evans, and Binh Tran-Nam

TABLE A2  Correlations between Constructs and Square Root of AVE

Model A−fulla Model A−partialb

PGC ATB SNO PBC ITC TCB PGC ATB SNO PBC ITC
PGC 0.66 PGC 0.72
ATB 0.12 0.61 ATB 0.27 0.63
SNO 0.01 0.09 0.62 SNO 0.06 0.09 0.64
PBC 0.11 0.06 −0.02 0.61 PBC 0.23 0.18 0.14 0.75
ITC −0.23 −0.46 −0.17 −0.21 0.71 ITC −0.39 −0.51 −0.24 −0.34 0.76
TCB −0.24 0.08 −0.07 −0.02 0.59 0.81

Model B−full Model B−partial

GCO PCO ATB SNO PBC ITC TCB GCO PCO ATB SNO PBC ITC
GCO 0.60 GCO 0.67
PCO 0.38 0.56 PCO 0.39 0.65
ATB 0.29 0.06 0.61 ATB 0.35 0.13 0.63
SNO 0.16 0.33 0.09 0.62 SNO 0.16 0.33 0.09 0.64
PBC 0.09 0.02 0.06 −0.02 0.61 PBC 0.11 0.06 0.18 0.14 0.75
ITC −0.30 −0.26 −0.46 −0.17 −0.21 0.71 ITC −0.31 −0.30 −0.50 −0.24 −0.34 0.76
TCB −0.18 −0.12 0.07 −0.07 −0.01 0.60 0.81

Model C−full Model C−partial

GTC PTC ATB SNO PBC ITC TCB GTC PTC ATB SNO PBC ITC
GTC 0.74 GTC 0.72
PTC 0.13 0.68 PTC 0.21 0.70
ATB 0.28 0.11 0.61 ATB 0.25 0.23 0.63
SNO 0.07 −0.04 0.09 0.62 SNO 0.12 0.10 0.09 0.64
PBC 0.05 0.04 0.07 −0.02 0.61 PBC 0.10 0.06 0.18 0.14 0.75
ITC −0.34 −0.17 −0.46 −0.17 −0.21 0.71 ITC −0.37 −0.30 −0.50 −0.24 −0.34 0.76
TCB −0.11 −0.10 0.07 −0.07 −0.02 0.60 0.81

Source: Calculated from the survey data.

Note: PGC = perceptions of general corruption; ATB = attitude towards tax underreporting; SNO  =
subjective norm towards tax underreporting; PBC = perceived behavioural control over tax
underreporting; ITC = intention to report actual income; TCB = level of reported income. GCO =
perceptions of grand corruption; PCO = perceptions of petty corruption; GTC = perceptions of grand
tax corruption; PTC = perceptions of petty tax corruption.
a
For full model, n = 223.
b
For partial model, n = 397.
 TABLE A3  Hypothesised Channels of Causality Between Perceptions of Corruption
and Tax (Non-)Compliance Behaviour: Path Coefficients and Model Fit Indices

Modified full model (n = 223) Modified partial model (n = 397)

MODEL A-full; Chi-square: 550.580 MODEL A-partial; Chi-square: 568.805

p-value: 0.128; χ2/df: 1.071 p-value: 0.064; χ2/df: 1.096
TLI: 0.981; CFI: 0.982; RMSEA: 0.018 TLI: 0.990; CFI: 0.991; RMSEA: 0.016

MODEL B-full; Chi-square: 723.682 MODEL B-partial; Chi-square: 713.259

p-value: 0.168; χ2/df: 1.052 p-value: 0.262; χ2/df: 1.034
TLI: 0.982; CFI: 0.983; RMSEA: 0.015 TLI: 0.996; CFI: 0.996; RMSEA: 0.009

MODEL C-full; Chi-square: 731.874 MODEL C-partial; Chi-square: 722.505

p-value: 0.120; χ2/df: 1.064 p-value: 0.190; χ2/df: 1.025
TLI: 0.981; CFI: 0.983; RMSEA: 0.017 TLI: 0.994; CFI: 0.995; RMSEA: 0.011

Note: * = significant at p < 0.05; ** = significant at p < 0.1; PGC = perceptions of general corruption; GCO
= perceptions of grand corruption; PCO = perceptions of petty corruption; GTC = perceptions of grand
tax corruption; PTC = perceptions of petty tax corruption; ATB = attitude towards tax underreporting;
SNO = subjective norm towards tax underreporting; PBC = perceived behavioural control over tax
underreporting; ITC = intention to report actual income; TCB = level of reported income. TLI = Tucker–
Lewis index; CFI = comparative fit index; RMSEA = root mean square error of approximation.
Source: Calculated from the survey data.
 TABLE A4  Effects Decomposition of Perceptions of Corruption on Intention
to Report Actual Income (ITC) and the Level of Reported Income (TCB)

Endogenous variables

Intention to Level of
report actual reported
Model category Exogenous variables & effects income (ITC) income (TCB)

Full models A. Perception of general corruption (PGC)

(n = 223) Direct effect – —
Total indirect effects −0.061** −0.034**
Total effects −0.061** −0.034**
B. Perception of grand corruption (GCO)
Direct effect — —
Total indirect effects −0.131* −0.073*
Total effects −0.131* −0.073*
C. Perception of petty corruption (PCO)
Direct effect — —
Total indirect effects −0.027 −0.015
Total effects −0.027 −0.015
D. Perception of grand tax corruption (GTC)
Direct effect — —
Total indirect effects −0.127* −0.071*
Total effects −0.127* −0.071*
E. Perception of petty tax corruption (PTC)
Direct effect — —
Total indirect effects −0.025 −0.014
Total effects −0.025 −0.014
Partial models A. Perception of general corruption (PGC)
(n = 397) Direct effect — NA
Total indirect effects −0.142* NA
Total effects −0.142* NA
B. Perception of grand corruption (GCO)
Direct effect — NA
Total indirect effects −0.167* NA
Total effects −0.167* NA
C. Perception of petty corruption (PCO)
Direct effect — NA
Total indirect effects −0.061* NA
Total effects −0.061* NA
D. Perception of grand tax corruption (GTC)
Direct effect — NA
Total indirect effects −0.121* NA
Total effects −0.121* NA
E. Perception of petty tax corruption (PTC)
Direct effect — NA
Total indirect effects −0.107* NA
Total effects −0.107* NA

Note: NA = not applicable. * = significant at p < 0.05; ** = significant at p < 0.10.

Source: Calculated from the survey data.
 Bulletin of Indonesian Economic Studies

ISSN: 0007-4918 (Print) 1472-7234 (Online) Journal homepage: https://www.tandfonline.com/loi/cbie20

Non-Tariff Trade Regulations in Indonesia:

Nominal and Effective Rates of Protection

Stephen V. Marks

To cite this article: Stephen V. Marks (2017) Non-Tariff Trade Regulations in Indonesia: Nominal
and Effective Rates of Protection, Bulletin of Indonesian Economic Studies, 53:3, 333-357, DOI:
10.1080/00074918.2017.1298721

To link to this article: https://doi.org/10.1080/00074918.2017.1298721

Accepted author version posted online: 20

Mar 2017.
Published online: 12 Feb 2018.

Submit your article to this journal

Article views: 1585

View related articles

View Crossmark data

Citing articles: 6 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=cbie20
 Bulletin of Indonesian Economic Studies, Vol. 53, No. 3, 2017: 333–57

NON-TARIFF TRADE REGULATIONS IN INDONESIA:

NOMINAL AND EFFECTIVE RATES OF PROTECTION

Stephen V. Marks*
Pomona College

Non-tariff regulations on imports and exports have spread in Indonesia since 2011. I
report findings of a study of many of these regulations, in which a variety of methods
were used to estimate the associated nominal rates of protection. These findings were
then used to estimate effective rates of protection (ERPs) across 140 tradable-goods
sectors in the Indonesian economy in early 2015, taking into account also the effects
of the most-favoured-nation and preferential-import tariff schedules, anti-dumping
and safeguard duties, export levies, duty drawbacks and exemptions, domestic sub-
sidies, and excise taxes. I find that the magnitude and dispersion of ERPs were higher
in 2015 than in early 2008, for which a similar study was previously conducted, and
that much of the variability was related to quantitative trade restrictions. In particu-
lar, the regulations examined boosted a measure of the cost of living by 7.6% in 2015,
compared with 2.5% in 2008.

Keywords: nominal and effective rates of protection, non-tariff trade measures

JEL classification: F13

INTRODUCTION
While import tariff increases have been constrained by multilateral and regional
trade agreements, non-tariff trade regulations have spread in Indonesia in recent
years. The share of tariff lines subject to non-tariff measures (NTMs) on the import
side grew from 42% in 2009 to 51% in 2015. On the export side, the share of tariff
lines subject to NTMs grew from 4% in 2009 to 10% in 2015. Many lines are now
subject to multiple measures.
This study estimates the effects of many NTMs in goods markets on prices
and incentives in the Indonesian economy, through the estimation of nominal
and effective rates of protection based on policies in effect in early 2015. Also

* Most of the work on this article was completed while the author was with the Australia
Indonesia Partnership for Economic Governance (AIPEG), a cooperative arrangement
between the Australian Department of Foreign Affairs and Trade and the Indonesian
Ministry of Trade. The author is grateful to Paul Bartlett, Ernawati Munadi, and Achmad
Shauki of AIPEG, Massimiliano Cali of the World Bank, participants at a seminar in the
Policy Dialogue Series at the Ministry of Trade on 19 August 2015, and several anonymous
referees for their comments on earlier versions of this work. The views expressed in the
article are those of the author alone.

ISSN 0007-4918 print/ISSN 1472-7234 online/17/000333-25 © 2017 ANU Indonesia Project

http://dx.doi.org/10.1080/00074918.2017.1298721
334 Stephen V. Marks

included in the analysis are import duties and export taxes, commodity subsidies,
and excise taxes.
Table 1 shows the product categories covered by the NTMs and export levies
included in this study.1 Given the extent of NTMs throughout the economy, I
focus on regulations that have the greatest potential to distort market conditions—
particularly quantitative restrictions on exports and imports, such as mandatory
licensing, quotas, and bans. Also included are some procedural measures that add
to the costs of importation or exportation.2
Some trade regulations, even quantitative restrictions, may have benign moti-
vations. For example, in addition to its effects on the commodities listed in the
table, Minister of Trade Regulation 44/2012 bans exports of endangered plant and
animal species. In any case, it is not the intention of this article to evaluate the
merits of the policies, but rather to summarise their effects on prices and incen-
tives within Indonesia.

ESTIMATION OF THE EFFECTS OF NON-TARIFF MEASURES

I measured the economic effects of NTMs through the associated nominal rates of
protection (NRPs), which indicate the percentage by which policies raise the pro-
ducer price inside the country relative to the border price. Given data limitations
and methodological concerns, I used a variety of methods to obtain estimates of
the NRPs associated with major NTMs.3 The remainder of this section outlines
these methods and provides a few notable examples of their application. More
complete details on the methodologies and findings are given in Marks (2015).
Throughout the analysis, I used the most defensible methodologies possible. I was
mostly not able to conduct sensitivity analysis in the traditional sense. However,
I was able to obtain estimates for some sectors by using more than one methodo-
logical approach. In such cases, I typically favoured the more reliable method a
priori, or simply the more conservative estimate.
Some NRP estimates were obtained through a limited number of interviews
with market participants, particularly for procedural trade regulations that applied
to multiple sectors (shown at the bottom of each of the two parts in table 1, and
discussed further below). In these cases, I typically had either multiple sources
for these figures, or detailed data (and a high level of confidence in those data)
from a single source.

1. The Indonesia National Trade Repository in the Indonesia National Single Window pro-
vides a comprehensive guide to the presence of various NTMs. See http://eservice.insw.
go.id/, particularly the bans and restrictions section: ‘Lartas—Larangan dan Pembatasan’.
2. I did not examine regulations on prescription drugs, narcotics and their precursors,
weapons and munitions, gemstones, nuclear power plants, waste products, and the like,
or bans or restrictions on imports of many used capital goods.
3. Deardorff and Stern (1998) provide a rationale and guidelines for a price comparison
approach. Kee, Nicita, and Olarreaga (2008) apply a theoretically grounded economet-
ric approach. Their estimates are detailed but in some cases questionable for Indonesia.
Gravity models have also been used to estimate the effects of NTMs, but are necessarily
applied at a highly aggregated level. A recent example is Devadason and Chenayah (2014).
 TABLE 1  Major Non-tariff Measures and Export Levies Examined in This Study

Commodity or group Type of regulation Recent regulation

Imported
Livestock & livestock
products Licensing Minister of Trade Regulation 46/2013
Horticulture Licensing Minister of Trade Regulation 40/2015
Cloves Licensing Minister of Industry and Trade Decree
528/2002
Milled, polished rice Licensing Minister of Trade Regulation 19/2014
Wheat flour Quota Minister of Trade Regulation 23/2014
Sugar Licensing Minister of Trade Regulation 19/2008
Alcoholic beverages Licensing Minister of Trade Regulation 72/2014
Salt Licensing Minister of Trade Regulation 58/2012
Baby clothes Product standards Minister of Industry Regulation 72/2012
Plastic raw materials Licensing Minister of Trade Regulation 36/2013
Cement Licensing Minister of Trade Regulation 40/2013
Steel Licensing Minister of Trade Regulation 8/2012
Optical media & equipment Licensing Minister of Trade Regulation 35/2012
Colour printers & copiers Licensing Minister of Trade Regulation 15/2007
Cellular telephones, tablet
computers Licensing Minister of Trade Regulation 38/2013
Children’s toys Product standards Minister of Industry Regulation 24/2013
Certain productsa Multipleb Minister of Trade Regulation 36/2014
Various productsc Quarantined Minister of Agriculture Regulation
12/2015
Exported
Milled, polished rice Licensing Minister of Trade Regulation 10/2014
Cocoa beans Tax Minister of Finance Regulation 75/2012
Rattan Ban Minister of Trade Regulation 35/2011
Palm oil products Taxes Minister of Finance Regulation 136/2015
Palm oil products Service fees Minister of Finance Regulation 133/2015
Minerals Bans, restrictions Minister of Trade Regulation 4/2014
Minerals Taxes Minister of Finance Regulation 153/2014
Tin Restrictions Minister of Trade Regulation 32/2013
Tin ores & concentrates Ban Minister of Trade Regulation 44/2012
Sand, clay, top soil Ban Minister of Trade Regulation 44/2012
Coal Licensing Minister of Trade Regulation 39/2014
Animal hides Taxes Minister of Finance Regulation 128/2013
Logs Ban Minister of Trade Regulation 44/2012
Wood products Taxes Minister of Finance Regulation 128/2013
Various primary productse Trade financef Minister of Trade Regulation 4/2015
a
  The regulation uses the term produk tertentu; covered are many processed foods and beverages,
personal care products, traditional medicines, apparel and other finished textile products, footwear
products, household electrical appliances, consumer electronics products, and children’s toys.
b
  Pre-shipment inspections, port restrictions, special importer registration, and additional docu-
mentation requirements.
c
  Live animals and products containing animal or plant materials, including wooden packaging of
imported goods.
d
  The length of the quarantine depends on the nature of the product.
e
  Crude palm oil and palm kernel oil, coal, crude oil, natural gas, certain petroleum products, and
certain mineral and metal products.
f
  Requirement to finance exports through a letter of credit from a bank.
336 Stephen V. Marks

Price Comparisons between Countries over Time

To estimate the NRPs associated with NTMs in Indonesia, Fane and Condon (1996)
used price comparisons between Indonesia and Singapore obtained by the World
Bank in 1993, while observing in an unpublished appendix that ‘the price com-
parison data are unavoidably subjective because of the difficulties of allowing
for differences in quality; transport, handling and storage costs; and other non-
pecuniary factors’.
An ideal way to correct for such factors is to look at the difference between
a domestic price in Indonesia and an external reference price, during a period
in which a policy was applied in Indonesia versus a period in which trade was
unimpeded. The intertemporal aspect of the comparison is intended to correct
for country fixed effects, and the cross-country aspect for international market
developments unrelated to the policy. Owing to data limitations, I was able to use
this difference-in-differences approach to estimate NRPs in recent years for only
four imported commodities: beef (for which the NRP was estimated to be 17.2%),
chicken meat (29.7%), shallots (61.9%), and wheat flour (22.0%).
I will focus on just one of these commodities, wheat flour, in this article, since
it offers a particularly clear illustration. Figure 1 shows the sharp increase in the
margin by which the retail price of wheat flour in Indonesia exceeded the whole-
sale price of wheat flour in the city of Minneapolis in the United States during
the period from late 2008 to 2015, relative to the period from 2000 up to that time.
This jump in the margin resulted from a series of policy actions taken by the
government of Indonesia. Flour millers in Indonesia initiated an anti-dumping
action against wheat flour from Turkey, Sri Lanka, and Australia in October 2008,
and in 2009 the government imposed anti-dumping duties ranging from 18.69% to
21.99% on wheat flour imports from Turkey. In 2012, a temporary 20% safeguard
duty was imposed on all imports of wheat flour. Temporary anti-dumping duties
were imposed on India, Sri Lanka, and Turkey in 2013, followed by temporary
across-the-board import quotas in 2014; wheat flour imports were 73.6% lower in
2014 than in 2011.

Retail Price Comparisons at a Point in Time

For many other consumer products, and a few construction materials, I and a team
from the Australia Indonesia Partnership for Economic Governance (AIPEG) and
the Indonesian Ministry of Trade compared retail prices of identical products in
Jakarta and Singapore in early 2015.4 Singapore has zero import tariffs on almost
all items and is relatively free of non-tariff trade regulations. However, it is also
an expensive city with relatively high labour and land costs for wholesale and
retail trade, and there are differences between the two countries in commod-
ity taxation: Singapore’s goods and services tax is applied at a uniform 7% rate,
whereas Indonesia’s value-added tax is applied at a 10% rate (but with exemptions
for many products).

4. Fane and Condon (1996) did this for a wide range of commodities, including industrial
inputs for which prices were not posted, but did not adjust the data in the way described in
this study (see below). A problem they encountered was that the prices of a number of prod-
ucts were lower in Indonesia than in Singapore, once the prices were converted into common
currency units, despite the presence of binding quantitative import barriers in Indonesia.
Non-tariff Trade Regulations in Indonesia 337

FIGURE 1  Wheat Flour Prices, Indonesia and the

United States, 2000/01–2014/15 ($/kg)

1.0

0.8

0.6
Average retail flour
price, Indonesia
0.4

Wholesale price
0.2 of bakery flour,
Minneapolis

0.0
2000– 2002– 2004– 2006– 2008– 2010– 2012– 2014–
01 03 05 07 09 11 13 15
Note: The US price data are from the Wheat Yearbook of the Economic Research Service, US Department
of Agriculture, and are given on a quarterly basis. The Indonesian data are quarterly averages of
monthly data from the Indonesian Ministry of Trade.

To make the comparisons as precise as possible, we compared prices at similar

types of retail outlets in Singapore and Jakarta. Nevertheless, there may have been
differences between similar types of stores: upscale grocery stores in Jakarta may
have had a greater degree of market power than those in Singapore, for example.
Thus, in addition to any necessary adjustments for differences in commodity taxes,
I adjusted for other differences between the two cities by comparing the prices
of regulated commodities with those of commodities that were not subject to
major NTMs in Indonesia. The comparators used for the adjustments—Xiang Lie
pears for non-durables and semi-durables (22.2% cheaper in Jakarta) and mono-
chrome laser printers for durables (11.5% cheaper in Jakarta)—were subject to zero
import duties in both countries. (In the case of the pears, the zero import duties in
Indonesia were under the ASEAN–China Free Trade Agreement.)
This alternative difference-in-differences approach offers a less robust way to
correct for country fixed effects than the one described above, considering the idio­
syncrasies of the various markets.5 Given the range of the price comparison data
obtained for many commodities, however, these comparators are conservative in

5. For example, household-type instantaneous water heaters were subject to procedural

trade barriers in Indonesia. In our survey, these heaters were 112.8% more expensive on
an adjusted basis in Jakarta than in Singapore. Because the most prominent brands were
manufactured in Singapore, however, that market may have been saturated, while in
Jakarta the heaters may have been a luxury item that commanded a premium price. In
some cases in which I had alternative data on price differentials for a given commodity, I
favoured the figures obtained by other methods; see note 12 below.
338 Stephen V. Marks

that their use tends to understate the amount by which the price of the regulated
commodity is higher in Jakarta than in Singapore. Among the sources of evidence
consistent with this is the International Comparison Program (ICP) coordinated by
the World Bank, and implemented in Asia by the Asian Development Bank (ADB
2014). In contrast to the present study, the ICP price comparisons were not confined
to identical products, and included prices from rural as well as urban areas. In
any case, for 2011, in common currency units, the ICP calculations implied that
the prices of all goods consumed by households were 42.8% lower in Indonesia
than in Singapore. The price differential was 47.1% for non-durables, 37.9% for
semi-durables, and 34.4% for durables.
In the present study, among the commodities subject to NTMs in Indonesia
were horticultural products such as carrots (for which the adjusted price differen-
tial used as an estimate of the NRP was 67.3%), mangoes (76.0%), oranges (98.0%),
tomato juice (79.5%), apple juice (49.6%), orange juice (9.9%), sausages (26.9%), and
sparkling mineral waters (39.5%).
I noted earlier that, with restrictive policies in effect, chicken meat was 29.7%
more expensive in Indonesia in recent years than in earlier years.6 I have consid-
erable confidence in the methodology used to calculate that figure, and so used
it as the estimate of the NRP for chicken meat and products. Nevertheless, the
adjusted retail price comparisons between Jakarta and Singapore revealed an
important nuance of the effects of Indonesia’s chicken import restrictions: the
relatively less attractive parts of the chicken by Western standards were over-
priced in Indonesia (leg quarters by 40.2%, necks by 49.4%, livers by 58.9%, and
gizzards by 64.0%), while the relatively more attractive parts were underpriced
(boneless breasts by 12.1% and wings by 24.8%). Whole chickens were overpriced
in Indonesia as well (village chickens by 37.3%, whole broilers by 6.0%). Given the
openness of the Singapore market, these figures are consistent with the poultry
market in Indonesia being isolated from global commerce. Cross-national differ-
ences in tastes may exist, but even so price differences would not persist in the
absence of barriers to trade. These barriers could include transportation costs,
although market participants downplayed this consideration for fresh food.
Among electronic and electrical products, laptop computers (adjusted price
differential of 16.2%), microwave ovens (27.2%), and video game consoles and
equipment (22.4%) were subject to procedural trade impediments in Indonesia,
and optical media (31.2%) to mandatory import licensing. By far the most eco-
nomically significant adjusted price premiums among durables in Jakarta relative
to Singapore were the 49.0% premium for smartphones and the 63.7% premium
for tablet computers. These figures are based on a great many price comparisons,
especially for smartphones, which all showed similar tendencies. No doubt these
premiums reflect the strict telecommunications import restrictions in Indonesia.7
For other commodities, the adjusted price differentials were not so large. In
particular, apparel and personal care products, subject to various procedural

6. The external reference price for chicken meat is a US East Coast port wholesale price
from the World Bank Pink Sheet (http://www.worldbank.org/en/research/commodity-
markets). The Indonesian price data are from the Ministry of Trade.
7. To some extent, the existence of a market for used cellular telephones may dampen the
effect of these price differentials on lower-income consumers.
Non-tariff Trade Regulations in Indonesia 339

trade impediments in Indonesia, were generally less expensive in Jakarta than

in Singapore.
I also carried out retail price comparisons for four construction products subject
to NTMs in Indonesia. Given that these products were largely sold in bulk transac-
tions rather than through retail outlets, I omitted the adjustment for distribution
margins and corrected only for commodity tax differences between Indonesia
and Singapore. I estimated the NRPs for cement (12.7%), an imported product in
Indonesia, and, among exports, for construction sand (−56.7%) as well as smaller
(−76.2%) and larger (−65.8%) split granitic rocks used in concrete.

Comparison of Domestic Prices and Border Prices

Sugar and rice have long been subject to NTMs in Indonesia in the form of quan-
titative restrictions on imports, in addition to import tariffs. For sugar, detailed
data are available from the Ministry of Trade on the costs and profit margins that
come between the border price and the retail price within Indonesia. I used these
data—with a few modifications and updates—to estimate an NRP for sugar of
54.8% in May 2015.
For milled, polished rice, the World Bank Jakarta office has compiled data on the
price of medium-quality rice at the Cipinang wholesale market and on the border
price for 15% broken long-grain rice from Vietnam. These data imply an NRP for
rice of 67.8% on average over the first six months of 2015.8
For rice and sugar, there is the question of how much of the domestic price
premium has been enjoyed by farmers. Under the sharing system between state-
owned sugar factories and independent sugar-cane farmers on Java, farmers are
supposed to receive 65% of the value of the sugar content of their cane. This implies
that, in principle, sugar cane should enjoy the same rate of protection as sugar. In
practice, however, farmers do not know the exact sugar content of their cane and
find it difficult to protect their interests vis-à-vis the sugar factories. If the share
of the value of the sugar-cane content going to farmers were to remain constant,
whether at 65% or less, the NRP for cane would still be 54.8%. It is only if the share
going to farmers goes down, in the presence of a rising domestic price of sugar,
that the NRP for cane would be less than that figure, but this seems plausible: the
higher the price of sugar, the stronger the incentive for sugar mills to cheat the
farmers, all else being equal.
In addition, the retail price of sugar could fluctuate independent of the auction
price at which sugar is sold at the state-owned sugar factories. In the absence of
quantitative information on these issues, I assumed that the 54.8% NRP for sugar
also accrued to sugar cane, but it may be viewed more properly as an upper bound
for the NRP for the latter.

8. This figure is based on the conservative assumption that freight and insurance costs
were $35 per tonne; under the assumption of $25 per tonne, the average NRP was 72.3%.
The Ministry of Trade also provides data on the retail and border prices of rice, based on
input from the National Logistics Agency (Bulog). However, there is a mismatch between
the quality of local rice for which retail price data exist, and the quality of imported rice,
so the application of similar methods of computation yielded an average NRP over the
first six months of 2015 of only 64.3%.
340 Stephen V. Marks

There were similar issues for rice: wholesale price fluctuations may not have
been passed on to growers of field rice, and the scales used by some rice millers
may not have been accurate. I similarly assumed that the NRP for field rice grown
by farmers was identical to that at the wholesale level for milled, polished rice, but
in practice there may have been a divergence.
Alcoholic beverages are now subject to quantitative import restrictions as well as
specific tariffs and excise taxes. The availability of data on prices for unregulated
and untaxed sales by the embassies of Australia and the United States in Jakarta pre-
sented an opportunity to estimate the border prices of these beverages. I used data
on retail prices in Jakarta to compare domestic producer prices with these border
prices, and thus estimate the NRPs, based on certain straightforward assumptions.
Among the alcoholic beverages for which I obtained NRP estimates were beers
and ales (119.4%), sparkling wines (130.3%), other wines (182.1%), gins (285.1%),
vodkas (283.8%), liqueurs and cordials (72.9%), and rums (72.1%). Imported rum
was unique in that the NRP was less than the sum of the import duty plus the
excess of the excise tax on imports over the excise tax on domestic output.

Imputation of Import Tariff and Export Tax Equivalents

Price data were not available for many imported and exported commodities sub-
jected to major new quantitative trade restrictions—notably imported horticultural
and livestock products and exported mineral products. For many of these items,
however, it was possible to use simulation methods to impute import tariff equiv-
alents or export tax equivalents of the NTMs. In effect, percentage changes in
quantities traded were translated into percentage changes in prices under certain
assumptions.
The analysis of mineral export restraints offers a clear illustration. Many unpro-
cessed and semi-processed mineral exports were banned or sharply restricted at
the start of 2014. Table 2 shows the value of exports of six major metallic ores and
concentrates from 2011 to 2014. It uses 2011 as the base year for the simulations
because export taxes of 20% were imposed on these products in 2012. The per-
centage change in the value of these exports between 2011 and 2014 ranged from
−62.5% for zirconium to −100.0% for manganese. The table also shows the estimated
export tax equivalents for the six ores and concentrates.9
In the simulations, the percentage change in the quantity exported between
2011 and 2014 was used to impute the export tax equivalent of the quantitative
restriction, based on an estimate of the price elasticity of export supply from
Broda, Limão, and Weinstein (2008).10 I deflated the border price of the commodity
by a broad international price index, but assumed that changes in the amount of
exports from Indonesia were too small to affect the external price. The border price
was proxied by the unit value of exports, which in some cases I replaced by the

9. For all of the commodities in table 2, I calibrated the data to an exponential inverse export
supply curve. This formulation allowed exports to be driven to zero by an implicit export
tax of less than 100%, unlike with a constant-elasticity export supply curve.
10. The export-supply elasticity data were not for Indonesia in particular, but for coun-
tries that exported to the United States (or in some cases to China, if data for US partners
were unavailable or problematic). Missing elasticity estimates for some products had to
be replaced by estimates for similar products.
Non-tariff Trade Regulations in Indonesia 341

TABLE 2  Imputed Export Tax Equivalents of Restrictions on

Exports of Selected Metallic Ores and Concentrates, 2011–14

Exports Change in Export

($ million) exports, tax
2011–14 equivalent
Description 2011 2012 2013 2014 (%) (%)

Copper 4,700.4 2,594.7 3,006.8 1,683.6 −64.2 28.7

Nickel 1,428.0 1,489.1 1,685.2 85.9 −94.0 15.4
Aluminium 773.2 626.0 1,349.7 46.4 −94.0 39.7
Iron 342.6 251.0 426.8 57.5 −83.2 25.8
Manganese 17.1 3.1 1.0 0.0 −100.0 92.2
Zirconium 56.7 81.6 42.8 21.3 −62.5 72.9

unit value of US imports or exports of the relevant commodity, if the Indonesian

data appeared suspect.
Tin ores and concentrates were subject to an earlier export ban, and exports
have officially been zero for some years. Application of a similar methodology for
these commodities with 2002 as the base year and 2004 as the policy year yielded
an export tax equivalent of 56.8%. This estimate may be too high, however, as the
effects of the policy have probably diminished over time as downstream process-
ing capacity has been built within Indonesia; if the export ban were to end, almost
certainly exports would not return to their pre-ban level.
The assumption that exports from Indonesia do not affect the external price
is questionable for nickel, tin, and aluminium ores and concentrates, given that
Indonesia produced more than 34% of the global output of nickel ore, 28% of tin
ore, and 18% of aluminium ore (bauxite) in 2013 (British Geological Survey 2015).
According to commodity price data from the World Bank, the nominal world
market price of nickel increased by more than 37% during the first five months of
2014, though most of the price increase dissipated over the next five months and by
March 2015 the nominal price was lower than it had been in January 2014. Increases
in the nominal prices of aluminium in 2014–15 and tin in 2004–06 may partly have
been related to export bans in Indonesia, although the price increases were mostly
reversed within a matter of months. In any case, had the world prices of certain ores
and concentrates increased as a consequence of the export bans—had downward-
sloping world demand for these exports been properly accounted for—the price
effect estimated from the simulations would have been even higher, so the NRP
estimates used in this article may be viewed as conservative in this respect.
I used a similar procedure to compute import tariff equivalents, based on the
drastic cutbacks in imports of some fresh and processed fruits and vegetables
caused by the horticultural import regulations enacted in 2013, and the fall in live-
stock and livestock product imports following the regulations enacted in 2011–13.11
I used detailed import demand elasticity parameter estimates for Indonesia from

11. For horticultural products, 2011 was the base year and 2014 the policy year; for livestock
products, 2010 and 2013 were used.
342 Stephen V. Marks

Kee, Nicita, and Olarreaga (2008) to do the simulations. Among the NRPs esti-
mated on this basis that I used in the calculation of effective rates of protection
(ERPs) were those for breeding cattle (21.8%), beef offal (130.6%), bananas (11.9%),
plantains (12.9%), pineapples (123.3%), mangoes (52.9%), oranges (12.5%), manda-
rins (38.2%), grapefruits (101.6%), durians (71.2%), longans (30.0%), pickled onions
(48.4%), frozen potatoes (153.1%), potato chips and sticks (138.8%), citrus preserves
and jams (22.9%), pineapple preserves (24.0%), longan preserves (50.0%), grapefruit
juice (51.9%), apple juice (14.9%), apple juice concentrate (49.9%), other juices (13.9%),
mixtures of juices (79.0%), and chilli sauce (28.4%).12

Other Quantitative Restrictions

I used price data from published sources and interviews with market participants
to calculate NRPs for a few other commodities. For example, field work conducted
by the Ministry of Trade (2012) allowed me to estimate the NRP from the complete
ban on exports of unprocessed and semi-processed rattan at −29.7%. Interviews
with knowledgeable market participants also revealed that restrictions on exports
of coal and on imports of cloves were not binding as of early 2015, and so the
associated NRPs were set to zero. Import restraints on salt were assumed to be
ineffective based on news reports.

Procedural Burdens on Importers and Exporters

As noted in table 1, regulations that impose procedural costs on importers or
exporters are among the NTMs enacted in recent years. In many cases I did not
need to estimate the costs of these procedural regulations, because the costs were
reflected in the price differentials estimated by other methods. Consider, for
example, horticultural imports. The impacts of the various procedural require-
ments—pre-shipment inspections in the country of origin, restrictions on the ports
that could be used in Indonesia (fruit and vegetables sold in Jakarta had to come
through Surabaya, for example), requirements that importers obtain permission for
shipments months in advance, and requirements that importers sell to distributors
rather than directly to retailers or consumers—were all accounted for in the price
differential calculations for these products.
Interviews with market participants allowed me to estimate the costs of other
procedural measures—not only direct monetary costs but also time costs, which
the interviewed parties tried to account for in monetary terms. For example, the
Ministry of Trade requires pre-shipment inspections, imposes certain port restric-
tions, and specifies special trader-registration procedures for imports of a wide
variety of consumer products. Based on detailed data from a large importer, I
estimated pre-shipment inspection costs to be 0.35% of the border price for the
imported items. Similarly, the Ministry of Agriculture now has quarantine require-
ments for virtually all products that contain plant or animal materials, including
the packing materials in which other imports are shipped. Based on interviews

12. To remain conservative, I used the figures from the simulations rather than those
obtained by retail price comparisons between Jakarta and Singapore, namely 12.5% rather
than 98.0% for oranges; 14.9% rather than 49.6% for apple juice; and 52.9% rather than 76.0%
for mangoes.
Non-tariff Trade Regulations in Indonesia 343

with market participants, I assumed that these costs were 1.5% of the value of the
shipment, though no doubt there is considerable variation across products, since
quarantine requirements vary by type of product. In addition, a container with
one type of fruit could incur lower quarantine costs than a mixed shipment of
several types of fruits.
Quantifying the costs of these mandated procedures is rather arbitrary because
the costs are often fixed, so that larger shipments incur lower costs per unit. For
example, the Ministry of Industry imposes the Indonesian National Standards on
an array of products, resulting in some cases in a requirement for advance testing
and clearance of products prior to shipment. A major importer of baby clothes put
the cost of these regulations at 15% of the value per shipment, because the fashion
season is short and shipments are relatively small.
Exports of coal, crude oil, crude palm oil, palm kernel oil, and an array of min-
eral products are subject to a Ministry of Trade requirement that the exports be
financed by letters of credit. The lowest estimate of this cost from industry insiders
was 0.215% of the border price, and that is the figure I used.
Finally, Indonesia offers seven preferential import tariff schedules, in addition
to the most-favoured-nation (MFN) schedule that it offers to members of the World
Trade Organization. Based on information from a major importer, I estimated the
documentation cost to obtain the preferential duty rates to be 0.53% of the border
price.

Commodity Subsidies
Various targeted commodity subsidies applied within Indonesia entail both
domestic and foreign trade restrictions. For example, in 2015 fertiliser subsidies
were intended to apply only to small farmers growing rice, corn, or soybeans—not
to other types of farmers, agribusiness, industry, or the fertiliser export market.
Among petroleum products, kerosene and diesel fuel are officially subsidised, as
is biodiesel fuel derived from palm oil, but low-octane gasoline and traditional
diesel fuel are unofficially subsidised in the form of losses incurred by the state
oil company, Pertamina, through selling at below-market prices.
Table 3 shows the estimated subsidy rates for the major subsidised commodities,
based on comparisons of their subsidised and unsubsidised prices. The analy-
sis that follows assumes that these subsidies were enjoyed only by the officially
targeted users, but no doubt there have been leakages of subsidised products to
other groups.

AGGREGATION ISSUES
The NRPs I have estimated apply at the level of tariff lines, but then must be
aggregated for the 140 tradable-goods sectors in the 175-sector input–output table
for 2005 in order to be incorporated into the calculation of ERPs.
This study aggregates rates of protection using free-trade tariff-line trade value
weights for each of the input–output sectors. A well-known issue with trade
weights is that the magnitude of the aggregated NRP may be biased towards zero,
to the extent that the trade weights have been affected by the policies in question.
For example, a prohibitive import tariff will be excluded from the aggregated
total, even though the NRP may be very high, because the import value is zero.
344 Stephen V. Marks

TABLE 3  Subsidy Rates for Various Commodities, Early 2015 (%)

Commodity Subsidy

Fertilisers
Urea 65.4
SP-36 (super phosphate) 63.6
ZA (ammoniated zinc) 46.1
NPK (nitrogen–phosphorus–potassium mix) 61.0
Organic fertiliser 53.4
Oil & gas
Kerosene 81.8
LPG in 3 kg containers 63.0
Diesel motor fuel 34.7
Biodiesel motor fuel 46.4
Gasoline (low octane) 17.8

Note: All rates are based on prices that applied some time between January and March 2015, except for
the biodiesel rate, which is based on the price in July 2015, when a new biofuel policy came into effect.

However, the trade data used in this study are from 2012, before the application of
many of the recent policies. For some commodity groups with NRPs that are high
in absolute value—notably meats, sand and related materials, and rattan—I have
replaced the 2012 trade weights with data from other years in which trade was
relatively unimpeded. The procedure is inelegant but effective.
For each tariff line, exports are used to weight the NRPs on the export side, and
imports are used to weight the NRPs on the import side. These figures are then
combined. This makes sense if exportable and importable commodities within
each tariff line are not substitutable with each other. For a commodity with a
high NRP on the import side, the presence of substantial exports will dilute the
overall NRP for the tariff line, given that all commodities have a zero or negative
NRP on the export side.13
Calculation of ERPs, to be defined below, requires a comprehensive database of
all trade and trade-related policies, not just the NTMs of particular interest. This is
complicated by the proliferation of tariff-cutting preferential trade agreements in
recent years—since ERP calculations require a single NRP for each sector. I have
resolved this issue by forming a composite of the various applicable import tariff
rates, which I call the marginal rate of protection (MRP). The MRP is defined as
the highest applicable import tariff rate (inclusive of anti-dumping and safeguard
duties, as well as procedural NTMs that add to the costs of importation) such that
imports from the particular region of origin are positive. Under the assumption
that imports of a given commodity from various countries are perfectly substitut-
able with each other and with the domestic import-competing product (standard

13. For example, the NRP for field rice (sector 1) in table 5 below is shown to be 67.2% rather
than the 67.8% reported for milled, polished rice (sector 57), due to the exportation of small
amounts of field rice from Indonesia.
Non-tariff Trade Regulations in Indonesia 345

assumptions in ERP analyses), but that the quantity of imports from the various
source countries is limited, the MRP will set the domestic price.14
Finally, if a quantitative restriction on imports results in a domestic–external
price differential that exceeds the MRP in magnitude, then the quantitative restric-
tion is treated as binding and the import NRP is taken to be that price differential
in percentage terms. A similar approach is used on the export side.

EFFECTIVE RATES OF PROTECTION

The effective rate of protection for a tradable item is defined as the percentage by
which value added per unit of output with trade interventions present exceeds
the level under free trade. Value added in a sector is defined as income to the pri-
mary factors of production (such as labour and capital) in that sector. It can also be
calculated as the difference between the value of output and the cost of produced
inputs used by the sector. Thus, a positive ERP for a sector indicates that policies
on balance raise the domestic price of output relative to the prices of the goods
and services used to produce it. The ERP can also be negative, indicating that on
balance policies raise the prices of inputs used by a sector relative to the price of
its output. Under standard assumptions—notably that input–output coefficients
are constant—the ERP shows the effects of the policies examined on the incentives
to produce more of a good.
Additional assumptions in ERP analyses are that the country is too small to
affect world market prices and, as noted earlier, that domestic and imported prod-
ucts are perfect substitutes. Under these conditions, the domestic price of each
tradable commodity is determined directly by its external price and the NRP for
that commodity.
All ERP studies must handle the problem of non-tradable items such as ser-
vices, however, since the prices of such items are determined within the domestic
economy. This article uses two approaches: the Humphrey (1969) method and the
Corden (1966) method.
The Humphrey method estimates the effects of the structure of protection on
the prices of non-tradables from the cost side. Increases in the prices of tradable
inputs caused by trade policies add directly to the unit costs of production of
non-tradables. In addition, value added in non-tradable sectors is separated into
the part that accrues to labour and the part that accrues to other primary factors
of production. Value added per unit from other factors is assumed to remain con-
stant, but money wages are assumed to adjust so as to hold real wages constant,
given the effects of trade policies on the prices of tradables and non-tradables,
so that value added per unit attributable to labour increases. This provides a
second channel by which trade policies boost the prices of non-tradable inputs.
A useful by-product of the Humphrey method is a cost-of-living index that sum-
marises the effects of the structure of protection on the prices of all tradables
and non-tradables.

14. Marks (2011) and Marks and Rahardja (2012) also use this method, and contrast it with
the average rate of protection, which simply calculates a weighted average import tariff
rate but is not as consistent in theoretical terms with ERP analysis.
346 Stephen V. Marks

The Corden method obviates the need to estimate the prices of non-tradable
items. Value added in a tradable sector is redefined to include not only the value
added contributed by primary factors in the sector itself, but also the value added
from the production of non-tradable inputs used directly or indirectly by the sector.
To implement this approach, one must go back through the input-output table to
decompose the prices of all non-tradable inputs into the prices of tradable inputs
and the value added of primary factors that contribute directly or indirectly to
the production of these non-tradable inputs.15

Included Policies
The ERP analysis accounts for the effects of the full structure of trade policies:
MFN and preferential import tariffs, anti-dumping and safeguard duties, specific
tariffs set in rupiah per unit rather than as a percentage of the unit price, and the
various NTMs and export levies noted in table 1.16 The analysis also accommodates
the commodity subsidies discussed above, as well as excise taxes on alcohol and
tobacco.
These subsidies and excise taxes do not affect the calculation of NRPs and thus
producer prices, but they do affect the estimation of ERPs, or more properly effec-
tive rates of assistance, by creating a differential between the producer price and
the user price, for users that are eligible for a subsidy or subject to an excise tax
on a given item.
I include two additional policies not discussed so far. The first is duty drawbacks
and exemptions for imported intermediate inputs and raw materials, which can
be claimed by exporters of manufactured goods in Indonesia. This policy applies
to all forms of import duties, including anti-dumping and safeguard duties. Duty
exemptions are one of the features of bonded customs zones within the country,
but duty drawbacks are available outside these zones.
Fane and Condon (1996) developed a simple approach to include these measures
at the level of input–output sectors in their ERP analysis; Marks and Rahardja
(2012) used a variant of that approach to study the duty drawbacks and exemp-
tions in effect in early 2008, as I do now as well. I prefer to understate rather than
overstate the effects of the measures. Thus, I winnowed in three ways the number
of sectors whose outputs would be eligible for duty drawbacks as inputs into
downstream sectors. First, I excluded sectors that primarily produced final goods
(capital goods and transportation equipment in particular), as per the regulations.
Second, I omitted sectors in which quantitative trade restrictions were binding.
Third, because the trade weighting method used in this article dilutes the aver-
age import tariff rate with the average export tax (the latter being equal to zero

15. In previous studies of Indonesia, Fane and Philips (1991) used the Humphrey method,
Fane and Condon (1996) used the Humphrey and Corden methods, and Warr (1992) used
the Corden and Balassa (1982) methods. The latter is similar to the Humphrey method but
excludes the wage effects.
16. Specific import duties in rupiah per minute for cinematographic film are replaced by an
approximate estimate (20%) of their ad valorem equivalent from the Ministry of Finance.
The need to calculate ad valorem equivalents for specific tariffs on sugar, rice, and alcoholic
beverages is obviated by the use of price comparisons to estimate the effects of trade poli-
cies, given the presence of quantitative restrictions on imports of these products as well.
Non-tariff Trade Regulations in Indonesia 347

for most sectors), I excluded sectors in which the value of imports was less than
the value of exports. This left the outputs of 32 of the 140 tradable-goods sectors
eligible for duty drawbacks as inputs into other sectors.17
Fane and Condon (1996) assumed that exporters of manufactured goods received
80% of the duty drawbacks and exemptions to which they were officially entitled.
I made the same assumption for comparability, but this policy parameter merits
empirical study.
The other policy is within-sector tariff escalation. Tariff escalation exists if
import tariffs are set at higher rates on downstream products than on upstream
products. In Indonesia, the MFN import tariff rate on most finished footwear
(sector 83) is 25% or 30%, while footwear parts have an MFN import tariff rate of
5%. MFN tariff rates are 50% on fully assembled motor vehicles (sector 133), but
typically only 10% for completely knocked-down kits ready for assembly and for
most vehicle parts. There are similar differentials for motorcycles (sector 134), for
which the highest MFN tariff rate is 40%, versus 10% for kits and parts.
These differences in import tariff rates in effect provide an input subsidy for the
sectors on their own outputs. It turns out that, because most finished footwear is
exported, the NRP for footwear inputs is actually higher than the NRP for foot-
wear outputs, and both are low, so I did not include this effect for footwear. For
motor vehicles, however, the sector effectively receives a subsidy of 16.4% relative
to other sectors that use motor vehicles, and I did include that. I similarly included
an effective subsidy of 12.2% for motorcycles relative to other sectors that use
motorcycles.

Findings of the Analysis

In order to reveal the effects of the major NTMs, I calculated ERPs for two sce-
narios: one in which all of the trade policies applied, and one in which quantitative
restrictions were excluded. All other trade policies, commodity subsidies, and
excise taxes were applied in both scenarios.18
To provide an overview, table 4 shows the NRPs, ERPs using the Humphrey
method (ERPH), and ERPs using the Corden method (ERPC), after combining the
input–output sectors to three different levels of aggregation using value-added
weights.19 The upper section shows 17 tradable-goods sectors, the lower one
focuses on five broader sectors, and the bottom row shows the calculations for
tradable-goods sectors overall.

17. All manufacturing sectors were treated as eligible users of duty drawbacks and exemp-
tions. Of the sectors in table 5, only the products of sectors 2, 6, 7, 11, 12, 16, 17, 27, 42, 46,
47, 59, 67, 68, 69, 72, 76, 81, 92, 94, 97, 98, 99, 103, 104, 108, 109, 114, 115, 116, 122, and 129 were
treated as eligible inputs.
18. As noted earlier, quantitative restrictions include policies such as import or export
licensing, quotas, and bans. Even though the domestic subsidy policies implicitly entail
non-tariff trade restrictions because of their targeted nature, these policies are treated
more as subsidies than as quantitative restrictions.
19. Fixed value-added weights based on free-trade value added calculated for 2008 using
the Humphrey method are used for all three sets of aggregations in the table. With variable
weights, the aggregation would mask the sharp increases in rates of protection between
2008 and 2015. Free-trade value added is the appropriate weight when working with ERPs.
 TABLE 4  Value-Added-Weighted Nominal and Effective Rates of
Protection, Three Levels of Aggregation, 2015 and 2008 (%)

2015 2008

All trade policies Quantitative All trade policies

in effect restrictions excluded in effect

Description NRP ERPH ERPC NRP ERPH ERPC NRP ERPH ERPC

Food crops 27.7 64.3 51.2 6.5 15.9 14.4 12.5 27.2 23.6
Estate & other crops 2.0 2.5 2.5 −0.4 −1.1 −0.6 −0.3 10.9 9.5
Livestock & their
products 8.8 12.0 11.2 3.4 5.4 4.9 0.9 1.4 1.3
Forestry −4.2 −5.0 −4.4 −2.9 −3.4 −3.1 −4.0 −2.5 −2.3
Fisheries 0.3 0.8 0.8 0.3 0.9 0.9 0.2 4.2 4.1
Oil & gas extraction 0.1 0.0 0.0 0.2 0.2 0.2 0.6 0.7 0.7
Other mining −22.1 −25.6 −24.1 0.7 0.3 0.4 −13.1 −14.4 −13.7
Food, beverages &
tobacco 10.6 19.6 13.9 1.4 2.3 1.7 3.0 3.5 2.7
Textiles, apparel &
leather 3.7 4.1 3.4 3.6 5.4 4.0 0.7 1.5 1.3
Wood products 0.6 0.8 1.2 0.5 0.8 0.8 −0.1 1.7 1.5
Paper products 1.4 1.7 1.9 1.4 2.5 2.1 0.7 2.3 1.9
Chemicals 3.2 33.0 24.2 3.0 8.8 6.7 1.7 25.1 18.0
Oil refining & LNG 0.4 0.6 0.7 0.5 0.7 0.7 0.2 −0.4 −0.4
Non-metal products 4.7 16.3 11.2 3.6 8.3 5.6 2.0 8.6 6.2
Metals & metal products 6.7 65.1 36.1 7.3 18.7 11.9 3.3 10.5 7.7
Machinery & transport
equipment 7.7 13.4 10.6 5.6 8.8 7.1 4.7 9.2 7.3
Other manufacturing 4.3 11.1 8.5 4.0 7.2 5.4 2.1 6.5 4.8
                   
Agriculture (excluding
forestry, fisheries) 18.6 41.1 33.1 4.4 10.2 9.3 7.5 19.0 16.5
Forestry & fisheries −1.0 −0.9 −0.7 −0.6 −0.3 −0.2 −1.0 2.3 2.3
Mining & quarrying
(excluding oil & gas) −22.1 −25.6 −24.1 0.7 0.3 0.4 −13.1 −14.4 −13.7
Oil & gas extraction
& refining 0.2 0.3 0.3 0.3 0.4 0.4 0.4 0.2 0.2
Manufacturing
(excluding oil & gas
refining) 6.6 17.8 12.3 3.4 6.3 4.6 2.7 6.7 5.1
                   
Overall 3.7 12.1 8.5 2.3 4.6 3.8 1.1 4.8 3.8

Note: NRP = nominal rate of protection; ERPH = effective rate of protection using the Humphrey
method; ERPC = effective rate of protection using the Corden method.
Non-tariff Trade Regulations in Indonesia 349

The first set of columns for 2015 in table 4 shows the calculations with all of the
trade policies studied in effect, while the second set excludes quantitative restric-
tions. (The set for 2008 will be discussed later.) Table 5 shows similar estimates for
each of the 140 tradable-goods input–output sectors, but only for 2015. (Note that
sector 34 does not produce tradable goods and thus is not included in the table.)
Table 4 indicates that trade policies transfer resources out of natural-resource-
intensive sectors—minerals and forestry—into manufacturing and especially
agriculture. Part of the story in political economy terms could be that the govern-
ment seeks to promote manufacturing employment, but also to support incomes
in agriculture, which have come under pressure due to relatively high productivity
growth in manufacturing. The International Labour Organization, for example,
notes that productivity per worker in industrial sectors in Indonesia is quadruple
that in agriculture (ILO 2015).
Table 4 shows that, in agriculture, the major beneficiaries of trade and trade-
related policies have been food crops. Among manufacturing sectors, the primary
beneficiaries of trade and commodity policies have been metals and metal prod-
ucts; chemicals; and food, beverages, and tobacco.
The first set of columns in table 5 shows that the protection of food crops is
concentrated in field rice (sector 1), fruits (sector 10), and vegetables (sector 9). The
impact of policies on estate crops is mixed: sugar cane (sector 13) is aided by the
protection of sugar, but oil palm (sector 15) and cocoa (sector 21) are hit by export
taxes. The protection for food, beverages, and tobacco is concentrated mostly in
just a few sectors—milled, polished rice (sector 57), alcoholic beverages (sector 70),
sugar (sector 62),20 wheat flour (sector 58), and processed tobacco (sector 72). In the
metals and metal products sectors, ERPs are particularly high in basic iron and
steel (sector 115) and other metal products (sector 122).
Table 5 shows that rates of protection do not necessarily escalate from upstream
to downstream. Despite the relatively high NRP for livestock and their products
(sector 25), for example, meat and viscera (sector 49) is well protected, but the pro-
cessed and preserved meats sector (sector 50) has negative ERPs. Similarly, wheat
flour (sector 58) is well protected, but bakery products (sector 60) and noodles
(sector 61) have negative ERPs. Although textile sectors such as spinning (sector
75) and weaving (sector 76) have positive NRPs and ERPs, labour-intensive apparel
production (sector 79) has negative ERPs.
Similarly, while basic iron and steel (sector 115) has a fairly high NRP from
protective import tariffs on steel products, the consequences of these tariffs are
borne by various downstream sectors whose NRPs are low and whose ERPs are
low or even negative—such as metal tools (sector 119), turbines and engines (sector
123), shipbuilding (sector 131), and railroad equipment (sector 132).21 With high
NRPs, motor vehicles (sector 133) and motorcycles (sector 134) are exceptions to

20. Sector 62 includes all sugars other than cane sugar, so that the NRP is diluted from the
54.8% estimated for cane sugar and sugar cane to 48.1% for all sugars.
21. Restrictions on imports of used capital equipment may be used selectively to protect
sectors such as shipbuilding. These restrictions are opaque, and have given rise to com-
plaints from industries ranging from cocoa processing to construction about the high costs
of importing new capital equipment or spare parts. These restrictions deserve further study.
 TABLE 5  Nominal and Effective Rates of Protection, 2015 (%)

All trade policies Quantitative restric-

in effect tions excluded

Sector Description NRP ERPH ERPC NRP ERPH ERPC

1 Field rice 67.2 204.3 152.4 8.4 35.6 30.9

2 Corn 6.2 27.9 25.1 4.8 25.9 23.1
3 Cassavas 0.5 0.2 0.3 0.1 0.0 0.1
4 Sweet potatoes 0.0 −0.1 0.0 0.0 0.0 0.0
5 Other tubers 1.3 1.2 1.3 1.1 1.1 1.1
6 Peanuts 6.4 6.8 6.6 4.9 5.3 5.1
7 Soybeans 1.5 10.8 10.1 0.0 9.0 8.3
8 Other beans & nuts 1.9 1.8 1.8 1.4 1.4 1.4
9 Vegetables 19.0 21.8 21.2 8.1 9.2 8.9
10 Fruits 24.5 27.2 26.4 7.9 8.6 8.4
11 Other grains 1.5 1.5 1.6 0.0 −0.1 −0.1
12 Rubber 4.6 5.6 5.4 3.5 4.3 4.1
13 Sugar cane 54.8 83.3 68.7 13.9 18.5 16.3
14 Coconut 0.1 −0.1 0.1 0.1 −0.1 0.0
15 Oil palm −8.5 −13.3 −10.4 −8.5 −13.2 −10.5
16 Fibre plants 2.0 2.0 2.1 0.5 0.5 0.5
17 Tobacco 5.2 9.7 8.0 4.0 7.8 6.2
18 Coffee 0.5 0.3 0.6 0.4 0.4 0.4
19 Tea 0.0 −0.2 0.0 0.0 −0.2 −0.1
20 Cloves 5.3 6.1 6.0 4.1 4.7 4.5
21 Cocoa −3.4 −4.3 −3.9 −3.6 −4.5 −4.2
22 Cashews 0.0 −0.2 −0.1 0.0 −0.2 −0.1
23 Other estate crops 0.7 0.1 0.5 0.5 0.0 0.2
24 Other agriculture 0.5 −0.2 0.0 0.4 0.2 0.2
25 Livestock & their products 13.4 16.3 15.3 0.6 0.4 0.4
26 Fresh milk 2.4 1.3 1.4 1.9 2.0 1.9
27 Poultry & its products 6.5 10.0 9.2 5.0 8.3 7.5
28 Other livestock products 2.9 1.4 1.5 1.4 1.1 1.1
29 Wood −2.5 −3.0 −2.6 −2.3 −2.7 −2.4
30 Other forest products −10.3 −12.1 −11.1 −5.1 −6.1 −5.6
31 Marine fish & products 0.5 1.6 1.6 0.5 1.7 1.7
32 Freshwater fish & products 0.1 −0.4 −0.3 0.1 −0.1 −0.1
33 Shrimp & prawns 0.1 −0.4 −0.2 0.1 −0.1 0.0
35 Coal −0.2 −0.7 −0.4 0.0 −0.4 −0.2
36 Crude oil 0.1 0.1 0.1 0.3 0.2 0.2
37 Natural gas & geothermal −0.2 −0.3 −0.3 0.0 −0.1 0.0
38 Tin ore −56.8 −58.3 −57.6 0.0 −0.2 −0.1
39 Nickel ore −15.4 −16.3 −15.9 0.0 −0.2 −0.2
40 Bauxite −39.7 −50.6 −45.9 0.0 −0.8 −0.5
41 Copper ore −27.6 −35.2 −32.2 0.0 −1.1 −0.8
42 Gold ore 0.5 0.3 0.4 0.5 0.3 0.3
43 Silver ore −90.1 −91.7 −91.0 0.0 −0.4 −0.3
44 Iron ore & sand −14.0 −19.5 −16.8 0.2 −0.2 0.0
45 Other mined metal products −44.6 −46.5 −45.7 1.5 1.5 1.5
46 Non-metallic minerals 1.2 1.0 1.3 1.2 1.0 1.0
47 Coarse salt 9.9 11.1 10.8 9.9 11.1 10.8
48 Other mined products −36.0 −41.3 −39.0 2.2 2.2 2.1
 TABLE 5 (cont.)

All trade policies Quantitative restric-

in effect tions excluded

Sector Description NRP ERPH ERPC NRP ERPH ERPC

49 Meat & viscera 37.4 116.5 81.5 4.8 9.4 7.8

50 Processed & preserved meat 8.1 −12.2 −6.6 2.0 −1.1 −0.3
51 Milk products 6.2 2.7 2.5 4.8 10.4 7.2
52 Processed fruits & vegetables 11.2 8.3 6.9 2.3 −1.3 −0.8
53 Dried & salted fish 1.7 2.6 2.5 0.3 −0.2 0.1
54 Processed & preserved fish 0.4 −1.7 −0.3 0.2 −1.1 −0.4
55 Copra 0.0 −1.1 −0.2 0.0 −0.6 −0.2
56 Edible oils −6.1 −8.0 −6.4 −6.0 −7.5 −6.2
57 Milled, polished rice 67.8 103.6 75.2 8.8 12.5 10.4
58 Wheat flour 13.6 47.3 42.2 3.1 9.6 8.9
59 Other flour 3.7 −10.9 −6.9 2.8 1.8 1.5
60 Bread, biscuits & related products 2.7 −18.7 −13.4 2.6 −1.1 −0.5
61 Noodles & related products 1.1 −15.3 −12.1 1.1 −3.1 −2.3
62 Sugar 48.1 87.6 55.4 12.5 18.0 13.2
63 Processed seeds & nuts 1.9 −7.4 −5.8 1.3 −0.8 −0.5
64 Chocolate & candies 2.4 −7.0 −5.4 2.3 0.4 0.5
65 Milled coffee 3.7 5.5 4.8 3.6 6.6 5.2
66 Processed tea 4.0 7.1 6.0 3.7 7.2 5.8
67 Soybean products 2.2 −3.3 −2.2 0.7 −1.3 −0.9
68 Other food products 4.8 −0.7 0.2 4.2 13.6 9.3
69 Animal feeds 1.8 −3.6 −1.9 0.6 −2.7 −1.7
70 Alcoholic beverages 53.7 245.6 103.6 39.9 179.1 83.9
71 Other beverages 11.5 11.7 8.6 7.0 11.8 8.1
72 Processed tobacco 9.7 23.1 17.0 8.4 20.9 15.1
73 Cigarettes & related products 0.8 −1.1 −0.6 0.8 −0.6 −0.3
74 Kapok 0.2 −3.3 −0.7 0.2 −1.5 −0.4
75 Spinning 4.6 8.3 6.6 4.6 9.7 7.1
76 Weaving 7.1 17.3 12.8 7.1 18.3 12.9
77 Textiles except apparel 6.1 10.5 7.9 6.1 11.6 8.1
78 Knitted items 3.3 1.8 1.9 3.3 2.2 2.0
79 Apparel 1.4 −3.4 −2.3 1.4 −3.0 −2.2
80 Carpets, cords & other textiles 5.4 8.5 7.5 5.4 9.8 8.4
81 Processed hides 1.5 −17.8 −12.7 0.4 −3.5 −2.3
82 Leather products 3.4 4.7 4.3 3.4 6.0 5.1
83 Footwear 1.6 −0.5 0.2 1.6 0.7 0.7
84 Sawn & preserved wood 0.4 1.5 1.8 0.1 1.0 1.0
85 Plywood & related products 0.4 −0.6 0.1 0.3 −0.1 0.1
86 Wood building materials 0.1 −1.2 −0.3 0.1 −0.6 −0.2
87 Household articles from wood etc. 1.1 1.9 2.0 1.0 1.9 1.6
88 Other articles from wood, cork etc. 0.8 8.8 7.0 0.7 4.7 3.7
89 Plaited items except from plastic 0.3 3.2 3.1 0.3 2.0 1.9
90 Pulp for paper 0.2 −2.7 −1.0 0.2 −0.9 −0.3
91 Paper & cardboard 0.6 −0.1 0.4 0.6 0.5 0.6
92 Paper & cardboard products 3.2 8.0 6.4 3.2 8.7 6.5
93 Printed materials 2.4 4.1 3.6 2.4 4.8 3.8
94 Basic chemicals except fertiliser 1.9 195.3 138.6 1.5 4.7 4.2
95 Fertilisers 0.5 0.8 1.0 0.5 0.8 0.8
 TABLE 5 (cont.)

All trade policies Quantitative restric-

in effect tions excluded

Sector Description NRP ERPH ERPC NRP ERPH ERPC

96 Pesticides 2.9 5.0 4.4 2.9 5.9 4.7

97 Plastics & synthetic resins, fibres 6.5 22.7 18.7 6.3 19.1 15.6
98 Paint, varnish, lacquers 5.0 11.7 8.9 5.0 12.7 9.0
99 Pharmaceuticals 2.1 1.6 1.9 2.1 3.7 2.5
100 Traditional medicines 0.1 −14.8 −8.1 0.1 −5.8 −3.0
101 Soaps & cleaning products 0.8 4.6 3.9 0.8 4.4 3.2
102 Cosmetics 7.7 29.7 18.3 7.5 30.9 18.2
103 Other chemical products 1.7 2.2 2.2 1.7 2.5 2.1
104 Oil refining products 0.8 1.2 1.2 0.8 1.2 1.2
105 Liquid natural gas −0.2 −0.4 −0.4 0.0 −0.2 −0.1
106 Crumb & smoked rubber 0.6 −7.8 −6.2 0.5 −5.7 −4.6
107 Tyres 4.5 8.6 6.5 4.5 9.8 6.8
108 Other rubber products 3.7 9.0 7.2 3.7 9.9 7.4
109 Plastic products 7.6 24.5 16.1 7.6 24.7 15.7
110 Ceramic & clay products 13.0 71.6 49.5 13.0 29.0 21.6
111 Glass & glass products 2.5 12.4 10.2 2.5 3.1 2.6
112 Ceramic & clay building materials 10.4 180.6 81.0 10.4 33.1 20.9
113 Cement 11.6 47.4 33.5 0.2 −0.5 −0.2
114 Other non-metallic materials −1.3 −0.7 −0.1 3.5 5.8 4.8
115 Basic iron & steel 17.1 140.5 70.2 17.1 96.8 52.5
116 Iron & steel products 13.1 34.9 21.2 13.1 36.5 21.4
117 Non-ferrous basic metals −4.8 7.3 6.5 0.7 0.5 0.6
118 Non-ferrous metal products 3.9 26.3 19.3 3.9 12.8 9.5
119 Metal kitchen, craft & farm tools 3.7 2.9 2.7 3.7 3.2 2.7
120 Metal household & office articles 9.8 16.5 11.8 7.2 11.1 7.8
121 Metal building materials 7.9 15.5 11.9 7.9 16.1 12.0
122 Other metal products 7.5 157.4 83.1 7.5 12.0 8.8
123 Turbines & engines for propulsion 2.8 −0.8 0.0 2.8 0.4 0.5
124 Machines & related equipment 4.9 6.7 5.4 4.9 6.8 5.1
125 Generators & electric motors 5.7 8.1 6.5 5.7 8.4 6.3
126 Electrical machines & equipment 4.7 5.8 4.8 4.7 6.3 4.8
127 Electronic products & equipment 8.0 14.7 10.6 1.1 −1.6 −0.8
128 Electrical household items 6.2 9.7 7.4 5.9 10.3 7.4
129 Other electrical equipment 3.0 2.8 2.6 3.0 3.4 2.6
130 Batteries 3.1 4.1 3.7 3.1 5.0 3.8
131 Ships & repair services 0.6 −3.1 −1.9 0.6 −3.2 −2.3
132 Railroad cars & repair services 0.7 −6.2 −4.1 0.7 −5.5 −3.9
133 Motor vehicles except motorcycles 13.0 26.6 21.2 13.0 26.8 21.0
134 Motorcycles 9.3 15.9 13.4 9.3 16.2 13.4
135 Other transportation equipment 5.4 7.1 5.6 5.4 7.8 5.7
136 Aircraft & their repair services 0.1 −3.0 −2.1 0.1 −2.8 −2.2
137 Optics, measurement instruments 3.6 4.6 3.9 3.6 4.8 3.6
138 Jewellery 2.2 21.3 14.8 2.2 6.3 4.6
139 Musical instruments 1.5 −1.6 0.2 1.5 0.3 0.6
140 Sports equipment 6.1 10.7 8.5 6.1 11.8 8.8
141 Other manufactured goods 6.7 16.3 12.4 5.9 11.6 8.7

Note: Sector 34 is not included in the table, because it does not produce tradable goods. NRP = nominal
rate of protection; ERPH = effective rate of protection using the Humphrey method; ERPC = effective
rate of protection using the Corden method.
Non-tariff Trade Regulations in Indonesia 353

this pattern, in part due to the more protective policies for assembled vehicles
than for kits and parts.
The playing field is much more level if quantitative restrictions are excluded.
Table 4 indicates that the negative NRP and ERPs in the mining and quarrying
sector overall go away, and the NRP and ERPs for food crops are much reduced.
The assistance to food crops comes from fertiliser subsidies as well as trade poli-
cies, however: with these subsidies excluded as well as quantitative restrictions,
the ERPH for food crops would drop further, from 15.9% to 7.5%, and the ERPC
from 14.4% to 7.1%.
It may seem ironic that the removal of quantitative restrictions causes the NRP
for metals and metal products to go up, from 6.7% to 7.3%, while the ERPH drops
from 65.1% to 18.7% (table 4). The explanation is that, although some of the prod-
ucts on which export bans or restrictions were imposed in 2014 are in the metals
and metal products sector, more consequential restrictions were applied in other
mining, which is a major input supplier for the metals sectors.
Table 5 reveals the effects of excluding the quantitative restrictions in more
detail. The large negative NRPs and ERPs for the mineral sectors (sectors 38–45
in particular) are no longer present. Similarly, the NRPs and ERPs for both field
and milled rice (sectors 1 and 57), sugar cane and sugar (sectors 13 and 62), fruits
(sector 10), and vegetables (sector 9) are reduced. However, the NRPs and ERPs
remain high for processed tobacco (sector 72), which is protected by relatively
high import tariffs.
With the exclusion of the bans and restrictions on exports of metallic ores and
concentrates, the ERPs on basic iron and steel (sector 115) as well as other metal
products (sector 122) are dramatically reduced, even though the NRPs in these
sectors remain unchanged. Even so, considering the increases in MFN tariff rates
for many steel products in 2015, as well as temporary anti-dumping duties for
various steel products, the ERPs for basic iron and steel remain relatively high,
behind only those for alcoholic beverages (sector 70). The ERPs for basic chemicals
(sector 94) and non-metal products, particularly ceramic and clay products (sector
110), similarly are much reduced with the exclusion of the quantitative restrictions
on mineral exports.
As a form of sensitivity analysis, table 6 shows the effects of duty drawbacks and
exemptions (assuming an 80% recovery rate) on the two measures of the ERP in
the input–output sectors in which these effects are largest. (Recall that these effects
are included in the estimates shown in tables 4 and 5.) The effects are shown as
percentages of free-trade value added, just as the ERPs are. The effects are gener-
ally miniscule in these terms, although in a number of cases they are substantial
as a percentage of the ERP without duty drawbacks or exemptions included.

EFFECT OF POLICIES ON THE COST OF LIVING

The government has focused many of its trade restrictions in recent years on
consumer products, especially fresh and processed food, presumably so as not to
interfere with trade in intermediate inputs. One concern, however, is that wage
increases commensurate with the implied higher cost of living could impact the
competitiveness of labour-intensive sectors. Moreover, import barriers, like com-
modity taxes in general, hit those at the lower end of the income distribution
354 Stephen V. Marks

TABLE 6   Increase in Effective Rates of Protection Due to Recovery of Duty Drawbacks

and Exemptions at an 80% Rate, with All Other Policies in Effect (percentage points)

Sector Description ERPH ERPC

12 Rubber 0.1 0.1

17 Tobacco 0.1 0.1
46 Non-metallic minerals 0.1 0.1
67 Soybean products 0.2 0.1
68 Other food products 0.3 0.2
69 Animal feeds 0.1 0.1
72 Processed tobacco 1.0 0.7
76 Weaving 2.0 1.3
92 Paper & cardboard products 0.4 0.3
94 Basic chemicals except fertiliser 1.5 0.8
97 Plastics & synthetic resins, fibres 1.0 0.8
98 Paint, varnish, lacquers 1.3 0.9
99 Pharmaceuticals 0.4 0.3
103 Other chemical products 0.8 0.6
108 Other rubber products 1.5 1.1
109 Plastic products 3.1 1.8
114 Other non-metallic materials 0.2 0.2
115 Basic iron & steel 0.5 0.2
116 Iron & steel products 3.1 1.6
122 Other metal products 7.2 2.7
129 Other electrical equipment 1.2 0.8

Note: NRP = nominal rate of protection; ERPH = effective rate of protection using the Humphrey
method; ERPC = effective rate of protection using the Corden method.

harder than others, because consumption, particularly of daily essentials such

as rice, tends to make up a larger share of the incomes of lower-income families.
With the Humphrey method fully solved, and with all of the examined policies
in effect, the money wage has to be 7.6% higher in the protected situation than
under free trade, in order for the real wage to remain constant. As noted earlier,
this adds to the costs of production in all sectors, particularly labour-intensive ones,
and raises the prices of non-tradable services. With all trade policies retained, but
subsidies and excise taxes eliminated, the cost of living is 7.8% higher.
If quantitative restrictions are excluded, the increase in the cost of living versus
free trade is only 2.9%. Thus, 62% of the increase in the cost of living is caused by
quantitative restrictions. Much of the increase in the cost of living with NTMs in
effect is driven by the import restrictions on rice; if these were allowed to lapse,
but rice import tariffs were retained along with all other policies, the increase in
the cost of living would only be 4.9%.

COMPARISON WITH PAST ESTIMATES

Indonesia had far fewer NTMs in 2008 than in 2015. In a similar analysis based on
the policies in effect in early 2008, Marks and Rahardja (2012) included estimates
Non-tariff Trade Regulations in Indonesia 355

of NRPs only for quantitative restrictions on imports of sugar and rice, and on
exports of logs, sand, and topsoil. Other quantitative trade restrictions were in
effect, however, notably for imports of cloves and salt and for exports of tin ores
and concentrates. Ad valorem import duties of up to 150% for alcoholic beverages
applied in 2008, but quantitative restrictions were introduced in 2009 and later the
ad valorem duties were replaced by specific tariffs. Export restraints were also
applied to some varieties of unprocessed rattan, but a complete ban on unpro-
cessed rattan exports was not implemented until 2011. None of the procedural
regulations examined in this article were in effect in 2008.
Marks and Rahardja (2012) found that the structure of nominal and effective
rates of protection was relatively flat, with the notable exceptions of rice, sugar,
and alcoholic beverages. The authors estimated the NRPs for milled rice and cane
sugar to be 36.9% and 35.6% respectively—considerably less than the rates of 67.8%
and 54.8% found in this study for 2015. Similarly, for alcoholic beverages, the NRP
on the import side was 118.7% in 2008 but 133.3% in 2015.
To do a more systematic comparison, I used the computational framework of
Marks and Rahardja (2012), after making several adjustments to bring it more in
line with the present study.22 First, the ERPs for sugar in the earlier study were
much higher than those in this study. The decrease in 2015 is not due to any reform
in trade policy, since the NRP is even higher than before, but rather the result of
a better (albeit imperfect) method in this study for allotting the gains from sugar
policy between sugar factories and sugar-cane farmers, as described above. There
is a similar difference between the two studies in the way that the NRP for field
rice is calculated. Thus, the method used this time for sugar cane and field rice
was applied to 2008 as well.
I have made a similar adjustment for oil palm. Marks and Rahardja (2012)
assumed that the 40% export tax on palm fresh fruit bunches determined the
local prices of the bunches, but palm fruit is not much traded internationally. In
the present study, the price is instead assumed to be determined by the regulated
arrangements between oil palm processors and independent farmers, under which
the price of the bunches is set as a function of the yields and prices of the crude
palm oil (CPO) and palm kernels obtained from the fruit. On this basis, data on
yields, prices, and export taxes on CPO, crude palm kernel oil, and palm kernel
cake can be used to impute a price for the fresh fruit bunches.23 Using this meth-
odology, I estimated the NRP for the bunches at −8.5% for 2015 and −9.9% for 2008
(in contrast to the figure of −36.0% used by Marks and Rahardja).24

22. Marks and Rahardja (2012) presented ERP calculations based on the Corden (1966) and
Balassa (1982) methods, but the present study replaces the latter with the Humphrey (1969)
method for data in both 2015 and 2008.
23. In this analysis, one must make an assumption about the processing margin for palm
kernels. Sensitivity analysis indicated that large variations in the margin did not have
much of an effect, mainly because palm kernel yields are much smaller than CPO yields.
I arbitrarily used a margin of 20% of the value of output. If the margin were 0%, however,
the export tax rates for the two years would round to the same values, and if the margin
were 100%, they would be −8.4% for 2015 and −10.0% for 2008.
24. The −40.0% NRP on the export side in 2008 was diluted by imports, such that the overall
NRP was −36.0%. News reports from 2015 stated that farmers in North Sumatra were being
356 Stephen V. Marks

Finally, the estimates of NRPs for tin ores and concentrates in this study are also
included in the database for the earlier period, since they were already in effect at
that time. Like Marks and Rahardja (2012), I assumed that duty drawbacks were
recoverable at an 80% rate, and I accounted for fuel and fertiliser subsidies.
The last set of columns in table 4 shows the calculated values of NRP, ERPH,
and ERPC for 2008, for the aggregated categories also shown for 2015.25 If this set
of columns is compared with the first set of columns, it is noteworthy that food
crops, and agriculture in general, had much higher nominal and effective rates
of protection in 2015 than in 2008, despite the reduction in the scope of fertiliser
subsidies in the interim. For mining other than oil and gas, on the other hand, the
rates were more negative in 2015 than in 2008. In manufacturing, the nominal and
effective rates of protection were higher overall in 2015 than in 2008, particularly
in metals and metal products and in food, beverages, and tobacco.
Overall, the value-added-weighted nominal and effective rates of protection
for all tradable sectors were lower in 2008 than in 2015, as shown in the bottom
row of table 4. In addition, the dispersion of NRP, ERPH, and ERPC, as indicated
by the weighted standard deviation of these measures for all 140 tradable sectors,
increased from 11.1%, 19.5%, and 16.3% respectively in 2008 to 17.9%, 43.0%, and
31.5% respectively in 2015. Finally, based on the Humphrey method calculations,
the increase in the money wage required to maintain a constant real wage with
distortive policies in effect was only 2.5% in 2008, versus the 7.6% figure noted
above for 2015. Overall, then, this study finds that the Indonesian economy is
subject to more widespread, and in many cases more heavy-handed, interventions
than in the not-so-distant past.

CONCLUDING REMARKS
Reducing multifaceted trade policies to a single nominal rate of protection is in
many cases clearly an oversimplification. Deardorff and Stern (1998) point out
that NTMs not only have price or quantity effects, but also can impact elasticities
of demand, supply, or substitution, such that the predicted impact of, say, a tariff
cut, not accounting for the changes in elasticities, will be inaccurate, even if the
assumed parameters of the model would otherwise have been correct. Moreover,
the interpretation of effective rates of protection rests on strong assumptions,
especially in a world of multiple preferential import tariff schedules that must be
reduced to a single tariff rate.
A just-finished AIPEG survey of businesses in certain sectors investigated fur-
ther the effects of some of the trade regulations included in this study. Among
the issues examined were the costs of utilising preferential import tariff rates,

paid a CPO price considerably below the price implied by the formula in the regulation
of the Minister of Agriculture. The issue in this case is similar to that for sugar cane and
field rice: if oil palm farmers receive a constant fraction of the mandated price, their rate of
protection due to the export levies is unaffected by the price they are actually paid being
below the mandated price. However, if their fraction goes down as export taxes go up, all
else being equal, their NRP will become more negative.
25. Calculations for all 140 tradable-goods input–output sectors in 2008 are available from
the author upon request.
Non-tariff Trade Regulations in Indonesia 357

pre-shipment inspection costs, the costs to exporters of claiming duty drawbacks

for imported inputs, and the effects of restrictions and bans on the importation of
used capital equipment. Although the survey responses are still being analysed,
one important finding is that import licensing requirements in the alloy steel sector
have not been mere formalities, but have led to reductions in the imports of some
firms. That sector merits additional study.

REFERENCES
ADB (Asian Development Bank). 2014. 2011 International Comparison Program in Asia and the
Pacific: Purchasing Power Parities and Real Expenditures. Manila: ADB.
Balassa, Bela, ed. 1982. Development Strategies in Semi-industrial Economies. Baltimore: Johns
Hopkins University Press.
British Geological Survey. 2015. World Mineral Production, 2009–13. Keyworth, Nottingham:
British Geological Survey.
Broda, Christian, Nuno Limão, and David E. Weinstein. 2008. ‘Optimal Tariffs and Market
Power: The Evidence’. American Economic Review 98 (5): 2032–65.
Corden, W. Max. 1966. ‘The Structure of a Tariff System and the Effective Protective Rate’.
Journal of Political Economy 74 (3): 221–37.
Deardorff, Alan V., and Robert M. Stern. 1998. Measurement of Nontariff Barriers. Ann Arbor:
University of Michigan Press.
Devadason, Evelyn S., and Santha Chenayah. 2014. ‘Proliferation of Non-tariff Measures
in China—Their Relevance for ASEAN’. Singapore Economic Review 59 (2): 1–28.
Fane, George, and Timothy Condon. 1996. ‘Trade Reform in Indonesia, 1987–95’. Bulletin of
Indonesian Economic Studies 32 (3): 33–54.
Fane, George, and Chris Phillips. 1991. ‘Effective Protection in Indonesia in 1987’. Bulletin
of Indonesian Economic Studies 27 (1): 105–25.
Humphrey, David B. 1969. ‘Measuring the Effective Rate of Protection: Direct and Indirect
Effects’. Journal of Political Economy 77 (5): 834–44.
ILO (International Labour Organization). 2015. ‘Indonesia: Trends in Wages and Productiv-
ity’. ILO, Geneva, January.
Kee, Hiau Looi, Alessandro Nicita, and Marcelo Olarreaga. 2008. ‘Import Demand Elastici-
ties and Trade Distortions’. Review of Economics and Statistics 90 (4): 666–82.
Marks, Stephen V. 2011. ‘Lao Economic Policies and Effective Rates of Protection’. Journal
of Asian Economics 22 (2): 115–27.
———. 2015. ‘Non-tariff Trade Regulations in Indonesia: Measurement of Their Eco-
nomic Impact’. Australia Indonesia Partnership for Economic Governance, Jakarta, 14
September. http://research.pomona.edu/stephen-marks/files/2016/05/Analysis-of-NTMs-
in-Indonesia.pdf.
Marks, Stephen V., and Sjamsu Rahardja. 2012. ‘Effective Rates of Protection Revisited for
Indonesia’. Bulletin of Indonesian Economic Studies 47 (1): 53–80.
Ministry of Trade. 2012. ‘Analisis kebijakan ekspor: Analisis dampak kebijakan pelaran-
gan ekspor rotan’ [Analysis of export policy: Analysis of the impact of the rattan export
prohibition]. Trade Policy Studies and Development Agency, Centre for Foreign Trade
Policies, Jakarta.
Warr, Peter G. 1992. ‘Comparative Advantage and Protection in Indonesia’. Bulletin of Indo-
nesian Economic Studies 28 (3): 41–70.
 Bulletin of Indonesian Economic Studies

ISSN: 0007-4918 (Print) 1472-7234 (Online) Journal homepage: https://www.tandfonline.com/loi/cbie20

Value Chain Connectivity in Indonesia: The

Evolution of Unbundlings

Fukunari Kimura & Lurong Chen

To cite this article: Fukunari Kimura & Lurong Chen (2018) Value Chain Connectivity in Indonesia:
The Evolution of Unbundlings, Bulletin of Indonesian Economic Studies, 54:2, 165-192, DOI:
10.1080/00074918.2018.1505412

To link to this article: https://doi.org/10.1080/00074918.2018.1505412

Published online: 26 Sep 2018.

Submit your article to this journal

Article views: 845

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=cbie20
 Bulletin of Indonesian Economic Studies, Vol. 54, No. 2, 2018: 165–192

Indonesia in Comparative Perspective

VALUE CHAIN CONNECTIVITY IN INDONESIA:

THE EVOLUTION OF UNBUNDLINGS

Fukunari Kimura* Lurong Chen*

Economic Research Institute for ASEAN Economic Research Institute for
and East Asia; and Keio University ASEAN and East Asia

This paper applies the conceptual framework of ‘unbundling’, proposed by Baldwin

(2016), to assess the value chain connectivity in Indonesia. Indonesia is a geographically
large country, and three different levels of unbundling co-exist. The first unbundling
aligns with the industry-wise international division of labour in plantation agriculture,
mining, and labour-intensive industries; the second aligns with where the task-wise
international division of labour dominates mainly in machinery industries; and the
third aligns with where face-to-face costs are reduced by the digital economy. We
argue that the further upgrading of unbundling and the more effective use of advanced
piecemeal technologies among different levels of unbundling will enable Indonesia to
achieve rapid and equitable economic development. Overcoming distance is key for
effectively using the mechanics of unbundling. We propose that three key elements
are be promoted: enhancement of connectivity, development of the service sector, and
improvement of the policy environment.

Keywords: international division of labour, distance, connectivity, global value chains,

information and communication technology
JEL classification: F10, F14, O14

INTRODUCTION
The concept of ‘unbundling’, or the notion that value chains between producers
and consumers are becoming longer through specialisation of production
(Baldwin 2016) is especially useful in formulating development strategies for a
large country such as Indonesia. How far economic activities can be unbundled
depends on connectivity to overcome distance, which could be uneven across a
country. This paper uses the unbundling approach and provides insights on the
location of industries and industrialisation in Indonesia.

* The authors would like to thank two referees for their helpful comments and suggestions.
We are also grateful for the organisers and participants of the 12th Sadli Lecture held in
Jakarta on 2 May 2018 and the Mubyarto Public Policy Forum in Yogyakarta on 3 May 2018,
particularly Zakir Machmud and Yos Adiguna Ginting.

ISSN 0007-4918 print/ISSN 1472-7234 online/18/000165-192 © 2018 ANU Indonesia Project

http://dx.doi.org/10.1080/00074918.2018.1505412
166 Fukunari Kimura and Lurong Chen

The older literature on international trade used the concept of a ‘lumpy’ country
in which the partial internal immobility of productive factors and the resulting
uneven internal distribution of productive factors of production generated
differences in international trade patterns across regions (Courant and Deardorff
1992). In particular, the theoretical model claims that a country tends to export
the goods that intensively use its unevenly distributed factor of production. This
model seems to explain, to some extent, differences in international trade patterns
across regions in Indonesia. If mineral resources and available land are treated
as factors of production, strong exports of mining products and palm oil by
Indonesia, particularly off Java Island, may be well explained.
However, this approach does not explain other important types of skewness
in the locations of industries. Why don’t cities in outer islands such as Makassar,
with its relatively large population, have substantial labour-intensive industries?
Why are machinery industries concentrated mostly in Java, rather than in cities
such as Medan? Indonesia occupies almost half of the total population of
ASEAN countries and lies in a huge archipelago from east to west. If we think
of patterns in the rest of ASEAN, it would be natural for Indonesia to have a
more scattered pattern of labour-intensive industries and multiple centres for
machinery industries. Within a country, people can move relatively easily, and
industries can also choose their location. Just like the chicken and the egg, it is
difficult to prove the causality of economic opportunities and the location of
resources. Nevertheless, the preparedness of each region for different levels of
unbundlings has affected the regional pattern of production and international
trade in Indonesia.
‘Unevenness’ is a key feature of the Indonesian economy. Average income in the
richest province, Jakarta, is over 10 times higher than the poorest province, East
Nusa Tenggara. Only 7 out of 34 provinces have a higher GDP per capita than
the national average. The Human Development Index ranges from 0.8 in Jakarta
to 0.58 in Papua. The country has not reached its full potential in participating
in international trade and value chains promoted by production unbundling.
There are over 1,000 seaports widespread across the country. However, the
interlinkages between them are rather weak, and the cost of domestic logistics
is high (Hill, Resosudarmo, and Vidyattama 2008). This has fragmented national
economic activities and created difficulties in economic growth.
Meanwhile, the digital economy has created another dimension of connectivity.
Access to the internet and smartphones by ordinary people has grown quickly
in Indonesia, perhaps due to its large young population. Interestingly, the gap
in digital connection between urban and rural areas, or between Java and outer
islands, seems to be relatively small compared with the large differences in
physical logistics links. This new wave needs to be considered when developing
strategies for the future.
This paper employs the concept of unbundlings and discusses the status and
the future perspectives of the Indonesian economy. The discussion centres on
three distinct levels of unbundling backed by technological advancements to
overcome geographical distance. Each level of unbundling requires different sets
of logistics and other economic infrastructure, policy environments, and human
resources. The readiness of a region also impacts on unbundling. Although a
region can simultaneously host various business models that use the mechanics
Value Chain Connectivity in Indonesia: The Evolution of Unbundlings 167

of different levels of unbundling, the scope for upgrading unbundlings may

be bounded by readiness in various aspects. In this analytical approach, policy
lessons can be drawn from investigating the preparation of the whole country
or specific regions for different levels of unbundling, which can help to take
advantage of new technologies and solve bottlenecks to advance unbundlings.
The analysis in the paper mainly uses international trade statistics, both at the
national and regional customs office levels. It assesses differences in the dominant
levels of unbundling across regions in Indonesia and provides ‘out-of-the-box’
considerations of national development from the perspective of technological
progress and globalisation.
The structure of this paper is as follows. It begins with a conceptual framework
of unbundlings, it then evaluates Indonesia’s performance of international trade
under the first two waves of unbundling and discusses Indonesia’s prospects
for the upcoming third unbundling, and it concludes by briefly discussing the
policy implications.

THE CONCEPTUAL FRAMEWORK

Our interpretation of Baldwin’s (2016) unbundling framework is summarised
in figure 11. We start from the pre-globalised world in which trade costs,
communication costs, and face-to-face costs are all high and a country basically
consumes what it produces—i.e. it is close to autarky. Around 1820, the mass
transport system represented by railways and steamships began to develop,
which reduced trade costs. This triggered the first unbundling where places
of production and consumption were unbundled and the industry-to-industry
international division of labour gradually came to dominate the world. In the
1980s, the information and communication technology (ICT) revolution began,
and long-distance communication costs were substantially reduced. This is the
beginning of the second unbundling that advanced the task-wise or production
process-wise international division of labour. And then, in the past few years,
a further ICT revolution reduced the cost of face-to-face interactions, which
characterises the third unbundling.
Traditional trade theories primarily deal with the first unbundling in which
countries specialise in industries in which they have a comparative advantage. Once
a country or a region is exposed to the world economy, it becomes increasingly
engaged in international industrial linkages or gradually evolving global value
chains (GVCs). The second unbundling, on the other hand, fragments tasks or
production processes into multiple production blocks and locates them in different
places. For this to happen, service link costs (Jones and Kierzkowski 1990) or
communication costs (Baldwin 2016) must be low enough to coordinate whole
production chains. That is why only a limited number of developing countries,
including some in East Asia and Eastern Europe as well as Mexico and Costa Rica,
engaged in the second unbundling, particularly in machinery industries.2

1. See Kimura (2018), for more a detailed explanation

2. Obashi and Kimura (2017) show that the penetration of the second unbundling widely
differs even among ASEAN member states.
168 Fukunari Kimura and Lurong Chen

FIGURE 1  Evolving Unbundlings to Overcome Distance

Pre-globalised First Second Third

world unbundling unbundling unbundling
(0) (1) (2) (3)

Trade costs High Lower Lower Lower

Communication costs High High Lower Lower
Face-to-face costs High High High Lower

What starts moving? None Goods Ideas People

Industry-wise
Task-wise Person-wise
International labour (fragmented
Autarky (fragmented (fragmented
division production &
industry) task)
consumption)

Dominant years –1820 1820–1990 1990–2015 2015–

It should also be emphasised that fragmentation of production goes with the

formation of industrial agglomeration (Kimura and Ando 2005). A production
network, once well developed, often consists of intra-firm transactions over long
distances and arm’s length (inter-firm) transactions over shorter distances, the
latter generating industrial agglomeration through vertical linkages. Industrial
agglomeration stabilises the industrial structure, provides chances for local firms
to link with multinational enterprises, and promotes technology transfer and
spillovers (Kimura, Machikita, and Ueki 2016). Job creation in manufacturing and
related services sectors accelerates poverty alleviation through internal labour
migration (Kimura and Chang 2017; Findlay and Pangestu 2016).
The third unbundling is associated with economic digitalisation. Under the
third unbundling, the face-to-face connectivity is more convenient than ever
before, with lower cost and higher quality. The division of labour is going to be
person-wise—a task will be unbundled into work for individual people, possibly
located in different countries, to complete. Although statistics on this type of
labour division do not exist, its practice is likely to flourish domestically and
internationally in the near future. The expansion of ‘working from home’ or
‘telecommuting’ are examples, as tele-presenting and tele-conferencing.
So far, the reduction in face-to-face costs makes the matching up of people easier,
not only for business-to-business (B2B) but also for business-to-consumer (B2C)
and consumer-to-consumer (C2C) transactions. As a result, various internet-based
digital platform businesses are mushrooming all over the world, including in
social media, e-commerce, business matching, the sharing economy, e-payments,
Value Chain Connectivity in Indonesia: The Evolution of Unbundlings 169

and fintech. On these platforms, even small businesses and individuals can be
sellers or buyers, and observers or participants. In this environment, services
can be outsourced to a substantial degree. These activities can be extended both
domestically and internationally through internet connections.
The ICT revolution has two aspects: information technology (IT) and
communication technology (CT) (Baldwin 2016). IT is represented by industry
4.0 and artificial intelligence (AI) that enhances the speed of data processing and
reduces the number of tasks. It may make current divisions of labour redundant,
lead to a replacement of labour by machines, and generate concentration of labour
forces. On the other hand, CT such as the internet and smartphones reduce the
cost of overcoming physical distance and are prone to encouraging the division
of labour by dispersion forces. It is not yet clear which force will dominate in
the coming new technological regime. From the viewpoint of less-developed
economies, it is relevant to consider how far they can take advantage of dispersion
forces and attract economic activities with better connectivity. In this regard, the
third unbundling will be a key concept for development strategies.
The third unbundling has already brought new opportunities as well as
challenges for less-developed countries. Digital platform companies tend to
enjoy network effects that create strong economies of scale. On the other hand,
the introduction of foreign platform giants such as GAFA3 from the US and
BAT4 from China would generate more business opportunities for local firms in
applications (Accenture 2016). With technology progress, entry costs for local
platforms and applications become very low so that even small to medium
enterprises (SMEs) can enter the market, although profit dissipation may be quick
as well. Consumers also receive benefits from digital connectivity (Chen 2017).
The impact of economic growth on equity issues could be different in the third
unbundling phase. In addition, how to deal with foreign platform giants may
become an important issue when considering competition policy and taxation
systems in individual countries. And how far domestic firms should be protected
may attract attention in industrial promotion.
To unleash the benefits from the third unbundling, soft and hard infrastructure as
well as human capital will be the key. For soft infrastructure, further liberalisation
and facilitation for trade of goods and services, as well as the movement of
people, will become even more important. ‘Almost’ free flow of data across
national borders will be crucial for the development of digital economy, but it
must be accompanied by a series of back-up policies for intellectual property
right protection, consumer protection, privacy protection, and cyber-security.
For hard infrastructure, internet connection is essential, though dependent on
quick and efficient logistics for goods transactions. Human capital is likely to
be a bottleneck for less-developed countries to expand their commitments to
the third unbundling, so that nurturing human capital as well as use of foreign
human resources may become crucial.

3. GAFA is an acronym for Google, Apple, Facebook, and Amazon.

4. BAT is an acronym for Baidu, Alibaba, and Tencent.
170 Fukunari Kimura and Lurong Chen

FIGURE 2  Key Features of Unbundling Regimes

Pre-globalised First Second Third

world unbundling unbundling unbundling
(0) (1) (2) (3)

Agriculture
Representative Subsistence Machinery Digital
Mining
industries agriculture industries economy
Manufacturing

Mass Supply chain Advanced

Self- production; management; information &
Key technologies
subsistence mass transport time-sensitive communication
system logistics technology

Concentration
Fragmentation
Comparitive & dispersion of
Geography Autarky & industrial
advantage innovation &
agglomeration
its applications

INDONESIA IN GLOBAL UNBUNDLINGS

Figure 2 is a simple illustration of representative industries, key technologies,
and geography that are associated with each unbundling regime as it applies
schematically to Indonesia in the past and the present. Mass production is the
backbone of the first unbundling. Typical industries in the first unbundling
include resource-based industries such as plantation agriculture and mining,
and manufacturing sub-sectors such as labour-intensive garment and footwear
industries in which industry-wise comparative advantage, based on resource
endowments, factor endowments, and production technologies, determine
the pattern of production and international trade with the support of a mass
transportation system. The second unbundling works mainly on manufacturing,
especially machinery industries. As the fragmentation of production and the
formation of industrial agglomeration proceed, the task-wise international
division of labour and parts and components trade are developed with efficient
and time-sensitive logistics links. The third unbundling is closely linked to the
digital economy. Digitalisation is leading a new wave of industrial revolution, B2C
and C2C matching, and perhaps in the near future the fragmentation of a task to
individuals over distance. How does this framework apply to Indonesia?

Trade Driven by the First Unbundling

Conceptually, the first unbundling is one of the key drivers behind the dynamic
‘flying-geese-formation’ process of region-wide sequential industrialisation in
northeast and southeast Asia where countries upgrade their industries from
the labour-intensive industries to the capital-intensive or human capital-
intensive industries (Akamatsu 1962; Kojima 2000; Kumagai 2008). As producers
in a country in a higher tier find new comparative advantages in the field of
 FIGURE 3  Indonesian Non-oil and Oil Trade (US$ billion and %)

Share of manufactures
200 in non-oil exports (rhs) 60

150

40

100

20

50

0 0
2010 2011 2012 2013 2014 2015

Non-oil exports Oil exports

Share of
200 100
manufactures
in non-oil
imports (rhs)
80
150

60

100

40

50
20

0 0
2010 2011 2012 2013 2014 2015

Non-oil imports Oil imports

Source: Data from Comtrade database.

172 Fukunari Kimura and Lurong Chen

TABLE 1  Top Indonesian Ports, 2015 (% in value)

Port Province % of trade % of exports % of imports

Tanjung Priok Jakarta 33.90 27.10 41.20

Tanjung Perak East Java 9.10 8.50 9.70
Soekarno-Hatta Jakarta 6.10 3.80 8.50
Dumai Riau 4.10 7.60 0.40
Tanjung Emas Central Java 3.70 3.50 3.80
Belawan North Sumatra 3.50 4.40 2.60
Batu Ampar Riau Islands 2.40 2.20 2.60
Bontang East Kalimantan 2.20 4.00 0.30
Balikpapan East Kalimantan 2.10 1.30 3.00
Cilacap Central Java 1.80 0.10 3.70
Merak Banten 1.70 0.20 3.30

Source: Data from BPS.

capital-intensive or human capital-intensive industries, their labour-intensive

industries will shift to countries in the tier below. In addition, the endowment
of natural resources will decide how far a country continues to depend on
agriculture, mining, and other natural resource-based industries, some of which
go with the first unbundling-type operations.
When ranking Asian countries in the four-tiered flying-geese formation up to
the 1980s, Japan could be regarded as the lead goose, followed by the four Asian
newly industrialised economies (NIEs) (the second tier), the four main ASEAN
countries (the third tier), and finally other late-developing economies such as
China and Vietnam. Indonesia was flying in the third tier, together with Malaysia,
the Philippines, and Thailand, although it did not follow the group as closely as
its companions.
The country is well endowed with natural resources such as oil and natural
gas, and it possesses comparative advantage in primary products in addition
to labour-intensive activities. In fact, Indonesia is the world’s main exporter
of minerals such as coal, nickel, copper, and gold, as well as products from
agricultural plantations, such as those farming palm oil and rubber. China and
Japan are among Indonesia’s largest export markets for palm oil and coal.
Oil and natural gas used to dominate the country’s export sector. The
combined export share of these products varied between 50% and 80% in the
1980s, depending on world prices. However, this has changed since the early
1990s as the Indonesian economy has moved towards industrialisation and it
has increased its exports of manufactures. More than 75% of exports are now
non-oil products, of which 50%–60% come from the manufacturing sector (figure
3). On the other hand, oil accounted for around 20% of Indonesia’s imports
during 2010–15. The imported oil products mainly comprised crude petroleum
oil, unleaded motor spirit, and automotive diesel fuel. Around the 66% of imports
were manufactured goods.
Value Chain Connectivity in Indonesia: The Evolution of Unbundlings 173

A port-level analysis shows that Indonesia has over 200 seaports that deal with
international trade. Most are quite small in terms of physical size, the value of
transactions, and the number of products traded. The top 10 seaports, together
with Soekarno-Hatta Airport in Jakarta, accounted for over 70% of international
trade (63% of total exports and 80% of total imports) of Indonesia (table 1).
Belawan in Medan, Tanjung Priok in Jakarta, Tanjung Perak in Surabaya, and
Makassar in South Sulawesi are the four hubs for domestic trade in Indonesia.
The first three have recently become gateways for international trade as well.
In 2016, the volume of international unloaded cargo exceeded that of domestic
unloaded cargo for the first time.
Some Indonesian ports are ‘import specialised’ while some others are ‘export
specialised’. For instance, Dumai Port in the Riau province accounted for nearly
8% of national total exports, but its imports were marginal. Over 90% of its exports
were categorised under codes 15 and 27 of the World Customs Organization’s
(WCO) international two-digit Harmonized System (HS).5 About 50% of total
exports were shipped to China, India, Japan, and the United States. Another
example is Bontang Port in East Kalimantan, which concentrated on exporting
fuel to Japan. In comparison, Cilacap in Central Java and Merak in Banten are
almost purely ‘import specialised. Cilacap specialises in importing oil and gas
from Saudi Arabia, Nigeria, and Angola, while Merak mainly imports oil and
chemical products from ASEAN countries, China, and South Korea.
The pattern of one-way, inter-industry trade provides two key pieces of
information. First, the provincial economy in the area surrounding a port is
sometimes heavily depends on just one or a few traded products such as mining
products and palm oil. This trade looks like it is first unbundling-driven, and
some regions are simply suppliers in this process. Second, since the Indonesian
economy is still impeded by the low efficiency of its domestic infrastructure and a
lack of integration and production coordination, provincial production activities
sometimes link directly with international markets rather than with supply and
demand in domestic markets.6
The spatial distribution of both industry and population is highly skewed
towards Java Island. It occupies nearly 80% of non-agriculture industries, provides
about 60% of total manufacturing jobs and contributes almost 60% to national
GDP. Jakarta hosts more than 30 million people,7 and it generated over 17% of
national GDP in 2015. It accounted over 30% of total exports and half of total
imports. Both the country’s largest seaport, Tanjung Priok, and the largest airport,
Sukarno Hatta Airport, are affiliated with the city’s economy. Manufactures are

5. HS code 15 covers animal or vegetable fats and oils and their cleavage products; prepared
edible fats; and animal or vegetable waxes. HS code 27 covers mineral fuels, mineral oils
and products of their distillation; bituminous substances; and mineral waxes.
6. Zen, Khoirunnurofik, and Yudhistira (2018) conducted a questionnaire and found that the
insufficient capability of soft and hard logistics infrastructure limits the scope of businesses,
particularly in outer islands.
7. Including those living in the surrounding metropolitan region (known as Jabodetabek).
174 Fukunari Kimura and Lurong Chen

FIGURE 4  Trade Driven by First Unbundling (by province)

Note: Darker shading represents relative manufacturing intensity and a higher likelihood of second
unbundling-driven trade.

1
Share of GDP in 2016. Data from BPS online database.
2
Share of population in 2010. Data from BPS online database.
3
Share of total exports in 2015 . Data from BPS online database.
4
Share of total imports in 2015. Data from BPS online database.

highly concentrated in the industrial clusters surrounding Jakarta.8 Nearly 70%

of international machinery trade occurs in this area.
Surabaya in East Java serves as the second hub of trade in Java. It accounts
for over 11% of total national exports and nearly 14% of imports. About
three-quarters of foreign trade is handled by Tanjung Perak Port, Indonesia’s
second-largest commercial seaport. In figure 4, most areas in Java are shaded
grey, showing the relatively higher intensity of manufacturing and the higher
likelihood that international trade from nearby ports is driven by either the first
or second unbundling. Jakarta and Surabaya are shaded dark grey to highlight
their status as trade hubs in Indonesia.
Outside Java, the Riau Islands province is another main source of exports. With
less than 2% of national GDP and 0.7% of the national population, Riau Islands
contribute over 20% of Indonesia’s total exports and nearly 10% total imports of
machinery products. Singapore dominates as a trading partner with this province,
accounting for 55% of exports and 40% of imports. Sub-regional cooperation

8. Some cities that are a part of greater Jakarta such as Bogor, Tangerang, and Bekasi have
been growing more rapidly, albeit from a low base (MGI 2012).
Value Chain Connectivity in Indonesia: The Evolution of Unbundlings 175

under the framework of Singapore–Johor–Riau Islands (SIJORI9) Growth Triangle

has played a role in shaping this trade pattern. The idea of SIJORI was proposed
in 1989 by Singapore, and the Indonesia–Malaysia–Singapore Growth Triangle
(IMS-GT) was formally created by the three countries in 1994. Singapore started
to transfer its labour-intensive industries to neighbouring areas, while Malaysia
and Indonesia tried to attract foreign direct investment and export-orientated
manufacturing.
In figure 4, the majority of Sumatra as well as South Kalimantan and East
Kalimantan are shaded light grey. The development of the local economies in
these regions has been connected to globalisation, but it has still primarily been
characterised by first unbundling-type activities.10 The region is well endowed
with natural resources, so companies in these regions have been global competitors
in exporting products from plantation agriculture and mining sectors.11
These areas host around 25% of Indonesia’s total population and 20% of its
national GDP, but their contribution to manufacturing trade is still marginal,
especially to machinery trade. Further expansion to the second or third
unbundling in these regions is likely to deepen Indonesia’s overall involvement
in the global economy.
The white areas in figure 4 are ‘pre-globalised’ regions. They cover the whole of
Sulawesi, the west and central parts of Kalimantan, and some other outer islands.
They do not have many direct links with the rest of the world. Among the four ports,
Makassar is the only one that still specialises in domestic shipments. The local
economy is close to being independent of international trade, although indirect
international connections exist via domestic trade. There may be opportunities for
these regions to expand their involvement in the first unbundling.

Trade Driven by the Second Unbundling

While the first unbundling emerged because of the international separation
between places of production and consumption, the second unbundling took
shape through the international fragmentation of production, especially in
complex industrial products. When the East Asian economies broke away from
their original order among the ‘flying geese’ in the late 1980s and early 1990s,
the connectivity among Asian countries was further enhanced by the second
unbundling, in which ‘formerly national production processes have been
unbundled and dispersed to the lowest cost location in East Asia’ (Baldwin 2006).
In the second unbundling, production becomes internationally fragmented
and is shared by producers in different locations. This mainly occurs with
manufactured goods. Typically, it involves two-way intra-industry exchange of
goods. Imported and exported goods could be of the same category of products,
even in the most disaggregated commodity classification, but with different
content of value-added. Technology plays a role in determining the stage of the
value chains in which a country retains its competitiveness.

9. SIJORI is an acronym using the first two letters of Singapore, Johor, and Riau Islands.
10. See Van der Eng (2015), Garnaut (2015), and Pangestu (2010) on natural resource-based
industries and their implications in Indonesia.
11. See also Woo and Hong (2010).
176 Fukunari Kimura and Lurong Chen

FIGURE 5  Indonesian Trade in Machinery Parts and Components (US$ billion and %)

80 Share of parts 60
& components in
machinery im ports (rhs)

Share of parts
60
& components
in m achinery 40
exports Share of
(rhs) machinery in
40 total im ports (rhs)

20

20 Share of machinery
in total exports (rhs)

0 0.0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Machinery exports Machinery imports

Source: Data from UN Comtrade database.

Note: The definition of machinery products and the corresponding parts and components
are based on that of Obashi and Kimura (2017).

According to the World Bank (2017), value-added in manufacturing contributed

21% to GDP in Indonesia in 2015. This was lower than in China (29%) and some
neighbouring ASEAN countries such as Malaysia (23%) and Thailand (28%). Less
than 20% of manufacturing value-added came from the machinery and transport
equipment sector. Over 50% of non-oil exports and 66% of total imports were
manufactured goods.
From 2000 to 2015, Indonesian machinery exports (HS84–92) steadily increased
from US$11 billion in 2000 to US$24 billion in 2012, and then decreased to US$21
billion in 2015. Machinery imports experienced even faster, but more fluctuating,
growth during this period. By 2015, Indonesian machinery exports doubled
in value while imports increased fourfold, compared with 2000. The share of
machinery exports in total exports declined from 18.3% in 2000 to 11.2% in 2011
and then increased to 13.9% in 2015. The share of machinery imports in the same
period increased from 28.6% to 38.4% in 2009 and then slipped back to 33.2%
in 2015. Trade in parts and components accounted for nearly half of the whole
machinery trade, showing the importance of GVCs in developing the country’s
manufacturing sector, especially on the import side. Indonesia is a net importer
of parts and components of machinery and transportation equipment. In 2015,
Value Chain Connectivity in Indonesia: The Evolution of Unbundlings 177

FIGURE 6.1  Global Value Chain Indicators for ASEAN Member States:
Foreign Value-added in Exports of Manufactured Goods and Services (% share)

60 Manufacturing
Cambodia
M alaysia
50 Vietnam Singapore
Thailand

40

Brunei
30 Philippines

20
Indonesia

10

0
0 10 20 30 40
Services

Source: Data from OECD (2016).

the total imported value of parts and components was twice as much as that of
exports (figure 5).
The data of trade in value-added confirms the somewhat premature nature of
the second unbundling in Indonesia. Figure 6.1 shows the ratio of foreign value-
added to gross exports (of both manufacturing and services). Figure 6.2 shows
the forward and backward participation indexes (FPI and BPI) for Indonesia and
other ASEAN member states. FPI is measured by the share of a country’s export
of goods and services that are used by its trade partners as imported inputs
for production of their goods and services for export. Higher values indicate a
larger contribution of the country’s domestic value-added to its trade partner’s
exports. BPI is the ratio of imported inputs to a country’s total exports. Figure 6.1
shows that the share of foreign value-added in Indonesian gross exports is very
small—less than 10% for services and less than 20% for manufacturing. Figure
6.2 shows that Indonesia has higher FPI values than BPI values, in contrast to the
patterns of neighbouring countries. This implies that Indonesia participates in
GVCs mainly as a material provider and that its role as a downstream producer
is still weak. Most Indonesian manufactured exports comprise resource-based
or low technology-based products, which has the nature of the first unbundling.
For Indonesian manufactured exports, the exchange of parts and components
is not as intensive as in countries such as Singapore, Thailand, or Vietnam. To
178 Fukunari Kimura and Lurong Chen

FIGURE 6.2  Global Value Chain Indicators for ASEAN Member

States: Forward and Backward Particpation (% share)

60 BPI

50 Singapore

40 Vietnam M alaysia Philippines

Cambodia Thailand
30

20

Indonesia
Brunei
10

0
0 10 20 30 40
FPI

Source: Data from OECD (2016).

date, the degree of Indonesia’s involvement in the second unbundling is not as

intense as in these countries.
However, the analysis of extensive margins—i.e. counting exported or imported
product–destination pairs, rather than looking at the value of exports or imports
or at a detailed commodity classification for machinery parts and components—
indicates Indonesia’s deepening involvement in the second unbundling.12 Trade
diversification was recently enhanced for both the exports and imports of parts
and components. Between 2012 and 2015, the number of exported product–
destination pairs of parts and components increased from 15,944 to 17,450,
while that of imported product-source pairs increased from 79,188 to 85,425. The
expansion of extensive margins in exports is mainly driven by increasing trade in
medium technology-based parts and components. Moreover, import data shows
that Indonesia tends to increase its diversification of importing high-technology-
based parts and components, while decreasing its import of low-technology-
based parts and components.

12. See Obashi and Kimura (2017) for the correlation between the dynamics of the second
unbundling and the changes in the number of exported product-destination/imported
product–source pairs of machinery parts and components in ASEAN Member States.
Value Chain Connectivity in Indonesia: The Evolution of Unbundlings 179

FIGURE 7  Trade Driven by Second Unbundling, 2015 (by province)

Source: Data from BPS online database.

Note: Bracketed data is from 2012.
Dark-grey shading is used in areas where international trade is driven by the second unbundling;
light-grey shading is used in areas that have recently become involved in the second unbundling.

Trade in parts and components is geographically concentrated in Jakarta, Riau

Islands, and Surabaya. Together, they accounted for 96% of the country’s total
exports and 88% of its total imports of parts and components in 2012. The combined
share of exports further increased to 97% in 2015, but that of imports declined to
82%. These three areas are shaded dark grey in figure 7. Ports located in Central
Java, Banten, East Kalimantan, and South Sumatra (shaded light grey) have started
to import parts and components. However, these areas have not begun significant
exports of these goods. In 2015, ports in Central Java, South Sumatra, and East
Kalimantan together handled less than 8% of the country’s total imports of parts
and components. Their contribution to exports is more marginal.
Delving deeper into the trade statistics, we find evidence of an evolving pattern
of trade in machinery parts and components trade. Indonesia shipped parts and
components to 189 countries in 2015. Individually, only two very specific products,
‘road wheels or hub caps for tractors’ and ‘wiring harness for motor vehicle’ were
widely shipped to more than the half of the country’s export destinations. For
most products, the number of export destinations was limited to less than 20
countries. In 2015, the top 10 exported parts and components, which accounted
for over 40% of total exports, were shipped to 153 countries or regions.
On the import side, Indonesia’s supply chains of manufactures worldwide
are spread across 157 countries. Products under the category of ‘apparatus for
electrical circuits; board, panels’ and ‘parts and accessories of vehicles’ accounted
 FIGURE 8.1  Number and Value of Parts and Components Exported from Indonesia

300 2,100
Singapore

Japan
Malaysia
Thailand Number of exported products (lhs)
China
United States Value of Indonesian exports (US$ billion; rhs)
200 Hong Kong 1,400
Australia
Korea
Vietnam

100 700
Myanmar

Brunei

Cambodia
Laos PDR
0 0

Source: Data from Comtrade database and BPS online database.

FIGURE 8.2  Number and Value of Parts and Components Exported from Jakarta

300 2,100

Singapore
Japan Number of exported products (lhs)
Thailand
Malaysia
United States
Value of Indonesian exports (US$ billion; rhs)
200 1,400
China
Hong Kong
Vietnam
Australia
India

100 700

Myanmar

Brunei
Cambodia
Laos PDR
0 0

Source: Data from Comtrade database and BPS online database.

FIGURE 8.3  Number and Value of Parts and Components Exported from Riau Islands

300 2,100

Singapore
Number of exported products (lhs)

Value of Indonesian exports (US$ billion; rhs)

200 1,400

Malaysia

United States
100 China 700
Japan
Australia
Thailand
Vietnam
Philippines
Brunei
0 0

Source: Data from Comtrade database and BPS online database.

FIGURE 8.4  Number and Value of Parts and Components Exported from East Java

300 2,100

Number of exported products (lhs)

Value of Indonesian exports (US$ billion; rhs)

200 1,400

East Timor

100 700
Signapore
Japan
Thailand
China
United States

0 0

Source: Data from Comtrade database and BPS online database.

 FIGURE 9.1  Number and Value of Parts and Components Imported by Indonesia
330 Singapore 6,000
China
United States
Japan
Malaysia
Hong Kong Number of imported products (lhs)
Thailand
Value of Indonesian imports (US$ billion; rhs)
220 4,000
Vietnam
Phillipines

110 2,000

0 0

Source: Data from Comtrade database and BPS online database.

FIGURE 9.2  Number and Value of Parts and Components Imported by Jakarta
330 China 6,000
Singapore
United States
Japan
Malaysia
Thailand Number of imported products (lhs)

Value of Indonesian imports (US$ billion; rhs)

220 4,000

Vietnam
Philippines

110 2,000

0 0

Source: Data from Comtrade database and BPS online database.

 FIGURE 9.3  Number and Value of Parts and Components Imported by Riau Islands
330 6,000

Singapore

China Number of imported products (lhs)

United States
Japan Value of Indonesian imports (US$ billion; rhs)
220 Malayasia 4,000

Korea

Thailand
110 2,000

Vietnam
Philippines

0 0

Source: Data from Comtrade database and BPS online database.

FIGURE 9.4  Number and Value of Parts and Components Imported by East Java
330 6,000

China Number of imported products (lhs)

Japan Value of Indonesian imports (US$ billion; rhs)

220 4,000
United States
Singapore

Korea

Malaysia
110 Thailand 2,000

0 0

Source: Data from Comtrade database and BPS online database.

184 Fukunari Kimura and Lurong Chen

for nearly 25% of Indonesia’s total imports of parts and components in 2015.
Imports of parts and components under ‘apparatus’ were from 106 countries and
regions, but most came from the three leading supply countries: China (28%),
Japan (17%), and Singapore (17%). Imports of parts and components under ‘parts’
were nearly of the same scale but with more geographic concentration: out of
83 suppliers, Japan and Thailand were the origin of 44% and 34% of Indonesia’s
imports of auto parts, respectively.
Figures 8 and 9 provide further insights into how far Indonesian parts and
components trade has diversified across foreign markets at both national and
regional levels. Overall, Indonesia’s exports to ASEAN countries and East Asia
are more intensive than those shipped to countries outside the region, in terms of
trade value and trade diversification. Singapore was the most important export
destination, with a large variety and scale of parts and components (288 types
in eight-digit HS code, and a value of more than US$2 trillion), which includes
possible trans-shipments through Singapore. Outside ASEAN, Japan was the
largest destination. More than 140 types of parts and components were exported
to Japan, generating more than US$1.4 trillion trade revenue in 2015. China,
excluding Hong Kong, was the fifth largest market for Indonesia’s exports of
parts and components, in terms of variety, behind Singapore, Japan, Malaysia,
and Thailand. However, the scale of exports to China was only about 25% of that
exported to Japan.
In 2015, over 60% of Indonesian exports of parts and components were
shipped from Jakarta, mainly via Tanjung Priok Port and Soekarno-Hatta Airport.
Exports of parts and components are generally intra-regional. Singapore was
the immediate destination for the largest variety of products; but in terms of
the trade value, Thailand, Japan, and Malaysia were the three top destinations
from these ports. The performance of manufacturing and therefore exports from
Jakarta tends to be closely linked with these three trade partners—with a wide
scope and on a large scale.
On the import side, China and Japan were the two largest suppliers of parts
and components to Indonesia. From ASEAN countries, over 80% of imports were
from Singapore and Thailand. The share of trade with Vietnam has substantially
increased—from 1.5% in 2012 to almost 5% in 2015. Jakarta, Riau Islands, and
East Java are also gateways for foreign-supplied parts and components that enter
Indonesia. Over 60% of parts and components imported by Indonesia in 2015
were destined for Jakarta.
Surabaya is the country’s third-largest hub of trade for parts and components.
It lags behind Jakarta and Riau Islands in terms of scale of trade and export
diversification. The main target market of the parts and components exported
through Surabaya is Japan, which absorbed over 50% by value.
The findings here are in line with the previous observations that Jakarta and
Surabaya are the two industrial production centres of Indonesia. Ports near these
cities handle most of the container boxes for international trade. The agglomeration
of manufacturing clusters and the relatively easy access to transportation contribute
to their roles as hubs for international and domestic economies.
Despite its relatively small economic size, the Riau Islands province is another
hub of international trade in Indonesia. In 2015, more than 25% of Indonesia’s total
exported parts and components and 15% of total imported parts and components
Value Chain Connectivity in Indonesia: The Evolution of Unbundlings 185

were shipped from and to Riau Islands. Singapore is Riau Islands’ largest trade
partner for both immediate imports and exports. Riau Islands has a trade surplus
with both Singapore and Malaysia. To Malaysia, Riau Islands is a net exporter
of final products but a net importer of parts and components. To Singapore,
Riau Islands is a net exporter in both categories. Between Singapore and Riau
Islands, is a two-way flow of resources. In 2015, Singapore imported more than
250 types of parts and components from Riau Islands directly, which represented
nearly 66% of bilateral exports. Shipments between Singapore and the ports in the
Riau Islands comprised over 50% of Singapore’s exports to Indonesia and nearly
66% of its imports of parts and components from Indonesia. Riau Islands also
plays a significant role in bilateral trade with Malaysia, accounting over 15% of
exports and nearly 25% of imports. Overall, the size of Riau Islands–Johor trade
is equivalent to less than the one-tenth of that of Riau–Singapore trade. In this
regard, although SIJORI is said to be a ‘triangle’ as it functions more like as a
combination of two bilateral links (Singapore–Riau Islands and Singapore–Johor)
in the hub-and-spoke system.
Indonesia’s performance in the first and second unbundling suggests that the
lack of efficient connectivity has restricted it from unleashing its potential to
maximise gains from the first two waves of globalisation. In particular, if the direct
or indirect commitment of foreign multinational enterprises is almost essential
to industrialisation, insufficient connectivity is fatal. Furthermore, the second
unbundling imposes even higher requirements of connectivity. For Indonesia, it
turns out that so far only few regions with relatively high accessibility have been
able to join international production sharing.
Three sources of inefficient connectivity exist in Indonesia — poor physical
infrastructure, insufficient services, and the policy environment. In World
Economic Forum’s Competitiveness Index, Indonesia’s maritime infrastructure
is ranked 12 out of the 14 Asian peers. The cost of logistics in Indonesia remains
high. In international trade, this implies extra trade costs and excessive shipping
time. Many Indonesian ports still need investment and further efforts to deal with
problems such as inadequate capacity and navigational aids, bunching of vessels,
limited cargo-handling facilities, high downtime of equipment, and lack of storage
space. The excessive demand for port services has caused congestion at ports and
therefore higher costs and longer transportation times. A survey by the Center
of Logistics and Supply Chain Studies, Badung Institute of Technology (ITB),
Asosiasi Logistik Indonesia, STC-Group, and the World Bank (2015) shows that
Indonesia was experiencing worsening traffic congestion between warehouses/
factories and seaports. The survey also notes that some government regulations
introduce unnecessary barriers to entry for end-to-end logistics services.13 For
example, Indonesia’s manufacturing sector finds it difficult to provide all-in-
one logistics services because it requires four separate licences (trucking, freight
forwarder, express delivery, and warehousing) to cover the whole process.

13. See also Pangestu, Rahardj, and Ing (2015).

186 Fukunari Kimura and Lurong Chen

FIGURE 10  Firm-level Technology Adoption in Indonesia

and Selected ASEAN Member States
5.5

5.0

Philippines
4.5
Indonesia
Thailand
World average
4.0 Vietnam

3.5
2011 2012 2013 2014 2015 2016

Source: Data from the WEF (2017).

Note: Technology-use values (1 = not at all; 7 = to a great extent) based on WEF survey responses.

Prospect for the third unbundling

The third unbundling as a form of the international division of labour, not directly
observed by statistics, is probably still small in scale in less-developed countries.
However, the basic elements of the third unbundling are already being prepared,
also in Indonesia. The reduction in face-to-face costs has started to change daily
life through the internet and smartphones.
To what degree will Indonesia harness benefits from the digital economy? A
natural concern is its shortage of human capital for service providers, system
engineers, and programmers whose work realises the lower face-to-face costs.
However, users of face-to-face connections do not necessarily need specialised
education or high levels of human capital. Platforms and applications offer
services that can be accessed by relatively small participants in business, such as
small businesses and especially individual consumers. The explosive expansion
of participation in various platforms is the very nature of the digital economy,
and the network externalities allow platform businesses to become economically
viable. In this sense, Indonesia has good potential to utilise the digital economy.
The country has the world’s fourth largest population. And, different from other
populous countries such as China, about the half of the population is under 30.
Indonesia’s middle class is expected to expand to around 20 million households,
with a disposable annual income of $11,300 or above per household by the end
FIGURE 11.1  Information and Communications Technology Use in B2B Transactions
6.0 Singapore
Malaysia

5.5
Vietnam
Brunei Philippines
Indonesia
China
5.0
Thailand
World average (4.73)
4.5 Cambodia India
Laos PDR

4.0

Myanmar
3.5

3.0

Source: Chen (2017). Data from WEF (2017); Executive Opinion Survey (2016).
Note: Technology-use values (1 = not at all; 7 = to a great extent) based on WEF survey responses.

FIGURE 11.2  Information and Communications Technology Use in B2C Transactions

6.0
Malaysia

Singapore
5.5 Indonesia
China
Vietnam
5.0
Philippines
Thailand
Brunei
4.5 World average (4.42)
Laos PDR
India
4.0

Cambodia
3.5
Myanmar

3.0

Source: Chen (2017). Data from WEF (2017); Executive Opinion Survey (2016).
Note: Technology-use values (1 = not at all; 7 = to a great extent) based on WEF survey responses.
188 Fukunari Kimura and Lurong Chen

FIGURE 12  Industrial Dynamism among Unbundling Regimes

Pre-globalised First Second Third

world unbundling unbundling unbundling
(0) (1) (2) (3)
‘Step-by-step’: sequential
upgrading of unbundlings

Export-based agriculture & fishing (0–1)

‘Leapfrogging’: bypassing
Service outsourcing, B2C & C2C (0–3)
of some regimes
AI-controlled farming & mining (1–3)

Internet of Things (3–2)

3D printers in garment industry (3–1)

‘Feedback’: possible
introduction of advanced Smartphones in agriculture (3–0)
piecemeal technologies to
old industries SCM (2–1)

Food value chains (2–1)

Large-scale agriculture (2–1)

of 2020 (World Bank 2017). These are positive factors for Indonesia to develop
its digital economy. It is assessed as having a relatively high speed of technology
adoption. According to the World Economic Forum (2017), the speed of firm-level
technology adoption in Indonesia is higher than the world average and higher than
many of Indonesia’s neighbouring countries (figure 10).
Although it is difficult to measure the current quantitative significance of the
digital economy at the macro level in Indonesia, it is obvious that various new
businesses have already been mushrooming in Indonesia.
First, people are connected through internet and social media. The Indonesia
Internet Service Provider Association (APJII) survey revealed that, by the end of
2017, the number of internet users in Indonesia reached 143.3 million (54.7% of
total population; a 7.6% increase from the previous year), of which 87.1% are active
social media users. There are more than 50 smartphones and tablets and more
than 25 personal computers for every 100 Indonesians. Users have relied more on
smartphones or tablets (83.4%) than laptops and desktop computers (43.8%) to
access the internet. We Are Social (2018) reports that as of January 2018, the share
of web traffic by device is as follows: laptops and desktops, 26% (-8% from a year
before); mobile phones, 72% (+5%); and tablet devices, 2% (-34%). For the frequency
of internet use for personal reasons, 79% of internet users go online ‘every day’, 14%
‘at least once a week’, 6% ‘at least once a month’, and 1% ‘less than once per month’.
Although connection speed may sometimes be a problem, Indonesians are very
Value Chain Connectivity in Indonesia: The Evolution of Unbundlings 189

familiar with the internet and social media. Urban–rural gaps in digitisation exist,
but they seem to be smaller than those of many other aspects of development. The
ratios of internet-user penetration are 72.4% in urban areas, 49.4% in rural–urban
areas, and 48.2% in rural areas. By island, the penetration ratio is 57.7% in Java,
54.2% by Bali-Nusa, 47.2% in Sumatra, 72.1% in Kalimantan, 46.7% in Sulawesi,
and 41.9% in Maluku-Papua (APJII 2018).
Second, old existing industries and businesses can be revitalised by reduced
face-to-face costs. E-commerce goes with, and beyond, traditional wholesale and
retail services. According to We Are Social (2018), as of January 2018, the total
number of people purchasing consumer goods in Indonesia via e-commerce was
28 million (+13% from the previous year), and the value of the consumer goods
e-commerce market (total annual sales revenue) amounted to US$7 billion (+22%
from the previous year). For e-commerce activities, APJII (2018) shows that 45% of
internet users searched and compared prices online (‘cari harga’), 37.8 % of internet
users collected information online before making purchase (‘informasi membeli’),
32.2% of internet users went shopping online (‘beli online’), and 16.8% of internet
users sold goods online (‘jual online’). Leading online e-commerce marketplaces
such as Lazada, Blibli, Zalora, Bhinneka, MatahariMall, Traveloka, Tiket.com,
and Pegipegi began to collect and integrate information from various sources and
provide users with service packages that include a transaction credit system to
help buyers and sellers gain trust. There are also trials to launch cross-border
insurance products to facilitate the transaction process. Transportation services
are another sector that has been further activated by the reduction in the matching
cost between businesses and consumers and between consumers and consumers.
The proliferation of companies such as Go-Jek, Grab, and Uber has been significant,
largely due to mobile connectivity. The adoption of social media, together with
online digital service platforms, such as WhatsApp, Line, Go-Jek, and Uber, start
to work on the soft infrastructure of connectivity and provide technical support to
supply chains by improving services with mega e-fulfilment centres, parcel sorting
centres (hubs), local parcel distribution centres for last-mile supply chains, local
city logistics depots, and returns centres (Financial Times 2015). Indonesia has a
higher score than many other Asian countries in the index of ICT use in B2B and
B2C business transactions. Particularly in B2C transactions, the popularity of ICT
in Indonesia is already very close to the level of Singapore or Malaysia (figure 11).
Third, the possibility of the services outsourcing via distance is expanding, and
the seed of the third unbundling emerges. In other parts of the world, working
at home or telecommuting has gradually become more popular, although much
of this work is still ‘domestic’ rather than ‘cross-border’. Various kinds of job-
matching platforms for employers and service providers have recently been
developed, which include Upwork, Witmart.com, and Amazon Mechanical Turk.
Although the penetration of service outsourcing may not yet be conspicuous in
Indonesia, it is believed to be just a matter of time before it expands.
The important factors in this discussion are connectivity, services, and the
policy environment. As for connectivity, fast internet connection is important, and
physical logistics are still key to expand the scope of people-to-people matching and
e-commerce. Services are at the core of the digital economy and the forthcoming
third unbundling. Policy environment for creative businesses is essential, together
with necessary back-up policies.
190 Fukunari Kimura and Lurong Chen

POLICY IMPLICATIONS
Three levels of unbundling can co-exist. Depending on industries, corporate
strategies, human capital, and other factors, firms will choose the optimal way of
unbundling. However, the extent of unbundling also depends on geographical
location, infrastructure, and the policy environment. In particular, a large country
such as Indonesia must think of expanding the capability of using higher levels
of unbundling.
Indonesia and its various provinces have great potential to upgrade their
capabilities to optimise higher levels of unbundling. Remote islands that still
operate in a pre-globalised context can initiate benefits from the first unbundling
by having a medium-quality logistics infrastructure and a conducive policy
environment. In particular, cities with relatively large populations such as Medan
and Makassar seem to have a good potential for the first unbundling, specifically
in labour-intensive manufacturing. Some areas in outer islands may even move
into the second unbundling with a higher-quality logistics infrastructure. Java can
deepen its involvement in the second unbundling by promoting the formation
of efficient industrial agglomeration. These traditional aspects of sustaining
economic growth are still relevant in much of Indonesia.
In addition, it is imperative for Indonesia to consider how to incorporate the
digital economy in the framework of its development strategies. The sound basis
of the first and second unbundling in terms of infrastructure, policy environment,
and human capital will help to expand the scope of the digital economy. But
an aspect that requires particular consideration in the coming era of the third
unbundling is human resources, which may become a considerable challenge for
the advanced portion of Indonesian society, given the existing challenges.14 For
the operation of production in the second unbundling, the shortage of middle
managers and engineers is a typical problem in effectively utilising abundant
labour and upgrading tasks in production networks. For the third unbundling,
programmers and entrepreneurs for platform providers will obviously be
essential. In addition, individuals may have to be qualified as independent
service providers in both domestic and international matchings. What sort of
human capital should be nurtured is a challenging question. Large and young
population in Indonesia must be utilised as a source of competitiveness.
The digital economy provides a novel angle for inclusive growth. The internet
can be accessed not only by influential and big participants in society and business
but also by small businesses and individual consumers. Even if a region does
not achieve a high level of industrialisation with physical connectivity, it may
have opportunities to leapfrog to a new paradigm of the third unbundling and
conduct service outsourcing. In addition, piecemeal technologies in the digital
economy are already finding applications in traditional sectors such as wholesale
and retail trade, transportation, tourism, and other sectors. Indonesia should find
ways to take advantage of various aspects of the digital economy.

14. Suryadarma and Jones (2013) provide a comprehensive review on education in Indonesia.
Value Chain Connectivity in Indonesia: The Evolution of Unbundlings 191

REFERENCES
Accenture. 2016. ‘Five Ways to Win with Digital Platforms’. Accessed 19 November 2017.
www.accenture.com/us-en/_acnmedia/PDF-29/accenture-five-ways-to-win-with-digital-
platforms-full-report.pdf.
Akamatsu, Kaname. 1962. ‘A Historical Pattern of Economic Growth in Developing
Countries’. The Developing Economies, 1(1): 3–25.
Asosiasi Penyelenggara Jasa Internet Indonesia (APJII). 2018. ‘Inforgrafis: Penetrasi &
Perilaku Pengguna Internet Indonesia 2017 Survey’. Accessed 29 May 2018. https://apjii.
or.id/.
Baldwin, Richard. 2006. ‘Multilaterizing Regionalism: Spaghetti Bowls as Building Blocks
on the Path to Global Free Trade’. The World Economy, 29(11): 1451–1518.
Baldwin, Richard. 2016. The Great Convergence: Information Technology and the New Globalization.
Cambridge, MA: The Belknap Press of Harvard University Press.
The Center of Logistics and Supply Chain Studies, Badung Institute of Technology
(ITB), Asosiasi Logistik Indonesia, STC-Group, and the World Bank. 2015. State
of Logistics Indonesia 2015. Accessed 12 May 2018. www.nestra.net/download/
StateofLogisticsIndonesia2015.pdf.
Chen, Lurong. 2017. Developing Asia in the Era of Cross-border E-commerce, ERIA Discussion
Paper series, ERIA-DP-2017–11.
Courant, Paul, and Alan Deardorff. 1992. ‘International Trade with Lumpy Countries’.
Journal of Political Economy 100 (February): 198–210.
Financial Times. 2015. ‘Go-Jek: Opening The Throttle in Indonesia’. Accessed 16 February
2018. www.ft.com/ content/d774419c-8a0f-11e5-9f8c-a8d619fa707c.
Findlay, Christopher, and Mari Pangestu. 2016. ‘The Services Sector as a Driver of Change:
Indonesia’s Experience in the ASEAN Context’. Bulletin of Indonesian Economic Studies
52(1): 27–53.
Garnaut, Ross. 2015. ‘Indonesia’s Resources Boom in International Perspective: Policy
Dilemmas and Options for Continued Strong Growth’. Bulletin of Indonesian Economic
Studies 51(2): 189–212.
Hill, Hal, Budy Resosudarmo, and Yogi Vidyattama. 2008. ‘Economic Geography of
Indonesia: Location, Connectivity, and Resources’. In Yukon Huang, Alessandro Magnoli
Bocchi, eds., Reshaping Economic Geography in East Asia, Washington DC: The World Bank.
Jones, Ronald, and Henryk Kierzkowski. 1990. ‘The Role of Services in Production and
International Trade: A Theoretical Framework’, in Ronald Jones and Anne Krueger, eds.,
The Political Economy of International Trade: Essays in Honor of Robert E. Baldwin, Oxford:
Basil Blackwell: 31–48.
Kimura, Fukunari. 2018. ‘“Unbundlings” and Development Strategies in ASEAN: Old Issues
and New Challenges’. Journal of Southeast Asian Economies, 35(1): 13-21.
Kimura, Fukunari and Mitsuyo Ando. 2005. ‘Two-dimensional Fragmentation in East Asia:
Conceptual Framework and Empirics’ in H. Kierzkowski (ed.), International Review of
Economics and Finance (special issue on ‘Outsourcing and Fragmentation: Blessing or Threat’,
14(3): 317–348.
Kimura, Fukunari and Mateus Silva Chang. 2017. ‘Industrialization and Poverty Reduction
in East Asia: Internal Labor Movements Matter’. Journal of Asian Economics 48: 23–47.
Kimura, Fukunari, Tomohiro Machikita, and Yasushi Ueki. 2016. ‘Technology Transfer
in ASEAN Countries: Some Evidence from Buyer-Provided Training Network Data’.
Economic Change and Restructuring, 49(2): 195–219.
Kojima, Kiyoshi. 2000. ‘The “Flying Geese” Model of Asian Economic Development: Origin,
Theoretical Extensions, and Regional Policy Implications’, Journal of Asian Economics, 11:
375–401.
Kumagai, Satoru. 2008. A Journey through the Secret History of the Flying Geese Model, JETRO-
IDE Discussion Papers (158).
192 Fukunari Kimura and Lurong Chen

McKinsey Global Institute (MGI). 2012. The Archipelago Economy: Unleashing Indonesia’s
Potential. Accessed 23 June 2017. www.mckinsey.com/featured-insights/asia-pacific/
the-archipelago-economy.
Obashi, Ayako and Fukunari Kimura. 2017. ‘Deepening and Widening of Production
Networks in ASEAN’. Asian Economic Papers 16(1): 1–27.
Organisation for Economic Co-operation and Development (OECD). 2016. Trade in value
added (Edition 2016). Paris: OECD.
Pangestu, Mari. 2010. The Challenges for Trade Policy in a Dynamic World and Regional Setting:
An Indonesian Perspective, Richard Snape Lecture Series, Melbourne: The Productivity
Commission.
Pangestu, Mari, Sjamsu Rahardj, and Lili Yan Ing. 2015. ‘Fifty Years of Trade Policy in
Indonesia: New World Trade, Old Treatments’. Bulletin of Indonesian Economic Studies
51:2, 239–261.
Suryadarma, Daniel and Gavin W. Jones, eds. 2013. Education in Indonesia. Singapore:
Institute of South East Asian Studies.
Van der Eng, Pierre. 2015. ‘Mixed Blessings: Mining in Indonesia’s Economy, 1870–2010’.
In Marc Badia-Miro, Vincente Pinilla and Henry Willebald, eds., Natural Resources and
Economic Growth: Learning from History, Abingdon: Routledge: 226–247.
We Are Social. 2018. Digital in 2018 in Southeast Asia. Accessed 26 May 2018. https://
digitalreport.wearesocial.com/
World Economic Forum. 2017. The Global Competitiveness Report 2016–2017. Geneva.
Woo, Wing Thye and Chang Hong. 2010. ‘Indonesia’s Economic Performance in Comparative
Perspective and a New Policy Framework for 2049’. Bulletin of Indonesian Economic Studies
46(1): 33–64.
World Bank. 2017. World Development Indicator 2017. Washington DC.
Zen, Fauziah, Khoirunnurofik, and Muhammad Halley Yudhistira, eds. Forthcoming.
Maritime Connectivity in South East Asia: Its Role and Challenges towards Integration. Jakarta:
ERIA.
 Journal of Contemporary Asia

ISSN: 0047-2336 (Print) 1752-7554 (Online) Journal homepage: https://www.tandfonline.com/loi/rjoc20

Indonesia’s Quest for Food Self-sufficiency: A New

Agricultural Political Economy?

Natasha Hamilton-Hart

To cite this article: Natasha Hamilton-Hart (2019): Indonesia’s Quest for Food Self-
sufficiency: A New Agricultural Political Economy?, Journal of Contemporary Asia, DOI:
10.1080/00472336.2019.1617890

To link to this article: https://doi.org/10.1080/00472336.2019.1617890

Published online: 02 Jul 2019.

Submit your article to this journal

Article views: 23

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=rjoc20
JOURNAL OF CONTEMPORARY ASIA
https://doi.org/10.1080/00472336.2019.1617890

Indonesia’s Quest for Food Self-sufficiency: A New

Agricultural Political Economy?
Natasha Hamilton-Hart
Department of Management and International Business, University of Auckland Business School, University
of Auckland, New Zealand

ABSTRACT KEYWORDS
The aspiration to achieve food self-sufficiency has returned to the Indonesia; agriculture; food
Indonesian policy agenda in the post-reformasi period. After being self-sufficiency; land
side-lined as a policy goal immediately after the transition to democ- ownership; reformasi
racy, political leaders since 2009 have increasingly cited self-
sufficiency to justify policies to raise domestic food production and
reduce food imports. These policies are often inefficient and at odds
with the goals of food security and decreasing poverty. The self-
sufficiency drive is thus often attributed to ‘political’ motives. But
what kind of politics is at work in Indonesia’s renewed self-sufficiency
drive: a broad-based politics that aims to foster support among
smallholder food producers, or an elite politics based on rent-
seeking by narrow constituencies? The distributional implications of
food self-sufficiency policies in the post-reformasi period suggest that
both dynamics are at work. Narrow elite interests have benefited, but
smallholder landowners account for a significant share of many
agricultural crops. Although smallholder landowners are increasingly
differentiated, they constitute a reasonably broad-based constitu-
ency. The valorisation of agriculture reflected in the self-sufficiency
drive and related rural sector policies may lay the foundations for
a more inclusive agricultural political economy.

The idea that Indonesia should become self-sufficient in food has been politically influential
since the country’s independence. After being given reduced priority in the early reformasi
period, self-sufficiency returned to prominence in the second term of President Susilo
Bambang Yudhoyono (2009–2014) and was pursued as a policy goal of the Joko Widodo
(Jokowi) presidency since 2014. Jokowi’s championing of food self-sufficiency is congruent
with his nationalist-developmental political strategy (Warburton 2016). This developmen-
tal strategy is itself consistent with an “illiberal turn” towards populist and economic
nationalist postures by Jokowi and his rivals (Aspinall 2015; Mietzner 2015; Hadiz 2017).
The return of food self-sufficiency as a policy goal raises a series of questions: what
kinds of interventions are being adopted, at what cost, and with what effects? Which
domestic interests does the self-sufficiency drive serve? This article assesses the self-
sufficiency drive in terms of its potential to consolidate a structural shift in the political
economy of agriculture in Indonesia. The valorisation of agriculture implied by the

CONTACT Natasha Hamilton-Hart n.hamilton-hart@auckland.ac.nz Natasha Hamilton-Hart, Department of

Management and International Business, University of Auckland Business School, University of Auckland, New Zealand.
© 2019 Journal of Contemporary Asia
2 N. HAMILTON-HART

food self-sufficiency programme represents a new turn in Indonesian development

thinking that is in line with moves in the global development community to value
agriculture as a progressive sector “for development” (World Bank 2007). Agriculture-
led development has the potential to lay the foundations for more inclusive political
economies (Trampusch and Spies 2014). Such a transformation, however, depends on
political context and structural features of the agricultural sector, including the dis-
tribution of landholdings, the organization of agricultural interests and the rights of
agricultural labour.
This article shows that Indonesia’s food self-sufficiency programme, along with
associated moves in agricultural and land policy, has inclusive potential. However,
inclusive elements of current policy confront the legacies of structural forces in the
agricultural sector and the political conflicts of the 1960s, which ended in the defeat of
leftist movements that called for land reform. Parts of the most recent self-sufficiency
programme can be read as attempts to overturn that legacy, but these attempts are
cautious and limited by countervailing forces. Although the meaning of food self-
sufficiency is contested, no element of the Indonesian self-sufficiency programme
represents a liberalizing force. Rather, food self-sufficiency policies represent
a confluence of illiberal currents in Indonesia’s political economy. The goal of food self-
sufficiency is for a country as a whole to produce sufficient staple foods to meet national
consumption (Agarwal 2014). That is, self-sufficiency aims to end most food imports. It
presumes that access to food is best achieved through national production and is
founded conceptually on distrust of international markets. It represents an illiberal
turn in economic policy thinking, favouring state intervention as a bulwark against free
market fluctuations (Neilson and Wright 2017).
Food self-sufficiency policies routinely attract the criticism that they are protectionist
rent-generating programmes operating under the cloak of nationalism. Their often
perverse implications mean that critics extend well beyond liberal economists like
Warr (2011) to include a variety of political perspectives and disciplinary orientations
such as McCarthy and Obidzinski (2017) and Neilson (2013). This article shows that
the self-sufficiency programme has created substantial rents through interventions to
subsidise agricultural production and that these rents have increased since 2009,
particularly since 2014. When it comes to identifying who has captured the rents
associated with food self-sufficiency policies, there is evidence that they have been
distributed more widely than some critics charge: large, elite-owned agribusiness
investors have benefitted, but so have many smallholders, including a set of “middle
farmers” who have consolidated their position.
The following section reviews the trajectory of food self-sufficiency as a policy goal
and discusses the potential for a renewed emphasis on agriculture to lay the foundations
for a more inclusive political economy. This is followed by a section that describes the
consequences of the self-sufficiency drive for food production, food imports and food
affordability. A subsequent section investigates patterns of landholding in agriculture,
presenting evidence from the post-reformasi period of an increasingly differentiated
smallholder sector that has developed alongside large-scale agribusiness interests. The
final section discusses the underlying political economy of agriculture in Indonesia,
drawing out the implications of long-term structural trends and political contests over
agricultural land.
 JOURNAL OF CONTEMPORARY ASIA 3

Food Self-Sufficiency and the Political Economy of Agriculture

Food self-sufficiency as a policy goal calls for the elimination of most food imports. It
reflects a rejection of economic liberalism’s call for free, internationalised markets. The
term is often conflated with the notions of food security – defined as access to sufficient
food at affordable prices – and food sovereignty, understood as a call for localised,
democratic empowerment of food producers. These agendas represent distinct and
often competing goals (see Agarwal 2014). Food self-sufficiency in the Indonesian
policy context is a national-level objective that is silent about distribution (McCarthy
and Zen 2013). Although it is often associated with ideas of improved outcomes for
poor farmers, the unit that is to be self-sufficient is the nation (McCarthy and
Obidzinski 2017, 345). Individual access to food, food affordability or other domestic
distributional considerations are relegated to the goal of increased national production
and reduced reliance on imported food (Neilson 2013).
The national-level framing of food policy means it has the potential to serve
a narrow set of elite agribusiness interests that have invested in large food estates
promoted under the self-sufficiency umbrella while further marginalising small-scale
farmers (McCarthy and Obidzinski 2017). If this trajectory is underway, the latest
iteration of food self-sufficiency policy would represent a continuation of the privileging
of elite interests and large-scale capital formation that has been fundamental to
Indonesia’s political economy since the advent of Suharto’s New Order in the mid-
1960s and which persisted in the form of oligarchic capitalism in the new landscape of
post-reformasi democracy (Robison 1986; Robison and Hadiz 2004; Winters 2013). In
this political economy, the structure of protection and the thrust of development policy
was biased in favour of manufacturing for much of the time (Fane and Warr 2008).
Indonesian policymakers appear to have viewed development as structural change out
of agriculture and towards the acquisition of manufacturing capacity. The targets of
“rural development” policy often viewed the process as a top-down extraction of
resources from rural areas by a centralising and repressive state (Dove and Kammen
2001). In the post-democratisation era, rural development “projects” proliferate, but
a long-term observer of rural livelihoods contends that they neglect the “core political-
economic questions concerned with the distribution of the means of rural livelihood”
such as land, jobs and wages, hence “set out to reduce poverty without addressing the
processes through which poverty was systematically produced” (Li 2016, 79).
There is much in Indonesia’s food self-sufficiency programme that reinforces these
views of a political economy driven by elite business interests working with state
powerholders. However, this article finds that there is also inclusive potential in the
current valorisation of agriculture and smallholders in the latest food self-sufficiency
drive. To the extent that smallholder interests have benefited both economically and
politically, it is possible that the socio-economic foundations of a more inclusive
political economy could emerge, helped by self-sufficiency policies. The potential for
agriculture to support both economic growth and empowerment of rural populations
has returned to development thinking (World Bank 2007). This reflects somewhat
belated recognition of the positive role that agricultural interests played in the devel-
opment of inclusive and economically dynamic political economies in some Western
countries as well as in Asia’s first and second generation of rapid industrialisers. A key
4 N. HAMILTON-HART

structural condition for this transformative potential for agriculture has been the
emergence of relatively small-scale farming operations, often organised in co-
operative ways (Trampusch and Spies 2014, 923). Labour-repressive agricultural poli-
tical economies in which large landholdings predominate have been associated with
support for authoritarian politics, but agricultural sectors in which land is distributed
more equally have been forces for democratisation (see Huber, Rueschemeyer, and
Stephens 1993, 76, 85). In many cases, a more equal agricultural structure was the result
of forceful land reform, to break large landholdings held by a rentier class. Such land
reform programmes in Japan, South Korea and Taiwan underlay the “Asian miracle,”
which rested on credible promises of shared growth (Campos and Root 2001). In
contrast, Indonesia and the other capitalist economies of Southeast Asia did not
experience any such fundamental restructuring of rural ownership rights, to the detri-
ment of their long-term abilities to generate dynamic, inclusive growth (Studwell 2013,
20–63). Rather than succeeding in comprehensive land reform, leftist movements that
challenged incumbent elites and called for land redistribution were forcefully repressed,
laying the foundations for Southeast Asia’s unequal development trajectories
(Hamilton-Hart 2012, 48–84). In Indonesia, this process was violent and extensive,
involving the military-orchestrated massacre of around half a million people over 1965–
1967 and extended incarceration and repression of many more.1
The socio-economic conflict that ended with the defeat of the left in the 1960s was in
many ways a rural conflict. In contrast to widespread visions of a homogenous and
cohesive rural sector, a recent overview emphasises that:

30–50 percent of the population of Java was already landless in the nineteenth century, and
there were deeply entrenched social divides. Villages were stratified into caste-like estates,
in which landholding families organised production by incorporating landless farm ser-
vants as permanent dependents and employed roving bands of “free” coolie labour, whose
livelihoods were anything but secure (Li 2009a, 632).

It was this agricultural political economy that Indonesia’s peasant-based communist

movement challenged, ultimately through unilateral land seizures in the early 1960s (White
2016). The New Order’s foundational objective was to end such grassroots mobilisation. In
addition to the mass killings of the 1960s, independent farmer organisations were banned. As
part of a general policy of bringing mass organisations into state-backed corporatist groups,
a single, state-controlled farmers’ organisation, the Indonesian Farmers’ Association
(Himpunan Kerukunan Tani Indonesia, HKTI) was established in 1973 to ensure the rural
sector remained politically acquiescent (Peluso, Afiff, and Rachman 2008, 382). Landlessness
among Javanese peasants increased dramatically in the 1970s, due to repression targeting
rural mobilisation (Farid 2005, 10–12).
It is important to recall this history of conflict and repression when tracing the
evolution of food self-security in Indonesia. In different periods, the policy has been
associated with different political emphases. Indonesia’s first president, Sukarno,
asked: “Why bother talking about political freedom if we don’t have freedom to
manage our rice, and always have to beg for help buying rice from our neighbouring
nations?” (cited in Neilson 2013). The main plank of Sukarno’s food self-sufficiency
policy was the Padi Centra Programme in Java, aiming to use the co-operative
movement to provide fertilisers, seeds and credit to spur rice production. Its failure
 JOURNAL OF CONTEMPORARY ASIA 5

was attributed to “the program’s attempt at widespread implementation along with

lack of funding, poor logistics and lack of trained and experienced staff” (Mears
1984, 125). It also took place in the context of rising political tension over land and
related peasant mobilisation, associating this first foray into food self-sufficiency with
redistributionist conflict.
Under the New Order, food production was prioritised on the understanding that
access to rice was politically crucial for regime survival (McCarthy and Zen 2013;
Nielson 2013). But this did not mean empowering food producers. Suharto’s govern-
ment began in the late 1960s to invest heavily in rice production and interventions to
stabilise the price of rice. This included mobilisation of capital inputs to raise yields
during the Green Revolution of the 1970s, but the imperative to ensure rice availability
meant that the government-controlled logistics agency, Bulog, became important as
a rice importer, buyer and distributor (Timmer 1996, 61). Subsidies to rice producers
were balanced through price interventions by Bulog, ensuring support for industrialisa-
tion efforts (McCulloch 2008). Bulog oversaw rice price stabilisation following a food
crisis and riots in 1974 and frequently intervened in rice markets. Through the 1970s,
rice self-sufficiency was secondary to the goal of availability and price stabilisation
through Bulog’s rice and wheat imports (Mears 1984, 132). Throughout the period of
authoritarian rule until partial liberalisation in the early 1990s, government rice policy
enabled intensive intervention in rural life and the distribution of subsidies cemented
patterns of patronage and control at the village level (Patrick 2004, 23).
The economic crisis of 1997–1998 meant that many people experienced food insecurity.
When the country was forced to borrow from the International Monetary Fund (IMF),
a common goal of the IMF and Indonesia’s economic technocrats was to end Bulog’s
import monopoly and liberalise the domestic rice market. A policy of free trade in rice was
maintained until 2004, when trade minister Rini Soewandi introduced a “seasonal” import
ban, which was then extended (McCulloch 2008, 46). Officials justified restrictions on
importing rice on the grounds that they wished to protect poor farmers from the effects of
cheap imported rice (Warr 2011, 6). President Yudhoyono elevated food self-sufficiency as
a national priority in the wake of the global food crisis of 2007–2008, which sparked unrest
in many countries as a result of a sharp rise in the price of staple foods, exacerbated by
export restrictions by key producers such as Vietnam (ADB 2012).
Following a surge in the cost of rice imports in 2011, the government introduced
a new Food Law in 2012, which formalised the goals of “food sovereignty” (kedaulatan
pangan), “food self-sufficiency” (kemandirian pangan) and “food security” (ketahanan
pangan). Although the law defines these separately,

the terms are, without exception, referred to collectively throughout the law. As such, the
Law conflates all three terms into a single objective. ‘Food sovereignty’ is employed
rhetorically in the law to strengthen the case for national scale food self-sufficiency to be
achieved through state interventions (Neilson and Wright 2017, 137).

In application, national-level self-sufficiency goals have dominated, sometimes at the

expense of local livelihoods (McCarthy and Obidzinski 2017).
President Yudhoyono laid out the goals of achieving self-sufficiency in six staple
foods, including rice, by 2014. President Joko Widodo has moved further to promote
food production and control food imports. Early in his presidency, Jokowi was reported
6 N. HAMILTON-HART

as saying that, “We do not want to import rice any longer. If we were self-sufficient [in
food production] there would be no more sugar or soy bean imports” (Jakarta Post,
January 31, 2015). The occasion was a visit to a rural area in East Java, where Jokowi
presented farmers with 852 hand tractors and 377 water pumps. He announced that the
government aimed to distribute farm equipment to 500 regencies during 2015, in order
to boost production. A protectionist, pro-farmer orientation was in line with his 2014
electoral campaign, which contributed to the “hyper-nationalist” atmosphere generated
by Jokowi and his rival for the presidency, Prabowo Subianto, who both expressed
illiberal views about the role of foreigners in the national economy and the free market
more generally (Aspinall 2015, 14–17). Under Jokowi, the self-sufficiency programme
has extended beyond import controls to include an increase in producer subsidies, land
allocation and state interventions that have given central government administrative
and enforcement agencies new discretionary authority.
Agricultural subsidies and protection afforded producers have increased dramatically
since 2009. These rents come in the form of producer subsidies (particularly for fertiliser,
credit and seeds), price supports (mandated price floors) and tariff protection. The
Organisation for Economic Co-operation and Development (OECD) calculates the com-
bined level of these interventions to generate a measure of producer support, which is
expressed as a percentage of gross farm receipts. This ratio of producer support averaged
10% of farm receipts between 2000 and 2007, rising to 17% in 2009–2014, reaching 29% in
2015 (OECD 2017b). This is well above the OECD average of 17% and, as a percentage of
GDP (4%), it is the highest level of agricultural protection recorded in the OECD survey of
developed and developing countries (OECD 2017a). The amounts involved represent sub-
stantial budget outlays. In 2016, fertiliser subsidies alone amounted to over Rp 30 trillion
(about US$2.27 billion), a 25% increase over 2014 (FAO 2017).
In addition to subsidies, the Jokowi government has instituted new policies to allocate land
for agriculture. One of his government’s flagship rural sector policies is an ambitious land
reform programme which aims to redistribute nine million hectares of agricultural land to
people in rural communities (Sulistyo 2017).2 The land reform programme is part of a much
broader set of development policies pursued under Jokowi, but it is explicitly linked to the
food self-sufficiency agenda by both officials and court decisions connecting land reform with
the 2012 Food Law and a 2013 law on the protection and empowerment of farmers (Syahyuti
2015). Land reform policy is not radical – the land slated for distribution is state land or land
allocated to state-owned companies and, as such, much is already contested by prior
community claims under customary law.3 To an uncertain degree, the programme consists
of granting official title to people who already have claims to the land in question. In the case
of much “forest” land, some local groups are sceptical about the programme’s benefits, given
de facto occupation of lands being handed over (Jong 2017). The programme has also been
criticised by independent farmer groups. The Chairman of the Indonesian Peasants Union
was reported as saying that, “Agrarian reform is treated as if it only means certification of
land. That is not Agrarian Reform, but only providing a legal basis to land that has already
been controlled by the people” (La Via Campesina 2017). Nonetheless, the formalisation of
millions of titles – 2.9 million from 2015–August 2017, with another 7 million targeted for
2018 – marks a major acceleration of Indonesia’s long-running land titling programme
(Sulistyo 2017; Lucas and Warren 2013). The formalisation of titles is often critiqued as
offering no substitute for land reform and as reinforcing unjust land settlements; but it is also
 JOURNAL OF CONTEMPORARY ASIA 7

often supported by otherwise marginalised groups whose land claims are insecure in the
absence of formal titles (Hall, Hirsch, and Li 2011).
Rice remains the most politically symbolic focus of the food self-sufficiency drive, but
other crops and types of food are also targeted. Corn production, for example, has been
supported through a programme announced in 2015 to allocate one million hectares of
land to the crop and $250 million in direct support, with the aim of reducing corn
imports to zero in 2017 (Jakarta Post, September 19, 2016). Beef imports have also been
subject to on and off restrictions, sharply curtailing imports from some sources and
subsequently relaxing these controls at times when beef prices have spiralled (Permani
2013; Jakarta Post, February 12, 2018). Indonesia’s restrictions on beef imports led to
a World Trade Organisation (WTO) dispute procedure brought by New Zealand, which
had seen its beef exports to Indonesia drop by 80%. Although the WTO panel ruled
against Indonesia at the end of 2016, Indonesia appealed the decision (MFAT 2017).
Regulations and other interventions targeting a wide range of other products have
been introduced or increased under Jokowi. Milk has also been a focus of an official
drive to reduce dependency on imported milk, with the government aiming to increase
production to meet 41% of industrial demand by 2021, up from 23% in 2016 (Jakarta
Post, May 4, 2017). Garlic is also subject to both price controls and controlled imports,
in an attempt to reduce price spikes while simultaneously maintaining the long-term
aim of reducing imports (Jakarta Post, May 9 and 17, 2017).
Beyond price interventions, subsidies and import controls, the government’s self-
sufficiency programme has seen it develop an array of bureaucratic powers that insert civilian
and military officials into food production and distribution systems. In the case of the rice
self-sufficiency programme, the government in 2015 issued a statement via the Ministry of
Agriculture that it planned to revive the “civic mission” of the national armed forces (TNI –
Tentara Nasional Indonesia), under the TNI Manunggal Masuk Desa banner, using non-
commissioned officers as “quasi-agricultural extension officers” at the village level (Lassa and
Priamarizki 2015; see Laksmana 2019). The actual activities of such officers remain uncertain,
given that the programme has a mandate to support government policy more generally.
According to official military sources, there were programme personnel in 61 districts across
Indonesia as of April 2017 (HukumOnline, May 30, 2017).
In May 2017, during the approach to the fasting month of Ramadan, when prices for many
food stuffs often rise, the National Police established a “food stability task force” to co-
ordinate with the Home Ministry, Agriculture Ministry, Trade Ministry, Business
Competition Supervisory Commission and Bulog to monitor food prices and “prevent cartel
practices in the food industry” (Jakarta Post, May 3, 2017). The threat of reprisals against
sellers was implicit in the report that should prices of staple foods exceed limits set by the
government, the task force would “track the source of the supply” (Jakarta Post, May 22,
2017). A ministerial decree issued by the trade minister required distributors of staple foods
to register and report their stocks to the government each month. According to an official of
the Trade Ministry, “If they do not comply with these requirements, we will take actions that
perhaps initially begin with an administrative measure” (Jakarta Post, May 22, 2017).
Although introduced with the explicit aim of safeguarding price stability during Ramadan,
there has been no indication that the task force has been disbanded. Even if dormant, it
remains a potential lever of bureaucratic discretion that creates opportunities for leverage
against commercial actors.
8 N. HAMILTON-HART

Is the Self-sufficiency Programme Succeeding?

Trends in food production serve as a basic measure of the success of the self-sufficiency
programme. For the politically important rice crop, annual production remained at near
constant levels in the first years of reformasi, despite an increase in population. However,
particularly since 2006, national production of rice has increased (see Figure 1). At the end
of 2016, with much fanfare, the government announced that it had achieved its goal of rice
self-sufficiency and the agriculture minister was reported as saying that Indonesia is ready
to become a rice exporter (Indonesia Investments 2017a). In fact, Indonesia had been close
to self-sufficient in rice for most of the time since 2004, maintaining a “self-sufficiency rate
above 96%” since then as well as having achieved full self-sufficiency in 1984–1985 and
1992–1993 (Lassa and Priamarizki 2015).
“Self-sufficiency” in the sense that annual production is sufficient to cover national
consumption does not mean that Indonesia has stopped importing rice. As shown in
Figure 1, rice imports have continued and when international rice prices rise, the cost of
importing rice becomes significant. In order to meet its goals of maintaining price stability in
the domestic market, Bulog, which still has a monopoly on international trade in rice,
maintains a stockpile of between 1.5 and 2 million tonnes of rice, often requiring it to turn
to international suppliers (Indonesia Investments 2017b; Jakarta Post, February 15, 2018).
For other foods, self-sufficiency is an ever-receding target, as Indonesia’s population
grows and becomes more affluent, leading to increased consumption of wheat and
animal products. Thus, although a review by a Ministry of Agriculture official shows
that production of food crops has risen over the past 10 years, notably rice, corn and
sugar (Sudaryanto 2016), Figure 2 shows that this has done nothing to reverse the
overall level of food imports. These import figures show gross imports, not net imports.
In terms of the agricultural sector as a whole, Indonesia has long been a net exporter. Its
agricultural exports, however, are dominated by non-staple foods and other agricultural
products such as rubber, palm oil, copra, coffee and tea (OECD 2012; Warr 2011).
Import trends for the major imported food categories (wheat, rice, maize, sugar,
meat, dairy and animal feed) show more volatility, but there is no sign that Indonesia

90 $1,600
80 $1,400
70 $1,200
60 Rice Production
$1,000
50 (tonnes,
$800 millions)
40
$600 Rice Imports
30 (tonnes,
$400 millions)
20
Rice Imports
10 $200
($US million)
0 $0

Figure 1. Rice production and imports

Source: BPS (2016, 2017).
 JOURNAL OF CONTEMPORARY ASIA 9

30 $16,000

$14,000
25
$12,000
20
$10,000

15 $8,000 Tonnes, millions

$6,000 US$, millions
10
$4,000
5
$2,000

0 $0

Figure 2. Imports of food and live animals

Source: BPS (2016, 2017).

12

10

8 Wheat
Animal feed
6
Rice
4 Maize
Sugar
2

Figure 3. Imports of staple foods, millions of tonnes

Source: BPS (2016, 2017).

will significantly reduce food imports in the near term. Figure 3 shows import trends
over the past decade for four major staple foods and animal feed (which supports food
production). In contrast to the recent decline in maize imports – which may indicate
success in the Jokowi government’s intensive drive to increase corn production –
imports of wheat and sugar have continued to rise. Animal feed imports have also
steadily increased, in line with the government’s drive to increase domestic production
of milk, meat and other animal-derived food.
Shifts in consumption habits undermine the food self-sufficiency programme,
although some industries targeted for intensive support and import restrictions are
becoming less reliant on imports. Domestic beef production, for example, has increased
(with some fluctuations) from 339,941 tonnes in 2000 to 524,109 tonnes in 2016
(Bappenas n.d.). Import restrictions on beef and live cattle have been in place since
2011, and beef imports have declined somewhat since then, but annual figures fluctuate
10 N. HAMILTON-HART

0.45
0.4
0.35
0.3
0.25 Dairy & eggs
0.2 Meat
0.15 Live animals
0.1
0.05
0

Figure 4. Imports of animal products, millions of tonnes

Source: BPS (2016, 2017).

as the government has intermittently relaxed import controls to stem rising prices
(Sudaryanto 2016; Jakarta Post, February 12, 2018). Indonesia’s overall imports of meat
and meat products show a clear upward trend since 2012 and imports of dairy and eggs
have steadily increased (see Figure 4). The disruption from the earlier trend lines for
both meat and live animals (2006–2012), however, suggests that the efforts to restrict
imports have had some effect.
If Indonesia has realised a mixed set of outcomes from its food self-sufficiency drive
in terms of production volumes and import trends, the verdict on implications for
efficiency and welfare is that self-sufficiency efforts have been welfare-decreasing over-
all. Economic analyses argue that the array of subsidies, import restrictions and price
controls have reduced agricultural productivity (OECD 2012; Armas et al. 2012). There
is a fundamental contradiction between the goals of price stabilisation, increased
production and efficiency. As noted by an economist reviewing the self-sufficiency
programme nearly 25 years ago, “Most countries have found it impossible to increase
agricultural productivity very rapidly without price incentives for farmers that matched
(or often exceeded) those available in international markets” (Timmer 1993, 154).
Increases in production have been achieved at the cost of higher prices, with domestic
producers shielded by both at-the-border and behind-the-border protection. Although
production has increased, productivity – increased output per unit of input – has not.
As summed up by another economist, “Productivity growth requires expanded invest-
ment in domestic agricultural research and infrastructure development. But this is not
happening on a sufficient scale” (Warr 2011, 12). In this context, government inter-
ventions have veered between policies that have raised prices to encourage production
and stabilisation efforts to reduce prices to address popular discontent over price spikes.
Sudden spikes in the price of goods such as beef, garlic or chilli have prompted
official interventions either to increase supply through relaxing imports or more heavy-
handed attempts to control prices by decree. More persistently, food self-sufficiency
policies have reduced food security, in the sense of access to food at affordable prices.
A key indicator is the price of rice on Indonesia’s domestic market compared to the rice
 JOURNAL OF CONTEMPORARY ASIA 11

price in other rice-producing countries. According to World Bank figures that tracked
rice prices for 12 months in 2015–2016, Indonesia’s rice prices were 73% higher than in
Vietnam, a country with minimal producer subsidies (Hamilton-Hart and Schulze 2016,
279–281; see also OECD 2017a). In one analysis, high prices alongside the direct cost of
producer support mean that Indonesians were “taxed” the equivalent of US$98 billion
between 2013 and 2015 (McBeth 2016).
The food security implications of higher staple food costs are serious. As put in
a recent overview, “If policy makers close the domestic market and use high prices to
guarantee that rice farmers continue to grow specific commodities, they support
national food self-sufficiency. But, they create suffering among the poor. Higher prices
affect household food security” (McCarthy and Zen 2013). For approximately half of
the Indonesian population that either falls under or close to the $2 per day poverty line,
food accounts for between 67% and 72% of expenditure (McCulloch 2008, 49). At the
bottom of the income distribution, food can account for 80% of household expenditure
(Warr 2011, 3). High rice prices are partially offset by the government’s “Rice for the
Poor” programme and cash subsidies for the poorest households, but these programmes
have high overheads and (particularly with the switch to conditionality in the cash
transfer programme), do not outweigh the effects of high food prices (Widianto 2013;
FAO 2017). In addition to direct pricing effects, to the extent that food self-sufficiency
policies favour large-scale land acquisitions, self-sufficiency can be associated with food
insecurity for the most marginalised rural people (McCarthy and Obidzinski 2017; Ito,
Rachman, and Savitri 2014).
The adverse effects on food security generated by food self-sufficiency efforts are
denied by the government and supporters of self-sufficiency. President Widodo has
frequently claimed to be concerned about food security. In one speech, he claimed that,
“Food security is important as the rise and fall of a country is dependent on the
people’s ability to have access to an affordable supply of food. The state must take on
the role of ensuring that the people do not find difficulty accessing food.” He also
affirmed that the “welfare of farmers should also be given serious attention so that they
could become more productive” (The Jakarta Post, June 14, 2017). The government has
also argued that security of food supply cannot be guaranteed through reliance on
international markets, where food prices of staple commodities can rise and fall sharply.
As discussed below, this distrust of international markets is widely shared by
Indonesian farmer groups who have supported the food self-sufficiency drive, albeit
from different vantage points and with different ultimate goals.
The divergent views between critics of the food self-sufficiency programme and its
supporters reflect a fundamental conflict between producers and consumers of food.
Increasingly, Indonesia is a nation of net consumers (Neilson 2016). Already by the early
2000s, only 28% of rural households were “net producers who would benefit from higher
prices” while across the country, only 18.5% of all households were net producers
(McCulloch 2008, 53). As discussed in the next section, increasing inequality in access to
land within the rural sector means that even fewer households now are net food producers.
The food self-sufficiency programme has had an additional adverse effect on poverty
and food security because of the gap between farm-gate producer prices and retail
consumer prices. Paradoxically, “restrictions on rice, maize and sugar imports and
various non-tariff measures applied to animal and horticultural products may have
12 N. HAMILTON-HART

afforded some protection from competition, but that hasn’t translated into higher prices
for farmers” (McBeth 2016). Rural incomes have stagnated. According to the finance
minister, real incomes in 2017 were 3% lower in real terms then in 2014; and the
government price floor for rice has not kept up with production cost increases (Reuters,
December 1, 2017).4 The gap between prices realised by producers and those paid by
consumers is comparatively large for many products because of bottlenecks in distribu-
tion and lack of market access by the poorest of food-producing households. Small-
scale fishers, for example, are often tied into dependent relationships with intermedi-
aries who effectively monopolise the selling opportunities of the fishers who rely on
them for credit and access to markets (Adhuri et al. 2016). For rice, some of the primary
benefits of import restrictions are captured by importers and millers (Warr 2011).
A third factor that limits the effectiveness of producer subsidies in ameliorating poverty –
including food poverty – is the extent of misappropriation and misallocation of subsidised
products. Food self-sufficiency policies give officials expanded scope for discretion and
control over resources, in the distribution of producer support funds or price stabilisation
task forces. Corruption scandals dogged Bulog in the New Order (Timmer 1996, 46). In the
post-reformasi period officials involved in the distribution of subsidised products continue
to have access to resources that can be deployed for both political or financial purposes
(Neilson and Wright 2017, 139). In 2016, an ombudsman report found that subsidised
fertilisers were being sold to plantations, rather than small farmers, at 40% above the official
subsidised price – but still well below the market price for non-subsidised fertiliser. The
investigator reported a lack of oversight of the country’s 44,000 state-approved farm
retailers allowed this diversion (Reuters, December 1, 2016).
Overall, if food self-sufficiency is interpreted to mean the ending of major food imports,
this goal has not been achieved and is unlikely to be achieved in the medium term. However,
this does not mean that the self-sufficiency policy has not achieved a measure of success. The
trend lines for some categories of import have shown slower rates of increase and in some
cases a decrease as a result of the interventions to restrict imports and increase production.5
This achievement has come at the cost of higher prices as well as the budget and adminis-
trative costs of subsidies and interventions. The counterfactual – what would things be like
without the self-sufficiency interventions? – remains contested. Free markets do not necessa-
rily self-stabilise costlessly. Global food prices tend to be volatile (ADB 2012). However, there
is no evidence that import restrictions and increasingly interventionist attempts to administer
prices are effective in reducing volatility or improving access. To the extent that the govern-
ment has been able to contain price volatility, it has been as a result of both high investment
(principally the maintenance of large rice stockpiles) and a willingness to relax import
restrictions to ease price spikes. Bulog has through these means been moderately successful
in reducing price volatility in rice, albeit at the cost of administrative leakage and vulnerability
to corruption (Timmer 1996). Notably, Bulog’s price-smoothing functions do not owe
anything to the self-sufficiency agenda of reducing imports. Increasing production may
increase the supply of food at the national level, but this in itself does not address food
insecurity and malnutrition, which are ultimately problems of poverty, not national produc-
tion. Poverty and malnutrition are concentrated in rural areas, and on some indicators, such
as child stunting, have worsened in the period between 2007 and 2014 (SMERU 2015).
 JOURNAL OF CONTEMPORARY ASIA 13

Who Produces Food?

Is the agricultural sector changing in ways that would make it structurally conducive to
a more inclusive political economy or do changes unfolding alongside the food self-
sufficiency drive speak to a continuation of an essentially elite-driven political economy?
This section approaches this question through an analysis of who produces food and other
agricultural products, with a view to assessing the changing structure of agricultural land-
holdings. It suggests that the agricultural sector can be divided into three broad groups:
large-scale corporate agribusinesses, small smallholders and landless farm workers and
a consolidating group of middle farmers, who have been the major beneficiaries of recent
agricultural policies, including the self-sufficiency drive. Agriculture, of course, is a much
broader sector than food production. It includes non-food production and the production
of crops that have food uses (from palm oil to coffee) but are not classed as staple foods.
However, it is necessary to take trends in the non-staple food sector into account when
assessing the food self-sufficiency programme, because often the land use directly competes
with staple food crops and many producers switch their production between food and non-
food crops. Some large “food estate” projects include non-food investments.
The structure of the agricultural sector and the role it plays in the national economy have
changed since democratisation. At the start of the reformasi period, in 1999, nearly half of
the national workforce worked in agriculture, although this figure was inflated by the
exonomic crisis, as many Indonesians returned temporarily to agricultural work (Hadi,
Madjid, and Indra 2003, 76). Since 2003, the share of agriculture in total employment has
trended steadily downwards, from 46% to 33% in 2015 (World Bank 2017b). Agricultural
value added as a share of GDP has been stable at around 13% between 2010 and 2017, down
from 23% in 1985 and 17% in 2000 (FAO 2003; World Bank 2017a). Total land area classed
as agricultural land remained static between 1961 and 1985, at around 38 million hectares.
Thereafter it began to increase, to 45 million in 1990, 47 million in 2000, 55 million in 2010
and 57 million hectares in 2015 (FAO 2018).
Large agribusiness ventures are not new to Indonesia, with many firms and planta-
tions established in the colonial era and subsequently taken into state ownership early
in the post-independence period. Private investors in plantation agriculture became
commercially significant during the Suharto period, often as part of multi-sectoral
conglomerate groups (Warburton 2018, 118–122). Despite a retreat for many of these
groups as a result of the 1997–1998 crisis, many have since re-established themselves in
a more plural corporate and political landscape (Robison and Hadiz 2004). By 2017,
there were 21 listed agricultural companies on the stock exchange, many of them still
linked to well-established conglomerate groups as well as new groups that have risen
during the reformasi period (Carney and Hamilton-Hart 2015). State-owned enterprises
also continue to be active in the agricultural sector. Corporate agriculture has thus
continued to have a close association with state elites.
Corporate agribusiness tends to be concentrated in export-oriented cash crops that can be
grown efficiently on large estates. They account for a significant share of oil palm planting,
which is the dominant crop behind the expansion in agricultural land in the last decade. The
average size of a private estate in palm oil was 4,000 hectares in 2013 (Daemeter 2015). The
private estate share of all land under oil palm dropped from 58% in 2000 to 50% in 2012 and
then rose to 55% in 2016 (DJP 2017). Over the long-term, these producers have enjoyed
14 N. HAMILTON-HART

substantial rents from state policies and actions, from collusion in the coercive acquisition of
land, access to finance and ability to flout environmental regulations (McCarthy, Gillespie,
and Zen 2012). More recently, in terms of net financial transfers, the palm oil sector is no
longer a net beneficiary of state interventions, with the introduction of a variable-rated export
tax on palm oil (rated at zero since 2014 due to relatively low prices) and an industry levy
imposed in 2015 to subsidise biodiesel, which was expected to raise approximately
US$700 million in 2016 (FAO 2017).
When it comes to food production supported under the self-sufficiency drive, the corpo-
rate sector has made some gains. Massive “food estate” schemes have attracted agribusiness
investors, with the most prominent being the 1.2 million hectare Merauke Integrated Food
and Energy Estate in West Papua (MIFEE), announced in 2010. MIFEE has created oppor-
tunities for agribusiness interests close to political actors (Ito, Rachman, and Savitri 2014;
Aspinall 2015; Neilson and Wright 2017). When launched, the crops targeted included rice,
palm oil, corn, soybeans and sugar. Although MIFEE is the flagship estate under the food self-
sufficiency policy, food crops (not including sugar and oil palm) make up only a small share
of the land allocation (Ginting and Pye 2011). An analysis of allocated development permits
and plans by named companies as of August 2013 showed that nearly all the area would be
taken up with sugar cane, oil palm and industrial forestry plantations (AwasMIFEE 2013).
Due in part to active resistance from local groups and problems securing land and environ-
mental impact assessments, the project had a slow start, with an official describing it as
“stalled” in 2015, when it was targeted for revitalisation led by state-owned enterprises
(Jakarta Post, April 16, 2015). Jokowi’s public statements on the project continue to headline
rice production for MIFEE, but the feasibility remains in doubt and the level of interest in
carrying out rice agriculture in the estate is low (AwasMIFEE 2015). Similarly, corporate
investors have been slow to plant rice in food estates in Kalimantan, where they have failed to
meet productivity targets and SOEs were again called on to invest, although also with limited
success (McCarthy and Obidzinski 2017, 348–349).
Smallholders appear to be capturing a significant share of the gains from agricultural
development and food self-sufficiency targets. Smallholders account for a rising share of
many traditional plantation crops. In oil palm, the smallholder share often increased in the
wake of expansion by large plantation companies (Budidarsono, Susanti, and Zoomers.
2013). In Riau Province, a major oil palm area, smallholders have displaced the historically
dominant large plantations to become the single largest category of oil palm producers
(Potter 2016, 320). In other areas, such as West Kalimantan, the plantation sector continues
to expand faster than the smallholder sector (Hamilton-Hart and Palmer 2017). The
smallholder share of the oil palm planted area stood at 28% in 2000 and rose to a peak of
43% in 2012, although it has declined since then to 39% in 2016 (DPJ 2017). The picture is
therefore mixed, but in some accounts the dominant trend across Southeast Asia is
a transition from large estates to smallholders for several cash crops, as smallholders are
often more cost-effective (Bissonnette and De Koninck 2017).
The relative advantage of the smallholder sector is subject to an important qualifica-
tion. The smallholder sector is becoming increasingly differentiated, to the extent that it
makes little sense to see it as representing a single set of interests (Jelsma et al. 2017,
287–289). Some smallholders have been dispossessed to make way for large estates in
crops such as oil palm, and many official “partnership” schemes linking smallholders
with estates have operated to the disadvantage of those who had previously farmed in
 JOURNAL OF CONTEMPORARY ASIA 15

the area (Cramb and McCarthy 2016). As noted above, food estates in Papua and
Kalimantan have made things worse for some small-scale farmers. More generally, the
commercialisation of agriculture has made life more precarious for those who are
unable to retain access to viable landholdings (Li 2009a, 2009b). But those smallholders
who have retained their land, often as a result of having secured better access to
subsidised inputs or credit, appear to be increasing their holdings and realising higher
incomes. A survey of Sumatran palm oil smallholders showed that a large part of
smallholder income gains, which were positive at all landholding sizes, were attributable
to increasing plot area (Euler et al. 2017). “Smallholders” may in fact not be that small.
Informal landholdings and the accumulation of multiple plots by the same owner
means that many oil palm smallholders are in fact moderately large-scale operators,
often local bureaucratic, political and commercial elites.6 While these actors include
local elites, they do not operate at the scale of large plantation estates.
Most staple food production continues to be in what is officially designated as the
smallholder sector (Hadi, Madjid, and Indra 2003). Although food crop landholdings are
smaller on average than those for tree crops, there is a similar trend towards consolidation.
One account of agricultural landholdings over the long term has shown that the share of
landless agricultural households has increased since 1983, and of farmers who did hold
land, the proportion with one hectare or more rose from 24% in 1963 to 29% in 2003. At the
same time, the proportion of “small peasants” with less than 0.5 hectares rose steadily from
44% in 1963 to 51% in 2003 (Bachriadi and Wiradi 2013, 43, 64, 54).
Recent changes are obscured by a shift in the basis for inclusion in the agricultural
census between 2003 and 2013. The figures reported for the 2013 census show a large
decline in the number of agricultural households, from 31.2 million to 26.1 million,
concentrated in households with very small plots of less than 0.1 hectare, as shown in
Figure 5. This decline is partly driven by a change in the definition of an agricultural
household, with the 2013 census excluding most non-rice growing farmers unless they
produced for non-household use (Neilson 2016, 256). This is consistent with the
otherwise unrealistically dramatic drop of millions of farms producing things like
chicken and bananas, which are commonly grown in small gardens for household

10
9
Number of households

8
7
6
5
4
2003 Millions
3
2 2013 Millions
1
0

Land per household, hectares

Figure 5. Agricultural households by size of land holding

Source: BPS (2013c)
16 N. HAMILTON-HART

consumption and reduced by nearly eight million and 10 million respectively

(Sudaryanto 2015). Nonetheless, there is clearly some real exit from agriculture of the
smallest farmers and increase in landholdings for those that remain. Population growth
and land fragmentation, compounded by increases in large-scale land conversions for
estate cropping, have led to greater inequality in land distribution over time (McCarthy
and Robinson 2016, 5–6; Neilson 2016). The average plot of rice-producing land
(sawah) increased from around 0.1 hectares in 2003 to 0.2 in 2013, for all agricultural
households, while for land-owning households, average rice plot size was 0.4 hectares in
2013 (BPS 2013a; 2013b). At the same time, the numbers reported as engaged in rice
farming increased by 3% (Sudaryanto 2015). These averages conceal wide variation in
landholdings – in the 2013 census, “small farmers” (petani gurem, those with less than
0.5 hectares) accounted for 55% of all agricultural households (BPS 2013c). In the same
census, households with plots above one hectare increased from 21% of all agricultural
households in 2003 to nearly 27% in 2013. Legally-formed farm businesses (those with
a formal company structure, thus representing larger operations) increased from
4 million to 5.5 million. Over the same period, total agricultural land increased from
51 million hectares in 2003 to 57 million in 2013 (FAO 2018).
Although landholding sizes vary dramatically by province, and are much larger for
tree crops than rice and other horticultural crops, the overall trend towards greater
differentiation holds across a variety of accounts. A study of Javanese villages in 2013
and 2015 shows the share of landless and marginal farmers with very small plots rising,
while local large farmers or absentee landlords became increasingly important – albeit
in most cases with holdings of less than five hectares (Ambarwati et al. 2016, 269–274).
This study is in line with the finding that some of the largest beneficiaries of import
controls that raise rice prices are landowners: “Agricultural product prices become
capitalized into the price of agricultural land. Raising the price of rice benefits anyone
who owns rice land, or land that could be used to produce rice” (Warr 2011, 10). Local
surveys and field studies have found active markets for land in many parts of Indonesia,
often with speculative characteristics (McCarthy 2010; Fortin 2011).
Animal production, which enjoys high (if intermittent) levels of protection under the
food self-sufficiency programme, remains largely in smallholder hands. Most dairy
farmers own only two to three cows. For beef and other livestock, most household
producers are small-scale – official figures estimate around 66% of households involved
in livestock production are smallholders with one or two animals; other estimates
suggest most beef producers have two to three animals (Sudaryanto 2016; Permani
2013). However, this means a third of producers are larger operations. For dairy and
beef, expansion subsidised through the self-sufficiency drive is likely to involve elite
interests. In the case of beef, influential beef cattle producers are widely presumed to
have been the “mafia” behind beef import restrictions in recent years (Informal com-
munication, meat industry investor, November 2016). Elite association with beef pro-
duction has a history in Indonesia, dating back at least to a World Bank (1972) proposal
to fund two 12,000-hectare cattle ranches for beef production. President Suharto was
known to have personal cattle ranch interests and Prabowo Subianto, the former
general turned post-reformasi era politician who vied with Jokowi for the presidency
in 2014, is reported to own a herd of 75 cattle (South China Morning Post, March 24,
2014). He has also served as chairman of the Indonesian Farmers’ Association, HKTI,
 JOURNAL OF CONTEMPORARY ASIA 17

a role that other members of Indonesia’s political elite have vied for (Kompas, April 10,
2017). In dairy, recent plans envisage expansion that will not be able to be met through
traditional smallholder farms with two or three cows each. Several new commercial
dairy-processing plants have been established, with the intention of increasing domestic
supply on a large scale (Jakarta Post, May 4, 2017; Tempo, May 4, 2017). It is difficult to
see how domestic supply will increase without a shift towards larger-scale production.
For now, however, the elite interests involved in animal production are mostly not
operating very large estates such as those owned by corporate agribusinesses in cash
crop production.

Prospects for Agriculture-Led Inclusion

The idea that economically illiberal food self-sufficiency policies associated with widen-
ing rural inequalities could in any way support more inclusive agriculture-led develop-
ment appears counter-intuitive. It is important to acknowledge the real political and
structural obstacles that hinder prospects for a more inclusive political economy.
However, the consolidation of a middle farmer group in Indonesia’s current political
and policy context could have inclusive potential. Rents generated by illiberal state
interventions have, by definition, static welfare-decreasing effects, but can lay the
ground for more dynamic transformations (Khan and Jomo 2000). The preceding
section argued that these rents have, to an extent, been captured by a broad “middle
farmer” category of landowners whose operations remain far smaller than the thousand
hectare estates of the agribusiness sector, but are significantly larger than those of
marginal farmer-smallholders and tenant farmers and, of course, landless farm workers.
The idea that the consolidation of a middle farmer sector could have inclusive
implications derives from the experience of economies where such agricultural interests
have played progressive roles in generating broad-based economic growth and demo-
cratisation, as described earlier in this article. Agriculture will not solve problems of
poverty but “has proven uniquely powerful for that task” (World Bank 2007, 1). This
claim is contested by accounts that emphasise the long-term diversification of rural
incomes away from agriculture and argue that the Indonesian land reform agenda
places too much emphasis on agricultural land as a productive resource (Neilson 2016,
261). Income diversification is indeed important, but rising incomes for larger small-
holders suggests that access to productive land is also important, as are spillovers from
farming in the wider rural economy (for example, Euler et al. 2017; Wardhana, Ihle,
and Heijman 2017). Land reform and the break-up of labour-repressive estates is
essential to any such progression. But it also requires landholdings large enough to
generate incomes that yield more than a marginal existence. Competition for land
under the dynamics of increasing agricultural commercialisation has created land
scarcity, which means that many of the smallest farm plots are not viable. The declining
share of agriculture in total employment over the post-reformasi period combined with
increasing average smallholder plots shows that many of the smallest farmers have
made an “exit from agriculture,” affirming the importance of non-farm incomes. Others
remain as landless farm labourers. Li’s (2009a) trenchant critique of the World Bank’s
new focus on agriculture makes the point that for those who have limited options to
“exit,” this process is one of dispossession and increasingly precarious existence,
18 N. HAMILTON-HART

vulnerable to exploitative labour practices by large agricultural estates. However, small-

holders who remain in agriculture and consolidate their holdings are in a much
stronger position.
Where this leads depends on the broader political dynamics that shape rural sector
policy and economic policy. The dynamics of consolidation in a commercial agricul-
tural context could mean further exploitation of landless and marginal farmers, albeit
practiced by “large smallholders” rather than estates.7 The legacy of repression of mass
movements and labour unions under the New Order means that, as put by Li (2009b,
85) in another analysis, the social forces behind a progressive rural political economy
that includes a “right to food” are “strikingly absent in Indonesia.” Although these
legacies have not disappeared, the social forces driving the broad thrust of rural sector
policy have shifted in Indonesia. Despite their limitations and perversities, the agricul-
tural politics associated with food self-sufficiency are much more broad-based than
suggested by a narrow focus on elite capture and corporate land-grabbing.
The first point of note is that the political weight of smallholder farmers is significant.
The New Order-era farmers’ association, HKTI, remains a target for members of the
current political elite, suggesting that the rural sector is important. Retired general and
HKTI president Moeldoko claims the organisation represents 60 million farmers (HKTI
Online, January 15, 2018). HKTI now exists alongside a large number of independent
farmer organisations, which multiplied and became increasingly assertive early in the
reformasi era (Lucas and Warren 2003; Peluso, Afiff, and Rachman 2008). These farmer
organisations, despite their very different political lineages, have all advocated for at least
portions of the food self-sufficiency programme, which is often attributed to the pressure
exerted by the country’s “mostly small-scale farmers” (Financial Times, October 11, 2012).
Thus “pro-farmer political groups” are described as lobbying successfully for the reintro-
duction of rice import controls under President Yudhoyono (Warr 2011). HKTI has long
argued that higher rice prices ensure food security by encouraging production, as well as
raising incomes in rural areas (McCulloch 2008, 46–47). The umbrella grouping of many
independent organisations, the FSPI or Federation of Independent Peasant Unions
(Federasi Serikat Petani Indonesia), has been supportive of the notion of food sovereignty
as promoted globally by the Via Campesina farmers’ advocacy group, which moved its
secretariat from Honduras to Indonesia in 2004 (Peluso Afiff, and Rachman 2008). To the
extent that “food sovereignty” resonates with Indonesian discourses of food self-sufficiency
in Indonesia, the food self-sufficiency programme is associated with a broader set of
interventions (Neilson and Wright 2017).
Second, self-sufficiency policies themselves can strengthen the political weight of farmers
whose economic interests they support. Policies aimed at national self-sufficiency in food
do not in themselves support the food sovereignty movement’s championing of land rights
and autonomy for farming communities, but the two movements share a common distrust
of international food markets and, to this extent, a common economic illiberalism. In
political terms, the food self-sufficiency programme represents an acknowledgement of the
status of farmers that has bolstered demands for land reform. In the first years after the fall
of Suharto, hundreds of land seizures by peasants occurred amidst a wave of peasant
organising and political action (Lucas and Warren 2003, 88–91). As independent farmer
groups developed their political organisation and focus, they emphasised calls for land
reform and “farmer rights,” in line with the food sovereignty movement (Peluso, Afiff, and
 JOURNAL OF CONTEMPORARY ASIA 19

Rachman 2008). Although land reform policy remains cautious, as described above, land
seizures by peasants have continued. One organisation reported 200,000 hectares of land
seized by small farmers and landless peasants between 2007 and 2014 (Neilson 2016, 251).
To the extent that official and unofficial action does succeed in enabling greater land access
by previously marginal farmers, it lays the ground for less precarious rural livelihoods and
enhanced opportunities for political participation.
Third, while the rural sector is extolled in the populist rhetoric of most political
contenders, it has gained particular political valorisation under the presidency of Joko
Widodo. Farmer organisations strongly backed his 2014 presidential campaign (Neilson
2016, 253) and, despite his rival Prabowo’s explicit invocation of rural themes in his
campaign, exit-polls suggested that Prabowo appealed most to wealthier, urban and
more educated voters, while Jokowi succeeded in winning the rural vote (Aspinall 2015,
25). While Jokowi was initially in a very weak political position, commanding only
a small minority of legislative seats, he greatly increased his political strength (Mietzner
2017). With a consolidated political position, he has expanded the land reform agenda
and elevated his commitment to rural support. In addition to increasing agricultural
subsidies described here, he has also expanded the “people’s credit” subsidised finance
scheme, village funds and rural infrastructure development (Hamilton-Hart and
Schulze 2016; Yusuf and Sumner 2015). There has been no revolution, but the political
status of farmers as agricultural producers has increased. Perhaps symbolic of this shift,
a former land rights and self-described “scholar activist,” Noer Fauzi Rachman, was
appointed by Jokowi to a presidential staff position with responsibility for monitoring
land reform (Kantor Staf Presiden n.d.).

Conclusions
One broad conclusion to be drawn from this analysis of food self-sufficiency in the
context of developments in Indonesia’s agricultural sector is that the intensification of
the food self-sufficiency drive forms part of a more general illiberal turn in Indonesian
policymaking, dating from around 2009 and intensified through increasing interven-
tions from 2014 (Diprose, McRae, and Hadiz 2019). Illiberalism in this context means
a distrust of markets, particularly international markets, and a willingness to adopt
intrusive state interventions to direct economic activity. It is consistent with the rise of
a populist, nationalist current in Indonesian politics (Aspinall 2015; Mietzner 2015).
A second conclusion, however, is that although Indonesian populism has in many ways
been used instrumentally in the service of an oligarchic elite (Hadiz 2017), the agri-
cultural sector populism represented by Indonesia’s illiberal food self-sufficiency poli-
cies is not entirely captured by elite interests. There are elite beneficiaries of the food
self-sufficiency programme, but the political foundations of Indonesia’s economic
illiberalism in the food sector are much broader-based than suggested by a narrow
focus on elite-owned agribusiness ventures. Farmers are much more politically mobi-
lised and influential than under the New Order and, even though many aspirations
unleashed at the start of the reformasi period are far from being met, farmers are
a politically significant constituency.
A third conclusion, albeit a speculative one, is that the consequences of illiberal
economic interventions can be more dynamic than a focus on static inefficiencies suggests.
20 N. HAMILTON-HART

Economic transformations can give rise to new socio-economic interests that may con-
solidate their position. Illiberal food self-sufficiency policies have encouraged processes of
differentiation in the smallholder sector and consolidated a “middle farmer” grouping.
This process was already underway as a result of agricultural commercialisation and the
development of rural land markets, but it has been further supported by policies associated
with the food self-sufficiency drive. Although this has worsened outcomes for the most
marginalised, it may simultaneously be laying the foundations of a more inclusive
agricultural political economy, through the strengthened economic position of middle
farmers. Such a pathway to inclusive growth faces countervailing forces; and the political
mobilisation of social forces in support of family-owned farms and rural workers remains
limited. Nonetheless, there are opportunities in the current political and economic land-
scape that favour a non-elite form of agricultural capitalism.

Notes
1. There is an extensive literature on the massacres and their long-term reverberations.
Notable works include Cribb (1990), Farid (2005), and Kammen and McGregor (2012).
On the ongoing impacts, see McGregor and Setiawan (2019).
2. In addition, 12.7 million hectares of designated as forest land are supposed to be redis-
tributed to rural communities under social forestry schemes.
3. Many works have documented the dispossession entailed by the imposition of state
ownership in different parts of Indonesia, along with ongoing contestation by local
communities (see, for example, Peluso 1992; McCarthy 2006; Kelly and Peluso 2015).
4. The decline in rural incomes (as opposed to agricultural wages) is consistent with the end
of the commodity price boom by around 2011–2012 (Yusuf and Sumner 2015).
5. These figures do not of course take into account unrecorded imports sold on the black
market.
6. Although Indonesia’s 2013 agricultural law prescribes a maximum landholding of two
hectares for smallholder status, for plantation crops the cut-off is 25 hectares (Jelsma et al.
2017). In some provinces, much land is planted without a permit (see Potter 2016;
Hamilton-Hart and Palmer 2017). A detailed land survey in a district in Riau found
seven different categories of “smallholder” ranging from very small holdings of around
one hectare to large “investor” plots of 10 hectares, with many investors holding multiple
plots (see Jelsma et al. 2017).
7. Such an outcome may be somewhat self-limiting, given that the smallholder advantage vis-
à-vis large estates lies in their more efficient use of labour (Bissonnette and De Koninck
2017). It is likely that family farms avoid the labour monitoring and evasion costs
associated with coercive labour practices.

Disclosure Statement
No potential conflict of interest was reported by the author.

References
ADB. 2012. Food Security in Asia: The 2007–2008 Food Price Crisis. Manila: Asian Development
Bank, August 30. Accessed August 23, 2017. https://www.adb.org/features/has-world-learned
-2007-2008-food-price-crisis.
 JOURNAL OF CONTEMPORARY ASIA 21

Adhuri, D., L. Rachmawati, H. Sofyanto, and N. Hamilton-Hart. 2016. “Green Market for Small
People: Markets and Opportunities for Upgrading in Small-scale Fisheries in Indonesia.”
Marine Policy 63: 198–205.
Agarwal, D. 2014. “Food Sovereignty, Food Security and Democratic Choice: Critical
Contradictions, Difficult Conciliations.” Journal of Peasant Studies 41 (6): 1247–1268.
Ambarwati, A., R. Harahap, I. Sadoko, and B. White. 2016. “Land Tenure and Agrarian Structure
in Regions of Small-Scale Food Production.” In Land and Development in Indonesia: Searching
for the People’s Sovereignty, edited by J. McCarthy and K. Robinson, 265–294. Singapore:
ISEAS.
Armas, E., C. Osorio, B. Moreno-Dodson, and D. Abriningrum. 2012. “Agriculture Public
Spending and Growth in Indonesia.” Washington, DC: World Bank Policy Research
Working Paper WPS 5977.
Aspinall, E. 2015. “Oligarchic Populism: Prabowo Subianto’s Challenge to Indonesian
Democracy.” Indonesia 99: 1–28.
AwasMIFEE. 2013. “Three Years of MIFEE (Part 3): As the Forest is Felled Where’s the Rice?”
October 23. Accessed May 17, 2018. https://awasmifee.potager.org/?p=584
AwasMIFEE. 2015. “Jokowi Relaunches MIFEE, Wants 1.2 Million Hectares of New
Ricefields Within 3 Years!” May 15. Accessed May 17, 2018. https://awasmifee.potager.
org/?p=1210.
Bachriadi, D., and G. Wiradi. 2013. “Land Concentration and Land Reform in Indonesia.” In
Land for the People: The State and Agrarian Conflict in Indonesia, edited by A. Lucas and
C. Warren, 40–92. Athens: Ohio University Press.
Bappenas. n.d. “Produksi Daging Sapi menurut Provinsi, 2000–2016” [Beef Production by
Province]. Bappenas Website. Accessed March 19, 2018. https://www.bappenas.go.id/down
load.php?id=14962?id=14962.
Bissonnette, J.-F., and R. De Koninck. 2017. “The Return of the Plantation? Historical and
Contemporary Trends in the Relation between Plantations and Smallholdings in Southeast
Asia.” Journal of Peasant Studies 44 (4): 918–938.
BPS. 2013a. “Rata-rata Luas Lahan yang Dikuasai per Rumah Tangga Usaha Pertanian menurut
Wilayah dan Jenis Lahan Tahun 2003 dan 2013” [Average Area of Land Controlled by
Agricultural Business Household by Province and Type, 2003 and 2013]. Agricultural census,
Badan Pusat Statistik/Statistics Indonesia. Accessed March 12, 2018. https://st2013.bps.go.id/
dev2/index.php/site/tabel?tid=27&wid=0000000000.
BPS. 2013b. “Rata-rata Luas Lahan yang Dikuasai per Rumah Tangga Usaha Pertanian menurut
Wilayah dan Jenis Lahan Tahun 2013” [Average Area of Land Controlled by Agricultural
Business Household by Province and Type, 2013]. Agricultural census, Badan Pusat Statistik/
Statistics Indonesia. Accessed March 12, 2018. at https://st2013.bps.go.id/dev2/index.php/site/
tabel?tid=28&wid=0000000000.
BPS. 2013c. “Jumlah Rumah Tangga Usaha Pertanian Pengguna Lahan dan Rumah Tangga
Petani Gurem menurut Wilayah Tahun 2003 dan 2013” [Total Agricultural Business
Households Owning Land and Small Farmers by Province, 2003 and 2013]. Agricultural
Census, Badan Pusat Statistik/Statistics Indonesia. Accessed March 12, 2018. https://st2013.
bps.go.id/dev2/index.php/site/tabel?tid=22&wid=0000000000.
BPS. 2016. “Import by SITC Commodity Groups.” Foreign Trade Statistics: Imports, Volume 3.
Jakarta: Badan Pusat Statistik/Statistics Indonesia.
BPS. 2017. “Import by SITC Commodity Groups.” Foreign Trade Statistical Bulletin: Imports,
May 2017. Jakarta: Badan Pusat Statistik/Statistics Indonesia.
Budidarsono, S., A. Susanti, and A. Zoomers. 2013. “Oil Palm Plantations in Indonesia: The
Implications for Migration, Settlement/Resettlement and Local Economic Development.” In
Biofuels – Economy, Environment and Sustainability, edited by F. Zhen, 173–193. New York:
INTECH.
Campos, J., and H. Root. 2001. The Key to the Asian Miracle: Making Shared Growth Credible.
Washington, DC: Brookings Institution.
22 N. HAMILTON-HART

Carney, R., and N. Hamilton-Hart. 2015. “What do Changes in Corporate Ownership in

Indonesia Tell Us?” Bulletin of Indonesian Economic Studies 51 (1): 123–145.
Cramb, R., and J. McCarthy, eds. 2016. The Oil Palm Complex: Smallholders, Agribusiness and the
State in Indonesia and Malaysia. Singapore: NUS Press.
Cribb, R., ed. 1990. The Indonesian Killings of 1965–1966: Studies from Java and Bali. Clayton,
Victoria: Monash University Centre of Southeast Asian Studies.
Daemeter. 2015. Overview of Indonesian Oil Palm Smallholder Farmers. Bogor: Daemeter.
Diprose, R., D. McRae, and V. Hadiz. 2019. “Two Decades of Reformasi: Indonesia and its
Illiberal Turn.” Journal of Contemporary Asia 49 (5).
DJP. 2017. “Tabel 2. Luas Areal dan Produksi Kelapa Sawit (Inti Sawit) Menurut Status
Pengusahaan Tahun 1970–2017 [Table 2. Area and Production of Palm Oil (Palm Kernel)
by Type of Business, 1970–2017]. In Statistik Perkebunan Indonesia 2015–2017, Kelapa Sawit
[Tree Crop Statistics of Indonesia 2015–2017, Palm Oil]. Direktorat Jenderal Perkebunan
Website. Accessed March 12, 2018. http://ditjenbun.pertanian.go.id/tinymcpuk/gambar/file/
statistik/2017/Kelapa-Sawit-2015-2017.pdf.
Dove, M., and D. Kammen. 2001. “Vernacular Models of Development: An Analysis of Indonesia
Under the ‘New Order’.” World Development 29 (4): 619–639.
Euler, M., V. Krishna, S. Schwarze, H. Siregar, and M. Qaim. 2017. “Oil Palm Adoption,
Household Welfare, and Nutrition among Smallholder Farmers in Indonesia.” World
Development 93: 219–235.
Fane, G., and P. Warr. 2008. “Agricultural Protection in Indonesia.” Bulletin of Indonesian
Economic Studies 44 (1): 133–150.
FAO. 2003. WTO Agreement on Agriculture: The Implementation Experience – Developing
Country Case Studies. Food and Agriculture Organization Website. Accessed May 25, 2018.
http://www.fao.org/docrep/005/y4632e/y4632e0l.htm.
FAO. 2017. Indonesia: Country Fact Sheet on Food and Agriculture Policy Trends. Food and
Agriculture Organization Website. Accessed February 26, 2018. http://www.fao.org/3/
a-i7696e.pdf.
FAO. 2018. Indonesia – Land. FAOSTAT. Food and Agriculture Organization Website. Accessed
March 8, 2018. http://www.fao.org/faostat/en/#country/101.
Farid, H. 2005. “Indonesia’s Original Sin: Mass Killings and Capitalist Expansion, 1965–66.”
Inter-Asia Cultural Studies 6 (1): 3–16.
Fortin, C. 2011. “The Biofuel Boom and Indonesia’s Oil Palm Industry: The Twin Processes of
Peasant Dispossession and Adverse Incorporation in West Kalimantan.” Paper, Conference on
Global Land Grabbing, University of Sussex, April 6–8.
Ginting, L., and O. Pye. 2011. “Resisting Agribusiness Development: The Merauke Integrated
Food and Energy Estate in West Papua, Indonesia.” Paper, Conference on Global Land
Grabbing, University of Sussex, April 6–8.
Hadi, P., M. Madjid, and M. Indra. 2003. “Indonesia.” In Development of Agribusiness
Enterprises, edited by R. Oliver, 76–86. Tokyo: Asian Productivity Organisation.
Hadiz, V. 2017. “Indonesia’s Year of Democratic Setbacks: Towards a New Phase of Deepening
Illiberalism?” Bulletin of Indonesian Economic Studies 53 (3): 261–278.
Hall, D., P. Hirsch, and T. M. Li. 2011. Powers of Exclusion: Land Dilemmas in Southeast Asia.
Singapore: NUS Press.
Hamilton-Hart, N. 2012. Hard Interests, Soft Illusions: Southeast Asia and American Power.
Ithaca: Cornell University Press.
Hamilton-Hart, N., and B. Palmer. 2017. “Co-investment and Clientelism as Informal
Institutions: Beyond ‘Good Enough’ Property Rights.” Studies in Comparative International
Development 54 (4): 416–435.
Hamilton-Hart, N., and G. Schulze. 2016. “Taxing Times in Indonesia: The Challenge of
Restoring Competitiveness and the Search for Fiscal Space.” Bulletin of Indonesian Economic
Studies 56 (3): 265–295.
Huber, E., D. Rueschemeyer, and J. Stephens. 1993. “The Impact of Economic Development on
Democracy.” Journal of Economic Perspectives 7 (3): 71–86.
 JOURNAL OF CONTEMPORARY ASIA 23

Indonesia Investments. 2017a. “Self-Sufficiency in Rice Achieved, Indonesia to Become Rice

Exporter?” Indonesia Investments, February 10. Accessed July 30, 2017. https://www.indone
sia-investments.com/news/todays-headlines/self-sufficiency-in-rice-achieved-indonesia-to-
become-rice-exporter/item7603.
Indonesia Investments. 2017b. “Rice.” Indonesia Investments, June 28. https://www.indonesia-
investments.com/business/commodities/rice/item183.
Ito, T., N. Rachman and L. Savitri. 2014. “Power to Make Land Dispossession Acceptable:
A Policy Discourse Analysis of the Merauke Integrated Food and Energy Estate (MIFEE),
Papua, Indonesia.” Journal of Peasant Studies 41 (1): 29–50.
Jelsma, I., G. Schoneveld, A. Zoomers, and A. van Westen. 2017. “Unpacking Indonesia’s
Independent Oil Palm Smallholders: An Actor Disaggregated Approach to Identifying
Environmental and Social Performance Challenges.” Land Use Policy 69: 281–297.
Jong, H. 2017. “As Indonesia Pushes Flagship Land Reform Program, Farmers Remain Wary.”
Mongabay, November 21. Accessed March 14, 2018. https://news.mongabay.com/2017/11/
indonesia-pushes-flagship-land-reform-program-farmers-remain-wary/.
Kammen, D., and K. McGregor, eds. 2012. The Contours of Mass Violence in Indonesia, 1965–
1968. Singapore: Asian Studies Association of Australia and NUS Press.
Kantor Staf Presiden.n.d. “Noer Fauzi Rachman.” Accessed March 14, 2018. http://ksp.go.id/
tentang-kantor-staf-presiden/struktur-organisasi/noer-fauzi-rachman/.
Kelly, A., and N. Peluso. 2015. “Frontiers of Commodification: State Lands and Their
Formalization.” Society and Natural Resources 28 (5): 473–495.
Khan, M., and K. S. Jomo, eds. 2000. Rents, Rent-Seeking and Economic Development. Cambridge:
Cambridge University Press.
Laksmana, E. 2019. “Reshuffling the Deck? Military Corporatism, Promotional Logjams, and
Post-Authoritarian Military Reform in Indonesia.” Journal of Contemporary Asia 49 (5). DOI:
10.1080/00472336.2019.1613556.
La Via Campesina. 2017. “Bring Agrarian Reform Program in Indonesia Directly under the
President’s Command: SPI.” La Via Campesina Website, October 6. Accessed March 14, 2018.
https://viacampesina.org/en/bring-agrarian-reform-program-under-the-presidents-command-
spi/.
Lassa, J., and A. Priamarizki. 2015. “Jokowi’s Food Sovereignty Narrative: Military in the Rice
Land?” RSIS Commentary 40, February 27. Accessed August 26, 2017. https://www.rsis.edu.sg/
wp-content/uploads/2015/02/CO15040.pdf.
Li, T. 2009a. “Exit From Agriculture: A Step Forward or a Step Backward for the Rural Poor?”
Journal of Peasant Studies 36 (3): 629–636.
Li, T. 2009b. “To Make Live or Let Die? Rural Dispossession and the Protection of Surplus
Populations.” Antipode 41 (S1): 66–93.
Li, T. 2016. “Governing Rural Indonesia: Convergence on the Project System.” Critical Policy
Studies 10 (1): 79–94.
Lucas, A., and C. Warren. 2003. “The State, the People, and Their Mediators: The Struggle over
Agrarian Law Reform in Post-New Order Indonesia.” Indonesia 76: 87–126.
McBeth, J. 2016. “Indonesia’s Troubled Quest for Food Self-Sufficiency.” The Strategist:
Australian Strategic Policy Institute, December 19. Accessed July 27, 2017. https://www.aspis
trategist.org.au/indonesias-troubled-quest-food-self-sufficiency/.
McCarthy, J. 2006. The Fourth Circle: A Political Ecology of Sumatra’s Rainforest Frontier.
Stanford: Stanford University Press.
McCarthy, J. 2010. “Processes of Inclusion and Adverse Incorporation: Oil Palm and Agrarian
Change in Sumatra, Indonesia.” Journal of Peasant Studies 37 (4): 821–850.
McCarthy, J., P. Gillespie, and Z. Zen. 2012. “Swimming Upstream: Local Indonesian Production
Networks in ‘Globalized’ Palm Oil Production.” World Development 40 (3): 555–569.
McCarthy, J., and K. Obidzinski. 2017. “Framing the Food Poverty Question: Policy Choices and
Livelihood Consequences in Indonesia.” Journal of Rural Studies 54: 344–354.
McCarthy, J., and K. Robinson. 2016. “Land, Economic Development, Social Justice and
Environmental Management in Indonesia: The Search for the People’s Sovereignty.” In
24 N. HAMILTON-HART

Land and Development in Indonesia: Searching for the People’s Sovereignty, edited by
J. McCarthy and K. Robinson, 1–31. Singapore: ISEAS.
McCarthy, J., and Z. Zen. 2013. “Food Security in Indonesia.” Inside Indonesia 114, October–
December, 2013. http://www.insideindonesia.org/food-security-in-indonesia-2.
McCulloch, N. 2008. “Rice Prices and Poverty in Indonesia.” Bulletin of Indonesian Economic
Studies 40 (1): 45–64.
McGregor, K., and K. Setiawan. 2019. “Shifting from International to ‘Indonesian’ Justice
Measures: Two Decades of Addressing Past Human Rights Violations.” Journal of
Contemporary Asia 49 (5). DOI: 10.1080/00472336.2019.1584636.
Mears. L. A. 1984. “Rice and Food Self-Sufficiency in Indonesia.” Bulletin of Indonesian Economic
Studies 20 (2): 122–138.
MFAT. 2017. Disputes with New Zealand as a Principal Complainant. Indonesia Importation of
Horticultural Products, Animals and Animal Products. Ministry of Foreign Affairs and Trade,
New Zealand. Accessed August 13, 2017. https://www.mfat.govt.nz/en/trade/trade-law-and-
disputesettlement/current-disputes/.
Mietzner, M. 2015. Reinventing Asian Populism: Jokowi’s Rise, Democracy, and Political
Contestation in Indonesia. Policy Studies 72. Washington, DC: East West Center.
Mietzner, M. 2017. “Indonesia in 2016.” Asian Survey 57 (1): 165–172.
Neilson, J. 2013. “A Matter of Life or Death for the Indonesian Nation?” Inside Indonesia 114,
October–December 2013. http://www.insideindonesia.org/a-matter-of-life-or-death-for-the-
indonesian-nation.
Neilson, J. 2016. “Agrarian Transformations and Land Reform in Indonesia.” In Land and
Development in Indonesia: Searching for the People’s Sovereignty, edited by J. McCarthy and
K. Robinson, 245–264. Singapore: ISEAS.
Neilson, J., and J. Wright. 2017. “The State and Food Security Discourses in Indonesia: Feeding
the Bangsa.” Geographical Research 55 (2): 131–143.
OECD. 2012. “OECD Review of Agricultural Policies: Indonesia.” October 19. Accessed May 25,
2018. doi: 10.1787/9789264179011-en.
OECD. 2017a. “OECD Agriculture Policy Monitoring and Evaluation 2017.” June 21. Accessed
February 26, 2018. http://www.oecd-ilibrary.org/agriculture-and-food/agricultural-policy-
monitoring-and-evaluation-2017/indonesia_agr_pol-2017-15-en;jsessionid=113c8w3mljjde.
x-oecd-live-02.
OECD. 2017b. “Producer Support (PSE), % of Gross Farm Receipts, 2000–2016.” Accessed
February 26, 2018. https://data.oecd.org/agrpolicy/agricultural-support.htm.
Patrick, I. 2004. Contract Farming in Indonesia: Smallholders and Agribusiness Working Together.
Canberra: ACIAR Technical Reports 54.
Peluso, N. 1992. Rich Forests, Poor People: Resource Control and Resistance in Java. Berkeley:
University of California Press.
Peluso, N., S. Afiff, and N. Rachman. 2008. “Claiming the Grounds for Reform: Agrarian and
Environmental Movements in Indonesia.” Journal of Agrarian Change 8 (2–3): 377–407.
Permani, R. 2013. “Rethinking Indonesia’s Beef Self-Sufficiency Agenda.” Inside Indonesia 114,
October 25. https://www.insideindonesia.org/rethinking-indonesia-s-beef-self-sufficiency-
agenda.
Potter, L. 2016. “How Can the People’s Sovereignty be Achieved in the Oil Palm Sector? Is the
Plantation Model Shifting in Favour of Smallholders?” In Land and Development in Indonesia:
Searching for the People’s Sovereignty, edited by J. McCarthy and K. Robinson, 315–342.
Singapore: ISEAS.
Robison, R. 1986. Indonesia – The Rise of Capital. Sydney: Allen and Unwin.
Robison, R., and V. Hadiz. 2004. Reorganising Power in Indonesia: The Politics of Oligarchy in an
Age of Markets. London: Routledge.
SMERU. 2015. “Food and Nutrition Security in Indonesia: A Strategic Review.” SMERU
Research Institute (January). Accessed March 16, 2018. https://docs.wfp.org/api/documents/
WFP-0000005506/download/.
 JOURNAL OF CONTEMPORARY ASIA 25

Studwell, J. 2013. How Asia Works: Success and Failure in the World’s Most Dynamic Region.
New York: Grove Press.
Sudaryanto, T. 2015. “Structural Change of Indonesian Agriculture: Evidence from Agricultural
Census 2003–2013.” Food and Fertilizer Technology Center for the Asian and Pacific Region,
March 25. Accessed May 25, 2018. http://ap.fftc.agnet.org/ap_db.php?id=398.
Sudaryanto, T. 2016. “Government Policy on Self Sufficiency to Achieve Food Security in
Indonesia.” Food and Fertilizer Technology Center for the Asian and Pacific Region,
September 9. Accessed August 17, 2017. http://ap.fftc.agnet.org/ap_db.php?id=680.
Sulistyo, E. 2017. “Nawacita Reforma Agraria” [Nawacita and Agrarian Reform]. Presidential
Staff Office, September 23. http://ksp.go.id/nawacita-reforma-agraria/index.html.
Syahyuti. 2015. “Farmland Policy in Indonesia for Young Generation.” Food and Fertilizer
Technology Center for the Asian and Pacific Region, June 4. Accessed May 25, 2018. http://
ap.fftc.agnet.org/ap_db.php?id=437.
Timmer, P. 1993. “Rural Bias in the East and South-East Asian Rice Economy: Indonesia in
Comparative Perspective.” The Journal of Development Studies 29 (4): 149–176.
Timmer, P. 1996. “Does Bulog Stabilise Rice Prices in Indonesia? Should It Try?” Bulletin of
Indonesian Economic Studies 32 (2): 45–74.
Trampusch, C., and D. Spies. 2014. “Agricultural Interests and the Origins of Capitalism:
A Parallel Comparative History of Germany, Denmark, New Zealand and the USA.” New
Political Economy 19 (6): 918–942.
Warburton, E. 2016. “Jokowi and the New Developmentalism.” Bulletin of Indonesian Economic
Studies 52 (3): 297–320.
Warburton, E. 2018. “Our Resources, Our Rules: A Political Economy of Nationalism in
Indonesia’s Resource Sectors.” Unpublished PhD Diss., Australian National University.
Wardhana, D., R. Ihle, and W. Heijman. 2017. “Agro-Clusters and Rural Poverty: A Spatial
Perspective from West Java.” Bulletin of Indonesian Economic Studies 53 (2): 161–186.
Warr, P. 2011. “Food Security vs Food Self-Sufficiency: The Indonesian Case.” Canberra: Arndt
Corden Department of Economics, Crawford School of Economics and Government, ANU
College of Asia and the Pacific, Australian National University, Working Papers in Trade and
Development, 2011/04. Accessed August 26, 2017. https://dev.crawford.anu.edu.au/acde/pub
lications/publish/papers/wp2011/wp_econ_2011_04.pdf.
Warren, C., and A. Lucas. 2013. “Indonesia’s Land Titling Program (LAP) – the Market
Solution?” In Land for the People: The State and Agrarian Conflict in Indonesia, edited by A.
Lucas and C. Warren. 93 – 113. Athens: Ohio University Press.
White, B. 2016. “Remembering the Indonesian Peasants’ Front and Plantation Workers’ Union
(1945–1966).” Journal of Peasant Studies 43 (1): 1–16.
Widianto, B. 2013. “The Political Economy of Social Protection Reforms in Indonesia.” In Social
Protection in Developing Countries: Reforming Systems, edited by K. Bender, M. Kaltenborn,
and C. Pfleiderer, 161–175. Abingdon: Routledge.
Winters, J. 2013. “Oligarchy and Democracy in Indonesia.” Indonesia 96: 11–33.
World Bank. 1972. “Report and Recommendation of the President to the Executive Directors on
a Proposed Credit to the Republic of Indonesia for a Beef Cattle Development Project.” Report
No. P-1151, December 26.
World Bank. 2007. World Development Report 2008: Agriculture for Development. London:
World Bank.
World Bank. 2017a. “Agriculture, Forestry, and Fishing, Value Added (% of GDP).” World Bank
Website. Accessed March 12, 2018. https://data.worldbank.org/indicator/NV.AGR.TOTL.ZS.
World Bank. 2017b. “Employment in Agriculture (% of Total Employment).” World Bank Website.
Accessed March 12, 2018. http://data.worldbank.org/indicator/SL.AGR.EMPL.ZS?locations=ID.
Yusuf, A., and A. Sumner. 2015. “Growth, Poverty, and Inequality under Jokowi.” Bulletin of
Indonesian Economic Studies 51 (3): 323–348.
 Renewable and Sustainable Energy Reviews 101 (2019) 231–247

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews

journal homepage: www.elsevier.com/locate/rser

Rethinking renewable energy targets and electricity sector reform in T

Indonesia: A private sector perspective
Martha Maulidiaa, , Paul Darguscha, Peta Ashworthb, Fitrian Ardiansyahc
⁎

a
School of Earth and Environmental Sciences, The University of Queensland, Brisbane, Australia
b
School of Chemical Engineering, The University of Queensland, Brisbane, Australia
c
IDH The Sustainable Trade Initiative Indonesia, Jakarta, Indonesia

ARTICLE INFO ABSTRACT

Keywords: Renewable energy targets announced in 2014 present an opportunity to reform Indonesia's electricity sector
Energy in Indonesia which is dominated by fossil fuels. In this paper we discuss Indonesia's current renewable energy policies and
Private sector investment future outlook for achieving the targets. This paper serves as a literature review of Indonesia's changing energy
Renewable energy policy landscape, as part of a broader research investigating renewable energy targets and the role of the private
sector. Despite Indonesia's wealth of renewable energy resources, numerous studies have identified multiple
constraints to the development of renewable energy, including geographical, institutional and investment fac-
tors. Influential groups are calling for the Indonesian Government to put in place a clear policy framework that
facilitates private sector investment. Therefore, interventions to facilitate investment in energy infrastructure in
Indonesia must address the monopolised power market system that oversees a changing, complex malaise of
electricity pricing regulations which make investment risky and uncertain. This study will enrich the existing
literature on renewable energy policy which emphasises the importance of engaging the private sector. It is
based on a rigorous qualitative assessment of Indonesia's changing policy that affects the progress of the re-
newable energy targets. The lessons from Indonesia's experience may provide insights for policymakers notably
in developing countries.Keywords: energy in Indonesia, private sector investment, renewable energy

1. Introduction
In 2014, Indonesia ambitiously revised its renewable energy targets
(RET) to 23% by 2025 (MRET) and 31% by 2050 (LRET) from a base of
only 6% of the current energy mix (2014) [1]. Achieving such targets
will require clear policy and regulations, a strict mandate and improved
coordination among institutions, not to mention billion dollars of in-
vestment. As a result of relentless economic and population growth,
rising living standards and rapid urbanisation, Indonesia's energy de-
mand is expected to continue to increase [2].

Abbreviations: AFOLU, Agriculture, Forestry, and Other Land Use; ASEAN, Association of Southeast Asian Nations; BAKOREN, Badan Koordinasi Energi Nasional
(National Energy Coordination Agency); BAPPENAS, Badan Perencanaan Pembangunan Nasional (National Development Planning Agency); BPP, Biaya Pokok Produksi
(Main production cost of electricity generation); CO2e, Carbon Dioxide equivalent; DEN, Dewan Energi Nasional (The National Energy Council); DKE, Dana Ketahanan
Energi (Energy Resilience Fund); EBTKE, Energi Baru dan Terbarukan dan Konservasi Energi (New and Renewable Energy and Energy Conservation), Directorate
General under MEMR; FIT, Feed-in Tariff; GDP, Gross domestic product; GHG, Greenhouse Gas; GT, Giga Tonne; GW, Gigawatt(s); IBEKA, Institut Bisnis dan Ekonomi
Kerakyatan (People-Centered Business and Economics Institute), a nonprofit organization; IEA, International Energy Agency; INDC, Intended Nationally Determined
Contributions; IPP, Independent Power Producer; LCOE, Levelised cost of electricity; KEN, Kebijakan Energi Nasional (National Energy Policy); KUBE, Kebijakan Umum
Bidang Energi (General Policy on Energy); MEMR, Ministry of Energy and Mineral Resources; MOF, Ministry of Finance; MRET, Medium-term renewable energy
target; MW, Megawatt(s); MWe, Megawatt(s) equivalent; LPG, Liquefied Petroleum Gas; RET, Long-term renewable energy target; MOPS, Mean of Platts Singapore;
NDC, Nationally Determined Contributions; NEPIO, Nuclear energy program implementing organization; OPEC, Organisation of the Petroleum Exporting Countries;
P2EBT, Tim Percepatan Pengembangan Energi Baru Terbarukan (Task force for accelerating the development of new and renewable energy); PLN, Perusahaan Listrik
Negara (State Electricity Company); PIT, Program Indonesia Terang (Bright Indonesia Program); PV, Photovoltaic; R&D, Research and Development; REPELITA,
Rencana Pembangunan Lima Tahun (Five-year national development plan); RES, Renewable Energy Sourced; RET, Renewable energy target (s); RUED, Rencana Umum
Energi Daerah (Regional/Provincial General Plan on Energy); RUEN, Rencana Umum Energi Nasional (National General Plan on Energy); RUPTL, Rencana Usaha
Penyediaan Tenaga Listrik (Power Supply Business Plan); SOE, State-owned enterprise; UNFCCC, United National Framework Convention on Climate Change
⁎
Corresponding author.
E-mail address: martha.maulidia@uq.net.au (M. Maulidia).

https://doi.org/10.1016/j.rser.2018.11.005
Received 21 February 2018; Received in revised form 12 October 2018; Accepted 3 November 2018
1364-0321/ © 2018 Elsevier Ltd. All rights reserved.
M. Maulidia et al.
A serious underinvestment in the electricity sector makes it chal-
lenging to meet the increasing energy demand [3]. The same reason
will make it impossible to grow the share of renewable energy to almost
four times that in the current mix in less than 10 years. Although these
targets may appear to be unrealistic, in light of the Paris agreement,
Indonesia, like many countries, must begin reforming its energy sector
to transition away from its reliance on fossil fuels. Increasing the use of
available renewable energy resources in Indonesia will reduce the
amount of fossil fuel imports in the future. Energy sector reform that
focuses on renewable energy at the domestic level, alongside global
research and development, will eventually bring down the cost of re-
newable energy. This would then ensure energy affordability and se-
curity for the Indonesian people.
The focus of the reform on long-term renewable energy will require
policy changes that enable Indonesia to mobilise funding, improve in-
stitutional coordination and facilitate technology transfer. To achieve
the medium-term target (MRET), Indonesia needs to provide sufficient
financial resources to build new electricity infrastructure. Due to lim-
ited public finance, private sector investment will be crucial for this
renewable energy development. The MRET of 23% renewables by 2025
requires an estimated investment of IDR 1.600 trillion or around US
$120 billion [4]. At current prices, this figure is around 76% of In-
donesia's government expenditure, US$157.5 billion in 2016 [5], and
13% of the country's gross domestic product (GDP), US$932.3 billion in
the same year [6] or around the same share of agriculture sector's
contribution to the GDP [7]. PLN1’s business plan (RUPTL) 2015–2024
estimates the need of private sector investment in electricity generation
of US$ 63 billion or 47% of the total investment need. PLN has limited
financial capability to fund electricity infrastructure projects in the
country. Accordingly, developing electricity infrastructure requires in-
novative financing schemes and a leap in private sector investment [4].
Eight out of 13 electricity systems in Indonesia have negative re-
serve margins, causing some regions to suffer from daily rolling
blackouts (Fig. 1). These systems cover Sumatra, Kalimantan and most
of eastern Indonesia. Only two systems – Java-Bali and South-and-West
Sulawesi – have reserve margins of more than 20%. Two other systems
experience power deficits, with a reserve margin of below 10% [8]. The
low electricity tariff determined by the government makes investment
in this sector unattractive.
Indonesia is an archipelagic country of 17,500 islands with a total
coastal line of 95,181 km, which creates huge geographical challenges
for electricity distribution. Building electricity infrastructure in remote
areas is not economically feasible, even for the national electricity
company, PLN that is subsidised by the government.
Indonesia's energy policy needs to ensure it is able to meet the
growing demand for energy, as well as ensure security of supply. At the
same time, Indonesia's strategy must also safeguard the environmental
sustainability of its energy supply [2]. Indonesia needs to formulate
policies that correctly consider three equally important aspects of en-
ergy supply – affordability, security and sustainability – known as the
energy trilemma [9]. Failure to balance the three dimensions will ex-
acerbate the problems the energy sector is facing. However, at the
implementation level, it is difficult to provide energy that is affordable,
reliable and sustainable. Available and cheap coal and biofuels are very
likely to cause environmental and social problems [9]. Renewable en-
ergy seems to be the only long-term solution to the energy trilemma,
although at present it is not competitive with coal.
To tackle the energy shortfall it is likely to face in the future,
Indonesia cannot delay the reform of the energy sector. Indonesia will
not only need to improve the institutional settings under which it op-
erates, but also the size of investment required for energy infra-
structure, its overall energy policy planning and implementation and
1
Perusahaan Listrik Negara, the state utility company who hold monopoly
rights of electricity business in Indonesia.
232
Renewable and Sustainable Energy Reviews 101 (2019) 231–247
how its energy markets and pricing policies are structured [2]. It must
take into account the energy trilemma into energy reform – and the
development and deployment of renewable energy will become crucial.
To assist in the development and deployment of renewable energy,
Indonesia needs to create a policy that makes it attractive for private
sector investment. The private sector requires a policy that can upgrade
its risk and return profile [10]. In well-developed regions, renewable
energy-sourced [11] electricity will compete with highly subsidised,
cheap coal power plants. For geographically remote areas, the total lack
of infrastructure makes electrification expensive. Without government
intervention, it is implausible to expect private companies to invest in
electricity for remote villages [12].
In reality, contradictive policies, substantial fossil fuel subsidies and
unclear mandates, combined with bureaucratic processes, lack of ac-
countability and limited institutional capacity, have hampered the de-
velopment of renewable energy in Indonesia. Although the Energy Act
2004 clearly states that renewable energy shall be prioritised, the reg-
ulations for implementation do not favour renewable energy over fossil-
fuel-powered electricity. Indonesia's old model of development that is
highly dependent on fossil fuels prevails, and seems to be the only
pattern that will continue in the next 20 years. The absence of a clear
policy and regulatory framework to support Renewable Energy Sourced
(RES)-electricity will make the MRET impossible to achieve within the
specified timeframe.
2. Indonesia's irony: rich in energy resources but poor in
electricity
Indonesia is known for its abundant energy resources – notably coal,
natural gas, and geothermal– but, ironically, it is facing an energy and
power crisis. Despite the efforts that have been made to improve the
performance of the electricity sector, Indonesia needs to work harder to
compensate for its under investment since the financial crisis in 1997.
Between 1997 and 2007, no single power plant was built. Moreover,
with almost 62% of the grid system having negative reserve margins,
Indonesia is at a crossroads for making the required changes to rectify
the root of the problem to be able to meet the high rising demand for
electricity, along with a growing population and need for economic
growth. This requires a thorough reform of the energy sector.
At the same time, Indonesia's GHG emissions have been rising due to
land use changes (deforestation and peat fires) and the fossil fuel
burning. The use of coal and diesel oil is expected to increase sig-
nificantly in the next few years if renewable energy is not rapidly de-
ployed. Renewable energy needs to be prioritised in the electricity re-
form agenda to avoid even higher GHG emissions and environmental
and health problems caused by coal and other fossil fuels. The
International Energy Agency (IEA) suggests that, in the next 5 years,
Indonesia could apply a bridging strategy (i.e. taking short- and
medium-term measures) before taking more impactful actions such as
retiring the least-efficient coal power plants, encouraging more in-
vestment in renewable technologies, and ending wasteful energy con-
sumption by implementing fossil fuel subsidy reform [2].
The acceleration of electricity infrastructure development requires
significant financial investment. The easiest, and possibly cheapest, way
to meet rising energy demand in the short term is to build large-scale
power plants that burn coal, due to its availability and low price.
Indonesia's coal production declined in 2014, 2015 and 2016 after
reaching the highest record of 474 million tonnes of produced coal in
2013 [13]. While Indonesia's production declined, other top coal pro-
ducing countries such as China, the US, Australia and India produced
more coal, placing Indonesia as the fifth highest coal producing
country [14]. Before 2016, more than 75% of Indonesia's coal was ex-
ported [15] mainly to China, India and Japan [16]. However, owing to
China's slowing economic growth, shifting of industrial and energy
focus forced by severe air pollution problem [17], more efficient energy
consumption per unit GDP and the shift towards solar [18], its demand
M. Maulidia et al.
Fig. 1. Electricity crisis and deficit in Indonesia, reused with permission from Ministry of Energy and Mineral Resources [8].
for Indonesia's coal has decreased [19]. Despite the slight increase of
coal export to China in 2016 as such, since 2015, the majority of In-
donesia's coal is being consumed in the domestic market [20]. The
declining domestic production also forced Indonesia to import coal
[20]. Because of this, a transition using coal in the short to medium
term may be necessary before leapfrogging to a more sustainable re-
newable energy regime. However, this transition phase needs to be
carefully planned, with the right balance of meeting energy demand
while minimising environmental impacts. Indonesia may learn from the
experience of Germany, which in 2011 began phasing out nuclear
power as part of its Energy Transition Program. The transition phase
still includes fossil fuels, but gradually changes into non-nuclear re-
newable energy. In the transition phase, before renewable resources can
replace the power from nuclear, Germany continues to use coal and
imports electricity from neighbouring countries [21]. By integrating a
social component into the analytical framework for adopting the nu-
clear phase-out policy, Germany opted to fulfil environmental justice
principles rather than energy security. Integrating the two goals in
practice is more difficult in practice [22]. It is estimated that the wel-
fare losses that the German society had to bear as a result of generating
power to replace their nuclear power amounted to €45.83 billion [23].
Another example of the so-called green policy transition is exemplified
in California's target of achieving 33% renewable energy by 2020. This
had been predicted to cause rolling blackouts, although the last re-
corded blackout happened in 2001. However, to be able to achieve the
target, the energy expended per unit of electricity generated is expected
to increase, which means electricity prices will also increase [24].
Despite Indonesia's long experience with the oil industry, it does not
appear to make the country any better at managing its resources.
Although Indonesia used to be one of the largest oil-exporting coun-
tries, it has been dealing with the problems of aging fields and lack of
new investment. The high revenue achieved from oil exports that lasted
for more than 5 decades was not proportionally used for reinvestment.
As a result, Indonesia changed dramatically from being a net oil ex-
porter to a net oil importer in 2004, forcing it to leave Organisation of
the Petroleum Exporting Countries (OPEC) in 2008. Indonesia's oil and
gas experience provides important lessons that can be applied to the
development of new and renewable energy policies. First, production-
233
Renewable and Sustainable Energy Reviews 101 (2019) 231–247
sharing contracts should be rewarded on a competitive basis to reliable
companies that do not necessarily provide the cheapest service, but can
offer high reliability. Second, the objectives of state-owned companies
tend to be intertwined with political agendas, giving rise to the in-
efficient market, and therefore this situation needs to be avoided.
Indonesia is blessed with geothermal resources, but does not use
them well. Located in the Pacific Ring of Fire, it has the world's largest
geothermal resources (40%), with various estimates of potential from
24,000 MW equivalent (MWe) [25] to more than 28,900 MWe and total
reserves of at least 16,580 MWe ( [26]. High-risk exploration combined
with forestry permits and tariff negotiations have hampered geothermal
development. By 2017, only a tiny fraction of the total proved reserves
or 1808.5 MW had been used. Indonesia needs to overcome these bar-
riers to succeed in optimising its potential in geothermal energy.
Modern energy services are essential for all economies, and the lack
of them can pose a threat to human security [27]. In Indonesia, around
5% of the population remains unconnected to the grid [11], and the
majority of people living outside of Java and Bali have no, or restricted,
access to cheap, reliable and sustainable electricity. Moreover, poor
infrastructure raises logistics costs of power generation significantly.
2.1. Fossil fuel and electricity subsidies
In contrast to its aspiration in climate change mitigation, Indonesia's
domestic energy policy is still troubled by the discourse over fossil fuel
and electricity subsidies. The subsidies reached US$18.3 billion, or al-
most 30% of the total state budget expenditure, in 2011 alone and
unpredictably leapt to US$28 billion in 2014, 24.1% more than the
initial budget [28]. This figure is predicted to continue to increase each
year as the demand for oil and gas and electricity surges. Before 2010,
the extensive government spending on fossil fuel subsidies was even
higher than the combined spending on defence, education, health and
social security [29]. In 2014, spending on fossil fuel subsidies was 31
times higher than spending on health (data from [30]). These subsidies
are unhelpful to the economy because imports put pressure on the
budget and the subsidies could be better used for spending on educa-
tion, health and infrastructure development [31]. A gradual removal of
fossil fuel and electricity subsidies, combined with correct reallocation
M. Maulidia et al. Renewable and Sustainable Energy Reviews 101 (2019) 231–247

Fig. 2. Fossil fuel and electricity subsidies in Indonesia [30,33–38].

schemes, will benefit the economy, producing an increase in real GDP

gains of 0.4% in 2020 [32]. The trend of increased spending on fossil
fuel and electricity subsidies as a percentage of government ex-
penditures from 2008 to 2014 is illustrated in Fig. 2.
In addition, pouring more money into fossil fuel subsidies may re-
sult in a government budget deficit. A budget deficit that approached
the legal limit of 3% forced the government to take action to gradually
reduce the level of subsidies, which raised the price of fuels and elec-
tricity tariffs in 2014. The reform of fossil fuel subsidies is beneficial for
the economy and the battle against climate change, but it has short-
term negative impacts on low-income households and small and
medium enterprises.
Since 2014, attempts have been made to reduce the fossil fuel
subsidies further. Due to a lower global market price for gasoline,
Indonesia has lowered the consumer price but no longer subsidises
‘premium’ gasoline. However, the government still subsidises kerosene,
diesel and LPG. In 2016 the government subsidies for electricity was
reduced by US$1.6 billion by removing subsidies to medium-sized
households. This is seen as a good beginning for a more structured
approach to removing the fossil fuel subsidy program.

2.2. GHG emissions profile, renewable energy, climate change mitigation

Indonesia's economic and population growth continues to cause

increased GHG emissions. The National Planning Agency (Bappenas)
predicts that, without any interventions, Indonesia's GHG emissions will
hike from slightly below 1.5 GT CO2e in 2010 to 1.8 GT CO2e by 2020
and reach 2.9 GT CO2e by 2030 (Fig. 3) [39].
In 2010, around 62% of Indonesia's total GHG emissions were from
land use change (deforestation and peat fires). This proportion has been

Fig. 4. GHG emissions share per sector 2010, 2020 and 2030, reused with
permission from Medrilzam, 2015 [39].

at least 60% since 1990. However, forecasts suggest that the energy
sector will see the highest surge (160%) and contribute the largest share
(45%) of GHG emissions by 2020 and 50% by 2030 (Fig. 4) [39].
In the short term, Indonesia's rising demand on electricity will be
fulfilled by two common options: cheap and abundant coal and self-
generation diesel plants. Diesel generators and power plants although
more expensive, are commonly found in remote areas in Indonesia
where there is no, or limited, access to the power grid. The increasing
use of coal and diesel will result in higher GHG emissions. Moreover,
although diesel has a lower GHG emissions factor than coal, it is heavily
subsidised and mostly imported, which means it places a high fiscal
burden on the country's economy.
Fig. 3. Indonesia's projected GHG emissions 2000–2030, reused with permis-
In the longer term, with Indonesia's economy and population
sion from Medrilzam, 2015 [39].

234
M. Maulidia et al.
growing so fast, the need for renewable energy becomes certain.
However, the country needs out the right strategy and policies in place
to prepare for the renewable era.
3. Slow transition from fossil fuels to renewable energy
Since the Dutch colonial era, Indonesia's energy policy has sup-
ported the oil and gas industry because the country's economy relied
heavily on these fuels. Even after Indonesia became a net oil importer in
2004, it has continued its support for the oil and gas sector, in sharp
contrast with the power sector where the government appears to have
neglected investment. This under-investment has resulted in a low
electrification ratio compared with other ASEAN countries.
Furthermore, PLN has used only a small fraction of the potential re-
newable energy resources such as geothermal, solar PV and biomass.
However, aging oil fields, stagnant exploration and depleting oil
and coal reserves are forcing Indonesia to formulate a new strategy for
energy supply and security. Domestic energy production cannot meet
Indonesia's growing demand; hence it requires an overall strategy and
regulatory framework to transition to renewables [40]. To catch up on
the slow utilisation, Indonesia needs firm commitment from the policy
makers to accelerate renewable energy development. This requires the
creation of achievable targets and enabling policies. Analysing In-
donesia's energy policy through the energy trilemma lens will reveal the
most effective policy that balances the three prongs: energy equity,
energy security and environmental sustainability.
3.1. Oil and gas
Dutch companies commenced the first oil exploration in 1884 [41].
Even before independence, US companies started investing in In-
donesian oil. Following independence in 1945, the government began
nationalising oil companies. Pertamin was founded in 1945 and 12
years later through a merger of several companies, Pertamina was es-
tablished. From the 1970s to 1980s, Indonesia enjoyed high revenues
from oil exports, with more than 70% of export value coming from oil.
It peaked in 1981 when economic growth reached 7.9% [41].
However, despite the government's efforts to diversify energy
sources, the increasing domestic energy demand forced Indonesia to
finally become a net importer of oil in 2004 and it subsequently left
OPEC in 2008. In spite of subsidy reform, falling oil production and
growing consumption (Fig. 5) may result in Indonesia being one of the
world's largest oil importers. The inadequate investment to diversify the
Fig. 5. Indonesia's oil production, consumption and reserves 1998–2015 [14,42–44].
235
Renewable and Sustainable Energy Reviews 101 (2019) 231–247
energy supply has been attributed to poor governance [45], inefficient
bureaucracy [46] and unclear regulatory frameworks [2,47].
The decreasing oil reserves have forced Indonesia to rethink its
strategy, particularly in light of recent oil exploration activity. Between
2009 and 2013, companies exploring the offshore basins in eastern
Indonesia have been reported to have lost US$1.6 billion [48]. To re-
vive its weakening oil production, Indonesia revised its government
regulation to offer tax elimination for oil exploration and tax facilities
for oil exploitation [49]. In addition, through the issuance of Minister of
Energy regulation in 2017, it changed the plans of development from a
block-basis to a national-basis and change the cost of recovery to gross-
split mechanism promising a higher share of return for investors. Apart
from complicated tax regime in the country, these measures are likely
to be more attractive to investors and may help to rebuild confidence in
continuing exploration for oil. However, their effectiveness is yet to be
seen.
3.2. Coal
Coal reserves are found mostly in Kalimantan and Sumatra.
Although abundant, the majority of Indonesia's coal is low grade lignite
with a high moisture content. Only 12% of Indonesia's total coal re-
serves can be classified as high-quality coal with a calorific value of
more than 6100 kcal/kg [50]. During the 1980s, in an effort to diversify
energy sources, Indonesia encouraged foreign investment to boost coal
production. To do that, Indonesia signed coal production-sharing con-
tracts that have favourable terms for companies. Apart from this di-
versification effort, Indonesia aimed to dominate the international
market and gain more revenue from coal production. To compete with
the subsidised oil, heavy subsidies have been given to domestic con-
sumers of coal. Weak, short-term oriented policy and a power struggle
between national and subnational governments have prevented longer
term investment, exploration and environmental impact manage-
ment [19]. All of the above will eventually lead to Indonesia's coal
production decrease.
Indonesia's policy on coal is mainly driven by external forces in
particular China's demand rather than domestic interest [51]. China's
economic boom between 2000 and 2011 required large amount of re-
sources, especially coal [19]. High coal exports to China have con-
tributed to Indonesia's economic growth, increased real GDP and led to
more coal consumption [52]. Indonesia was the largest coal exporter
during the boom [19]. Indonesia's high coal exports had several mac-
roeconomic impacts that were not always positive. High coal prices
M. Maulidia et al.
contributed to economic growth, but also to high expenditure. After
coal prices started declining, Indonesia had to deal with a budget def-
icit, but still needed to continue spending on critical activities [19].
Indonesia has benefited from both the oil boom in the 1970s and
1980s and the coal boom from 2000 to 2011. The main policy has been
to support production and provide subsidies for consumption. This
appears short sighted, with a lot of important aspects, such as re-
investment, being neglected. Exploration of oil has been stagnant since
early 2000s (data from [42]). An oil reinvestment fund was suggested
by the Ministry of Energy and Mineral Resources (MEMR) and other
non-government actors, but it was never realised. The proposed fund
intended to levy premium oil exports for reinvestment in exploration.
The government's requirement for foreign companies to divest their
shares on mining and a seemingly plunging demand for coal around the
world, since 2015 Indonesia has been using most of its coal for domestic
purposes. New exploration of coal has not been a priority since the
global coal price dropped in 2014. However, Indonesia needs coal to
supply the power plants under the 35 GW electricity program. Without
new coal exploration, Indonesia might use up all its existing coal re-
serves between 2033 and 2036 before the 35,000 MW program finishes
[53].
3.3. Nuclear
Indonesia has three research nuclear reactors and one cobalt ra-
diator for agriculture but no large-scale nuclear power production fa-
cilities. In the National Energy Plan 2014, the government decided that
nuclear will be the last option to consider, at least until 2050 [1]2.
In the previous National Energy Plan 2006, President Yudhoyono
sought to build a nuclear power plant within 10 years after the reg-
ulation was issued [54]. Muria, Regency of Jepara, Central Java was
selected as the first nuclear power site. But it received opposition from
the local people [56]. Indonesia's Nuclear Energy Agency (BATAN)’s
preliminary studies on site selection of nuclear power plants in 2011
and 2013 came up with two new feasible sites: Bangka Belitung and
Banten [56,57]. Nuclear continues to be a polemic among the stake-
holders. Some parliament members are known to be the proponent of
nuclear and regretted Indonesia's position towards nuclear as stated in
the Presidential Regulation on National Energy Plan in 2014 [58].
In 2016, President Widodo instructed the National Energy Council
(DEN) to formulate a nuclear development roadmap. DEN came up with
the roadmap, however it was later found that the mandate is not
aligned with DEN's main duties and functions. Although a few members
of (DEN) appears to oppose the development of nuclear power plants,
some agreed to start developing nuclear in the near future considering
the advantages for emissions reduction and energy security while
prioritising renewable energy [59].
The roadmap was taken over by the Ministry of Energy and Mineral
Resources (MEMR), who together with the Committee of National
Economy and Industry (KEIN) established a working group to work on
the roadmap. This working group recommends to build nuclear power
facilities immediately and urges the government to make a firm deci-
sion to go nuclear and issue a regulation to establish a Nuclear Energy
Program Implementing Organisation [60]. The nuclear development
roadmap 2018–2025 is now in the final draft version as at December
2017 (Fig. 6).
The roadmap details the action plan required to develop Indonesia's
2 tion. It proved to be successful in achieving intensification, but the
Paragraph 2, Article 11 Point 2 and 3 state that “… the priority of national
energy development will be based on (a)…, (b)…, (c) optimising the utilisation of gas
and renewable energy (point 2) except nuclear that will be utilised by considering the
security of supply of national energy in a large scale, carbon emissions reduction and
by prioritising new and renewable energy according to the economic value and
consider it (nuclear) as the last option while strictly paying attention to the safety
factor.”.
236
Renewable and Sustainable Energy Reviews 101 (2019) 231–247
first nuclear power plant by 2025. It also contains strategy to meet the
remaining 3 issues that Indonesia has not fulfilled yet out of 19
Infrastructure Issues as set out in the International Atomic Energy
Agency (IAEA) Milestones Approach of the Guidance Document for
Countries Introducing Nuclear Power [61]. In 2018, the Commission 7
of the House of Representatives has been having a series of discussion
regarding the plan of development of nuclear power facilities with
BATAN and Indonesia's Nuclear Energy Regulatory Agency (BAPETEN)
[62]. It shows that the plan to build a nuclear facility in the near future.
Depleting oil resources and low electrification rate are the two
factors that motivated the push for alternative energy including nuclear
in Indonesia [63]. Slow adoption of renewable energy has forced the
policymakers to rethink the possibility of nuclear development, even
though the 2014 National Energy Policy mandated otherwise. The
roadmap produced by the Ministry of Energy which favours the accel-
eration of nuclear power plants development is a further evidence that
nuclear discourse is no longer taboo but necessary to address the energy
insecurity. The IAEA and nuclear technology provider countries such as
Russia, France, China, South Korea and Japan have been promoting
nuclear to Southeast Asian countries. This lobbying move seems to
convince the Ministry of Energy, the National Energy Council and the
parliament. Malaysia, Indonesia, Thailand and Vietnam are among
Southeast Asian countries who have expressed their interests in devel-
oping nuclear power plants, with Vietnam being the most ad-
vanced [64]. The regional push is a logical rationale behind the move to
pursue nuclear together with other countries in the region. If a few
countries in the same region are going ahead with nuclear, the risks can
be shared proportionally while enjoying the same benefits [65]. How-
ever, Indonesia's plan still has many questions left unanswered such as
the high cost of nuclear technology. The electricity pricing should be
within the range that does not require another form of subsidies. For
many years, Indonesia does not move forward with nuclear energy due
to the issue of public distrust. Indonesia's notorious corruption culture,
inefficiency and low coordination among government institutions are
the common perception among the public that results in low acceptance
of nuclear [63]. Only if the current government shows solid evidence in
the holistic reform of judicial, executive and legislative branches, In-
donesia's nuclear plan may be realised.
3.4. Renewable energy in Indonesia's energy policy
Although not necessarily being implemented entirely, Indonesia's
energy policy has been evolving from what was known as the mining
law in the old colonial era in 1899, then a petroleum-focused policy, to
a policy that gives more consideration to conservation, diversification,
clean and renewable energy since 1987. To improve coordination
across the energy sector, in 1980 the President mandated the estab-
lishment of the Energy Coordination Agency (Badan Koordinasi Energi
Nasional/ BAKOREN). Later in 2008, the National Energy Council
(DEN) was established with similar tasks as BAKOREN's but headed
directly by the President to enable cross-sectoral coordination (DEN
2016). Compared with BAKOREN, DEN has more representatives from
the private sector and academia. In 1981, BAKOREN formulated the
energy sector policy (Kebijakan Umum Bidang Energi/ KUBE), with the
main aim of maximising the utilisation of energy resources. This policy
was revised in 1987, 1991 and 1998 (See Fig. 7).
As discussed earlier, the energy policy during the oil boom era in the
1980s and 1990s simply focused on the intensification of oil produc-
tion, diversification of energy resources to coal and energy conserva-
government's lavish and inefficient fossil fuel subsidies hindered the
third objective of this policy. The increasing energy consumption and
the declining oil production motivated the shift from oil to coal and
other resources. In 1998, KUBE included energy pricing and environ-
ment as additions to the previous three objectives. KUBE 1998 re-
cognised that Indonesia had to develop clean energy to minimise the
M. Maulidia et al. Renewable and Sustainable Energy Reviews 101 (2019) 231–247

Fig. 6. Roadmap of nuclear energy development in Indonesia 2018–2025 [61].

Fig. 7. Evolution of Indonesia's Energy Policy, modified from [66].

237
M. Maulidia et al.
negative impacts on the environment. Although categorising clean and
renewable energy solely as a solution to environmental problem is an
understatement, this was Indonesia's first acknowledgement of the
importance of clean energy.
The national energy policy (Kebijakan Energi Nasional/ KEN), en-
acted as a presidential regulation in 2006, was a critical milestone that
set out a measurable energy mix that included not only fossil fuels but
also renewable energy sources. The revised version of KEN formalised
in 2014 mandates an even higher share of renewable energy. KEN ap-
pears to prioritise new and renewable energy, but the drive behind this
policy is simply to optimise the energy mix. This document stipulates
that renewable energy should be at least 23% of the energy mix by 2025
[1]. It projects a significant growth of renewable energy and a decline
in the share of fossil fuels in the energy mix. In the 2025 projected
energy mix, coal's share will be kept at 30%, becoming the largest en-
ergy source for Indonesia. By the same year, oil and gas will make up
25% and 22% of the energy mix, respectively.
Indonesia's proven coal reserves are declining and similarly, oil and
gas reserves have seen a decline in the past decade and the trend is
likely to continue. These important facts seem to be neglected in the
policy design of KEN. The policy predicts that the energy mix in 2025
will still be dominated by fossil fuels, but Indonesia's depleting fossil
fuel reserves make energy imports unavoidable. Importing fossil fuels
will likely do more harm than good from the perspective of energy
security [27].
KEN 2014 has implementing regulations at both the ministerial and
sub-national levels. At the practical level, KEN is detailed in the na-
tional general plan on energy (RUEN) and provincial general plan on
energy (RUED). The detailed renewable energy targets in this document
are determined using a modelling exercise undertaken by the MEMR.
Indonesia is committed to take various steps to overcome the bar-
riers to geothermal development. The law on geothermal has been re-
vised in 2014 to include new article on profit sharing to local govern-
ment. Also, the new law does not categorise geothermal as a mining
activity, so that exploration and development can be done on conserved
forest areas [67]. In addition to that, Indonesia adopted pricing policy
to support private sector investment in renewable energy including
geothermal, before the minister changed the regulations in 2017.
Besides the national energy policy, between 2014 and 2016, the
Indonesian Government launched a few initiatives to push renewable
energy development. These included: the acceleration of electricity
infrastructure of 35,000 MW; setting up a task force for accelerating the
development of new and renewable energy (P2EBT); the Clean Energy
Centre of Excellence; an Energy Security Fund (DKE); the Eastern
Indonesia electrification/ Bright Indonesia Program (PIT); the Bali
clean energy region; micro-grids; and one-stop-shop permits [68].
These initiatives complemented regulations on FITs for geothermal,
solar-, hydro- and biomass-sourced electricity in 2015 and 2016.
Unfortunately, the change of Minister of Energy and Mineral
Resources in mid 2016 shifted away from the focus on renewable en-
ergy. It began by the launching of a new ministerial regulation on re-
newable energy tariffs that were seen as a discouragement for private
investors [69]. The new regime was also marked by the cancellation of
FITs, dissolution of strategic pro-renewable units within the ministry
[70] and termination of programs including the Energy Security Fund,
the Bright Indonesia Program and the Centre of Excellence for Re-
newable Energy.
3.5. Climate policy
Indonesia has taken important steps in climate change policy,
mainly driven by the international negotiations (Fig. 8). In 2016, In-
donesia submitted its voluntary greenhouse gas (GHG) emissions re-
duction plan to the United Nations Framework Convention on Climate
Change (UNFCCC). In the document known as Nationally Domestic
Contributions (NDC), Indonesia pledged to reduce its GHG emissions of
238
Renewable and Sustainable Energy Reviews 101 (2019) 231–247
29% against 2010 baseline to be achieved by 2030. Twenty three per-
cent of the total emissions reduction goal will have to come from the
energy sector [39]. Development of clean coal power plants and the use
of renewable energy sources for electricity production are among the
six mitigation measures that Indonesia put forward in the document.
Indonesia's NDC targets 19.6% of electricity production to be
sourced by renewable energy, compared with the use of coal power
plants in the business as usual scenario [71]. It requires the develop-
ment of additional renewable energy power generation of 7.4 GW by
2030. This target has simply been copied from the general plan of
power generation development (RUPTL) 2016–2025 document, devel-
oped by PLN. It indicates that this target will be achieved in any case
because it is a part of an approved business plan. There is no additional
emissions reduction resulting from the development of 7.4 GW of re-
newable energy-sourced power, so Indonesia's pledge to ‘reduce emis-
sions’ is misleading. In general, the basis of determining the targets
which includes baseline, methodology and stakeholder involvement is
questionable [72]. Indonesia also lacks the tools to monitor and verify
the emissions reduction goal in energy sector as set out in the NDC [73].
In the NDC document, Indonesia also planned to pursue low-emis-
sion coal technology, such as supercritical and ultra-supercritical power
plants. It suggests that 75% of all coal power plants in Indonesia will
use clean coal technology by 2030. Although this target is not legally
binding, Indonesia needs to connect its international ambition and the
domestic reality. It has to prepare the policy and regulations and pro-
vide sufficient funds to achieve the announced emissions reduction. The
right instruments, such as pricing policy and incentives, may attract
private project developers to use more expensive, but cleaner, tech-
nology.
4. Renewable energy targets: a wishful thinking?
The renewable energy target set out in the national energy policy
(Fig. 9) is only one of several important milestones set by the govern-
ment. The others include provision of primary energy, utilisation of
primary energy per capita, provision of electricity generation capacity,
electrification ratio, energy conservation target and GHG emissions
reduction [74]. At least 45.2 GW of electricity infrastructure capacity
will need to be developed to meet the MRET of 23% of the total energy
mix in 2020 [74]. Most of the energy resources that will be used to meet
this target are large hydro and micro-hydropower (46%). Geothermal is
expected to contribute to 15.3% of the overall target. Other energy
resources include solar, bioenergy (biomass), mixed diesel and biofuel,
wind and ocean. While this move should be applauded, challenges to
achieve the target still lie ahead. This target requires a large-scale in-
vestment in a relatively short period of time.
Indonesia's renewable energy targets are important because it shows
Indonesia's awareness of renewable energy as a possible solution for
electrification, security of supply and reducing environmental impacts.
The targets also represent more forward thinking in policy evaluation
that moves from abstract to measurable targets. With quantification, it
is easier to measure the impact of the policy. Although not perfect, the
renewable energy targets are a stepping stone towards more cohesive
and implementable energy policy. Fig. 10 below shows a clear dis-
tinction between the actual shares from year to year vs the target.
The task to fulfil the energy demand appears to be onerous. Under
RUEN, the existing capacity of geothermal power plants will be in-
creased five-fold (from 1438 MW in 2015) in less than 10 years.
Hydropower capacity will be increased four-fold, while other RES
power plants capacity will be increased from negligible to three-fold
(bioenergy/biomass), over 80-fold (solar) and almost 600-fold (wind)
of their current installed capacities. (Table 1).
The allocation of renewable energy targets was developed through a
modelling exercise by the MEMR after a series of consultations with
stakeholders. The choice of renewable energy types seems to be based
on the potential resource or proved reserves of each energy source .
M. Maulidia et al. Renewable and Sustainable Energy Reviews 101 (2019) 231–247

Fig. 8. Indonesia's climate change policy.

Geothermal, hydro and bioenergy are given higher targets propor-

Fig. 9. Indonesia's current and targeted energy mix [74,75].
tionate to their respective potential resources.
The barriers to developing renewable energy projects seem to have
been overlooked in the process of defining the target allocation. The
policy makers could have learned from the experience in developing
renewable energy in the past, which shows an observable pattern that
can be used as a basis for policy design. This pattern will be useful in
determining more realistic future scenarios. The targets need to con-
sider several important aspects to make them implementable.
Most renewable energy projects take time to develop, especially
geothermal plants that will require a few years in the exploration stage.
A typical geothermal power plant in Indonesia will need at least 7 years
before it can start supplying electricity. Large hydropower plants will
have to deal with the relocation of multiple villages, which have critical
social impacts. Therefore developing large hydro projects will take
more time and efforts to comply with more stringent social and en-
vironmental standards. Similarly, the development of biomass power
plants faces both similar and specific challenges. Biomass power will
require substantial amount of land for the plantation of biomass feed-
stocks. An integrated system of plantation and the power plant is es-
sential to ensure the sustainability of the supply. Uncertainty in pricing
policy and negotiation of tariffs with PLN (the single buyer) may delay
the project and lead to financial consequences.
All the issues above relate to two critical challenges faced by re-
newable energy projects: time and cost. Renewable energy project de-
velopers have to manage specific risks in addition to the usual elec-
tricity project risks. The time period is obviously not properly taken into
account in the design of the renewable energy targets. It took 3 years
from the policy design to detailed action plans launched in 2017.
Likewise, the financing sources and implementation mechanisms are
not considered in the renewable energy targets design.
The renewable energy targets policy does not differentiate two
contrasting situations in Indonesia. Densely populated islands in the
west and less developed regions in the east may require two different
approaches in developing RES electricity. The western part of
Indonesia, consisting of Java, Bali and Sumatra, has a relatively ade-
quate electricity supply and more reliable grid systems. In contrast, the
middle and eastern provinces in Indonesia are facing electricity crises,
with power deficits and negative reserve margins, causing the regions
to suffer from daily rolling blackouts. To supply electricity in the rural
regions, two existing options are costly grid extensions or the use of
dirty and expensive, but easy to install, diesel generators. In theory,
renewable energy is a viable substitute for these two practices. RES
mini grids, isolated from the centralised grids, can improve electricity

239
M. Maulidia et al. Renewable and Sustainable Energy Reviews 101 (2019) 231–247

Fig. 10. Renewable energy target vs actual share [75].

supply efficiency and in turn reduce GHG emissions. (Fig. 11).
The legal basis of RET is Law No. 30/2007 on Energy and
Government Regulation No. 79 Year 2014 on National Energy Policy
(KEN). Law No. 30/2007 clearly states that energy security should be
achieved through diversified new and renewable energy resources [77].
Meanwhile, in Government Regulation No. 79 of 2014, the Indonesian
Government emphasises the need to maximise the use of renewable
energy in national energy development depending on the economic
level. The regulation also recommends minimising the use of petro-
leum, but still encourages coal as a main resource for a dependable
supply of national energy.
The term “the economic level” in Government Regulation No. 79 of
2014 is not clearly defined. It is not clear at what cost can the state
afford to develop renewable energy. A technical level regulation,
Presidential Regulation No. 4 Year 2016 on Acceleration of Electricity
Infrastructure Development, tries to answer this question. This pre-
sidential regulation mandates the use of new and renewable energy to
be prioritised in order to achieve the target proportion of the new and
renewable energy in the energy mix. This regulation also mandates
measures to support renewable energy such as: fiscal incentives; sim-
plified licences and procedures; setting the electricity purchase price;
establishing of a separate business entity to supply power; and the
provision of subsidies for renewable energy [78]. The new Minister of
Energy's regulation on renewable energy [79]deviates from the pre-
sidential regulation especially with regards to electricity purchase price
and the provision of subsidies.
The spirit of this regulation is to provide strong support to renew-
able energy by not merely focusing on economic benefits but also by
taking into consideration energy security and energy sustainability.
Although the main actor mandated by this regulation is PLN (the state-
owned company holding monopoly rights in electricity business
chains), the complex interconnections among government agencies in-
volved – including the Ministry of Energy, Ministry of Finance and the
Ministry of State-Owned Enterprises to which PLN reports – may be a
handicap in realising the acceleration in a timely manner.
4.1. Implications for the private sector
There is no dispute that private sector participation is imperative to
address the underinvestment in electricity infrastructure. Table 2 below
summarises problems in renewable energy sourced (RES) electricity in
Indonesia.
Private sector participation is not without hurdles. Difficulties in
engaging the private sector in Indonesia are not limited to renewable
energy but are rather generic. The private sector's role in the power
sector is limited due to challenges with bureaucracy, especially the
process of permit issuance. Different licences from different level of
authorities are required to build electricity infrastructure. In addition,
power generation from renewable energy such as hydro and geothermal
require even more scrutiny due to their location in forest areas and
potential competing land use issues [8]. Moreover, long winding tariff
negotiations have hampered the development of electricity generation.
As a result, the planned acceleration program to increase Indonesia's
power capacity has been unsuccessful [8].
Other challenges include land acquisition and overall quality of the
performance of contractors and/or developers [8]. Learning from the

Table 1
Renewable energy sources and targets [76].
Renewable energy source Potential resources (GWe) Baseline (2015) Updated (2018) installed Targeted Percentage of Increase from
installed capacity (GW) capacity (GW) capacity (GW) total target current capacity

Geothermal 29.544 1.4385 1.9245 7.2 15.9 5 times

Large, mini- and micro-hydro (MMH) 75.091; 19.385 (MMH) 4.827; 0.197 (MMH) 4.928; 0.303 (MMH) 18 46.5 4 times 15 times
3 (MMH) (MMH)
Bioenergy 32.654 1.671 1.841 5.5 12.2 3 times
Solar 207.898 (4.8 kWh/m2/day) 0.0785 0.09 6.5 14.4 83 times
Wind (and wind-diesel hybrid) 60.647 (≥ 4 m/s) 0.00307 0.0781 1.8 4 581 times
Others (bioenergy-diesel hybrid; ocean) 17.989 (ocean) 0.0003 (ocean) 0.0003 (ocean) 0.307 3.1 6.9 8 times
0.307 (total with (total)
hybrid bioenergy)
Total 45.2 100

240
M. Maulidia et al. Renewable and Sustainable Energy Reviews 101 (2019) 231–247

Table 2
Summary of problems in renewable energy sourced electricity in Indonesia.
Problems Ideal situations Possible solutions

Monopolised power sector [80]
Substantial fossil fuel subsidies [28,81]
Under-investment in new power generation capacity [3]
Limited institutional capacity [82]
Contradictive policies; unclear mandates and responsibilities;
unclear regulatory framework
Rigid bureaucracy, extortion
Competitive wholesale market
No or limited fossil fuel subsidies
Scale up public and private
investment
High performance of institutions
Clear policies and regulations;
Renewable Energy Act;
more attention on legal aspect of a
policy; legal certainty
Simplified licence and non-licence
procedures
Good governance practices of PLN; separate business entity from PLN for
renewable energy businesses
Subsidy restructuring; feed-in-tariff for renewable energy
Fiscal incentives; renewable energy funds; temporary subsidies for RE;
feed-in-tariffs; clear and transparent electricity purchase price
Improvement of quality of services; capacity development
Commitment from all level of government
Streamline bureaucracy, transparency

“10 GW crash programs”, some renewable energy projects may have
technical problems such as the performance of contractors and devel-
opers and road access [83].
To further complicate the issue, energy inequality is seen in
Indonesia. Therefore, the private sector needs to be encouraged to in-
vest in geographically remote areas [10]; [12]. In addition, community-
based initiatives, such as the IBEKA scheme- a community empower-
ment movement beyond supplying electricity from micro-hydro- [84],
should be replicated in hard to reach areas [8].
Ideally, good governance practices and clear long-term policy and
regulations should help to attract private sector investment. However,
even with support from the government, investment in renewable en-
ergy in Indonesia is still considered a risky investment due to the lack of
infrastructure to support the industry. Private investors in RES elec-
tricity expect a clear pricing signal that guarantees a return on invest-
ment. Specifically for geothermal, the exploration phase entails high
risks that should be compensated in the electricity price. Drilling wells
in the exploration phase requires high capital, and often the actual
energy production is lower than the initial estimates.
Private sector participation is important, as shown by experience in
countries that are considered successful in reforming their electricity
sector, including Germany and the UK. However, although Norway did
not implement large-scale privatisation, its market reform was con-
sidered a success [85]. Effective public policies are the main driver of
private finance flows into renewable energy projects [111]. However,
the type of policy instruments has different impacts. Pricing instru-
ments such as FITs are proven to increase investment volume, while
quantity-based instruments fail to do so because pricing policy gives
more tangible incentives [111]. Electricity sector reform requires a
change in policy and regulations, whereas private sector participation is
an important aspect of ideal reform structure [85]. Well-designed pri-
cing and taxation policies also offer potential in achieving effective
emissions reduction. However, these types of policies are prone to
power struggles, with negative economic and social impacts [81,86].
The network of stakeholders as shown in Figs. 12 and 13 [10] does
not necessarily imply that private sector actors are included in decision-
making processes. On the other hand, PLN, as a state-owned enterprise
(SOE) does not have a direct coordination line under the Ministry of
Energy. There has been a case when PLN issued a regulation that was
not in line with the Ministry of Energy's policy. In 2015, the Ministry
issued a ministerial decree that regulates the electricity tariff of micro-
hydropower plants; a year later PLN issued a decree with lower tariff for
micro-hydropower plants [87,88]. Conflicts like this have hampered the
progress of the electricity infrastructure development.
Government institutions need to engage better with both private
and SOEs to unlock the potential of private finance and eventually
connect private sector business investment with the national programs
[89]. The government needs to understand private sector interests and
provide practical options to reflect priorities with respect to investment
opportunities.

Fig. 11. Installed capacity 2014 compared to total target by 2025 [74].

241
M. Maulidia et al.
Fig. 12. Main governmental stakeholders in the electricity sector, modified from Schmidt et al. [10].
Looking at the unrealistic targets that Indonesia has to attain in less
than 8 years, the MRET seem to be an aspirational, rather than an
implementable, target. The targets require policies, regulations and
incentives that will encourage extensive investment in the electricity
infrastructure. In the near and medium term, incentives such as FITs
may attract private sector investment. Beyond the MRET, in the long
term, renewable energy is the key to a transformation towards a green
economy. This will require far-reaching structural, regulatory and
market reform.
Climate finance is crucial for both climate change mitigation and
adaptation efforts. Large mitigation infrastructure projects, such as re-
newable energy projects require larger-scale funding. The private sector
is needed to bridge the gap provided by public sector funding. The
private sector's capacity in finance and technology typically exceeds
those of the public sector. Public climate finance may be directed to
provide incentives for and compliment private investments [90]. The
private sector should be engaged in the financing incentives design and
implementation [91]. The private sector can benefit from climate fi-
nance through increasing project revenue (such as FITs), reducing
242
Renewable and Sustainable Energy Reviews 101 (2019) 231–247
project costs or mitigating investment risk. These approaches play an
important role in enhancing the investment viability of projects [92].
There are numerous challenges in using these sorts of climate finance
mechanisms in countries like Indonesia. Firstly, there is a gap between
financial needs and the availability of domestic and international
funding support [93]. Secondly, from the perspective of the recipient,
there seems to be a lack of readiness to access finance both at gov-
ernment and company level. Recipient countries like Indonesia need to
improve its governance systems, data management and transparency,
streamline country program and pipelines, and step up the private
sector's involvement [93]. At the company level, large private compa-
nies and state-owned enterprises in Indonesia, despite their high po-
tential to invest in mitigation projects need to resolve barriers such as
political and power issues, weak governance and capacity [94] Inter-
national donors make decisions to give aid in response to recipient
countries’ needs on sustainable energy. The transition to sustainable
energy will depend a lot on private sector's capacity in technology,
hence providing incentives to involve the private sector are highly
encouraged [23].
M. Maulidia et al.
Fig. 13. Main industrial stakeholders in the electricity production value chain, modified from Schmidt et al. [10].
4.2. Structural, regulatory and market reform
The pattern of power sector governance across Asian nations shows
strong state control and regulation [95]. This, however, does not ne-
cessarily lead to desirable outcomes. Indonesia's performance in the
electricity sector is poor when compared with other countries in the
region. Indonesia's electrification ratio is one of the lowest compared to
other ASEAN countries such as Malaysia, Philippines, Thailand and
Vietnam which already achieve ratio over 99%. It has improved from
86% in 2015 to 94.9% in 2017 [11]. Moreover, inequality across pro-
vinces in the country is stark, with Papua and other eastern provinces
having a much lower electrification ratio. Only 50% of Papua's [96],
53% of Eastern Nusa Tenggara's and between 66% and 69% of the
population of two other eastern provinces [97], are connected to
electricity.
Although a reform is necessary, restructuring of Indonesia's elec-
tricity sector is bound by laws and regulations. Also, specifically for
Indonesia, liberalisation may be viewed as contrary to the Constitution.
The monopoly status of the electricity, oil, gas and other natural re-
source sectors stems from the rigid interpretation of Article 33 of the
Constitution, which requires economic sectors crucial to the welfare of
243
Renewable and Sustainable Energy Reviews 101 (2019) 231–247
the people to be controlled by the state. Two years after it came into
force, the Law on Electricity No. 20/2002, which allowed increased
involvement of the private sector, was annulled by the Supreme Court
for violating the Constitution.
The current Law on Electricity No. 30/2009 emphasises the role of
the state in controlling the business of power supply. This law also
regulates the consumer power tariffs that the government sets peri-
odically upon consent of the Parliament. The consumer tariffs vary,
depending on geographical regions, and are set to achieve a balance of
interests among government, regions, power producers, and consumers.
Philosophically, the tariffs are controlled to ensure that even the
poorest of Indonesia's society can have access to electricity. After the
subsidy reform in 2014, which sets gasoline price following the fluc-
tuating gas oil market price (Mean of Platts Singapore, MOPS), elec-
tricity is the only commodity whose price is still determined by the
government.
PLN holding the vertically integrated monopoly receives high gov-
ernment subsidies to compensate for losses from government-regulated
low tariffs. Regulated tariffs in general are inefficient and discourage
new investment [95]. Currently, PLN and its subsidiaries control
around 79% of power generation in the country [98]. Independent
M. Maulidia et al.
Power Producers (IPPs) own the remaining 21% of electricity-gen-
erating facilities. PLN has an onerous task, mandated by Law, to pro-
vide electricity infrastructure to the whole country, including nego-
tiating tariffs with IPPs.
Under the new Law No. 30/2009 on Electricity, the push for elec-
tricity sector restructuring remains. A number of civil society organi-
sations and the PLN labour union opposed the controversial article
regarding the unbundling of PLN, which includes vertical unbundling,
(involving integrated business entities that may have different tariffs
from PLN) and horizontal unbundling (splitting business entities based
on geographical regions).
Market liberalisation and electricity reform can be viewed through
the lens of five different aspects: privatisation, wholesale competition,
retail competition, unbundling, and the introduction of independent
regulation Thomas 2006 in [99]. Since the 1990s, privatisation – in the
sense of inviting independent power producers (IPPs) to build power
plants – has occurred in many South-East Asian countries, including
Indonesia. However, after almost 18 years since the first Independent
Power Purchase Agreement (PPAs or IPPAs) was signed, only a few
private companies have actually built power plants and supplied elec-
tricity into the grid in Indonesia. Indonesia may want to learn from its
neighbours. Malaysia is the region's pioneer, with extensive privatisa-
tion of its electricity business. It began poorly, involving ‘crony capit-
alism’ [100] in the 1990s, but later tariff renegotiation and competition
resulted in lower costs and efficient productivity [101].
Another form of privatisation – that of privatisation of state-owned
utilities – does not typically occur in South-East Asian countries. This
type of privatisation has been facing strong opposition from labour
unions and civil society, so the electricity sector is still owned and
controlled by the state. Although in theory competition is good for
consumers, different cases show different outcomes. Only a complete
and thorough privatisation can benefit consumers [95]. For example,
Indonesia's water sector privatisation showed no significant improve-
ment in service quality and access for the poor [102], while tele-
communications privatisation showed completely the opposite [103].
Unlike electricity/ energy or water, telecommunications are not cate-
gorised as sectors of production that are important for the country and
affect the life of the people as legally mandated by Article 33 of the
Indonesian Constitution. The power industry has different character-
istics and therefore whether privatisation in this sector will be bene-
ficial needs to be explored in more detail.
Liberalisation of the electricity sector in other countries has pro-
vided mixed lessons. Some cases have resulted in improved perfor-
mance, but liberalisation has a high potential risk of failure if it is not
done correctly [112]. The most important goal of electricity sector re-
form is to establish institutional arrangements that ensure economic
efficiency for the benefit of consumers [95]. The reform needs to in-
clude all aspects of electricity provision from structure, regulatory and
market liberalisation. Only those reforms that followed the ‘textbook’
model have eventually succeeded in improving service quality and la-
bour productivity [95]. Liberalisation is supposed to help achieve ef-
ficient markets and create a level playing field for renewables because
the increased competition among companies will trigger them to in-
novate and provide the best service for the consumers. A case study in
the UK electricity reform has shown that R&D activities increased
sharply after liberalisation [104].
4.3. Feed-in tariffs for renewables
FITs are a mechanism that could be used to deploy more renewable
energy. FITs are a temporary solution to promote renewable energy in
the currently distorted market that tends to favour fossil-based elec-
tricity. As the volumes of renewables increase, it is expected that they
will get cheaper. Indonesia first implemented FITs when it introduced
small hydropower plants in 2009. Following this, Indonesia developed a
ceiling price scheme for geothermal and FITs for biogas and biomass in
244
Renewable and Sustainable Energy Reviews 101 (2019) 231–247
2012 and for mini-hydro (up to 10 MW) in 2014. The rates of FITs are
different from one region to another, and are determined on the geo-
graphic location and the size of installed capacity [105]. In addition, a
special allocation fund, provided through the Ministry of Energy and
Mineral Resources, is dedicated to build renewable infrastructure in
remote areas – in particular biogas, micro-hydro and solar PV.
The implementation of FITs in Indonesia is not without criticism.
The regulations on FITs seem to lack a degression rate during the pri-
mary years and appear not to take into account inflation and project
profitability. All of these factors potentially deter investors [105]. Fair
estimates on the degression rate based on best practice during the im-
plementation stages would help to improve the scheme [105]. Although
FITs look like a promising solution, different types of renewable energy
would require different mechanisms.
The new energy policy regime has revoked regulations on FITs/
ceiling price and issued a new regulation on the price for RES electricity
[79]. The new regulation based its tariff on the regional and national
main production cost of electricity generation (BPP) as a reference
price. The criticism to this approach is that the BPP only covers gen-
eration cost and excludes transmission and distribution costs, and the
BPP is a figure that changes with time [106]. BPP does not reflect the
LCOE of renewable energy, which is estimated from capital and oper-
ating expenditures, discount rate and electricity sold [107]. Thus, BPP
is not a solid basis for tariff negotiation and increases investment risk.
Moreover, the new regulation does not consider tariff escalation,
something that private investors demand to cover future inflation
[106].
5. Policy implications and the way forward
Indonesia has recognised the importance of renewable energy since
1998, when it included clean energy as a priority in energy policy. It
was translated into a more concrete policy in the national energy policy
in 2006 and was revised in 2014. However, the development of re-
newable energy has been facing many challenges – the biggest being the
large fossil fuel subsidies that distort the market and don’t allow re-
newable energy to compete. With the notion of sharing the export
revenues from the extensive oil resources with the people, fossil fuel
subsidies started in 1969 and continued to be the largest percentage of
central government expenditure until 2014. After Indonesia became a
net oil importer in 2004, no major policy breakthrough has been made
to minimise the budget pressure resulting from fossil fuel imports until
2016 when subsidies reform in transportation sector was introduced
and 2017's electricity subsidies gradual decrease.
The second biggest challenge to achieve the RET is both the
monopolistic nature of Indonesia's electricity landscape that has led to
inefficiencies and little recognition of private sector's role in the
country's policy. The role of PLN in Indonesia's electricity landscape is
dominant but it is not equipped with sufficient financial resources. Its
monopolistic nature as mandated by the Law has created inefficient
markets and has not met the purpose of serving the public's interest.
Discourses to reform PLN vary from reforming PLN's structure while
retaining its monopoly rights to market liberalisation of the country's
electricity market.
While the discussion is mostly focused on the supply side of elec-
tricity, there is a need to create market demand on renewable energy. In
the future, RETs are likely to be achieved if complemented with a
market-based approach such as tradable certificates and other means
where private companies can participate directly.
In general, Indonesia's energy policy is short sighted, does not take
into account the depleting domestic fossil fuel resources and does not
consider long-term benefits such as energy security and environmental
sustainability. Revoking feed-in-tariffs and other pro-renewable energy
regulations and replacing them with regulations that have weak basis to
reflect private sector's investment risks is seen as a policy uncertainty
that deters investment in renewable energy in the country. The
M. Maulidia et al.
misinterpretation of ‘cheap’ energy as instructed by the President
should be corrected by internalising other aspects of energy provision.
Indonesia has homework to depoliticize its energy policy to eliminate
constraints in mobilising private sector funding to build electricity in-
frastructure, hence maximise the benefit for the people.
Indonesia's energy policy should be formulated to correctly tackle
three equally important issues: energy insecurity, energy poverty and
climate change mitigation – also known as the energy trilemma [9].
Failure to balance the three dimensions will worsen the problems the
energy sector is facing [9].
A distorted policy that skews towards only one or two of the three
prongs of the energy trilemma creates injustices to society. The
economy is the dominant aspect in law and policy formulation. The
energy trilemma problem has to be resolved by proposing laws and
policies that are efficient and effective in balancing the three objectives
and thus provide justice to the society [108]. Tackling the energy tri-
lemma can be seen as an opportunity for a reform. Although the mo-
tivation to resolve climate change problem is more external, Indonesia
can reap the benefits of developing renewables to address its domestic
energy problems [109].
The private sector is an important actor that has been neglected in
policy making. To determine the most effective energy policy for
Indonesia, policy research needs to be undertaken through the lens of
the private sector. A policy formulation based on both energy justice
and energy trilemma frameworks will inform decision makers from a
different angle from the currently dominated economic costing.
Energy justice is a conceptual framework that can be used as a tool
to analyse energy policy, energy production, energy consumption, en-
ergy security, climate change, energy trilemma and political economy
of energy [113]. This concept looks at injustices in energy planning, a
complement of the business as usual approach using economic costing
as the dominant consideration in decision making [108]. Energy justice
recognises the dominance of economists in policy formulation. It has a
narrow focus on new energy infrastructure development and therefore
is easy to quantify and feeds directly into economic models [108].
Achieving renewable energy targets can be seen as a starting point
to accomplish Indonesia's electricity sector reform. For long-term con-
sideration, energy justice provides a broader socio-economic frame-
work to analyse types of energy regulation that is suitable for Indonesia,
monopoly or competition taking into account both universal economic
concept and local context. In addition, different types of energy tech-
nologies including renewable energy and non-renewable energy can be
assessed by looking at the role of certain social groups and the eco-
nomic and environmental costs of certain policies and technologies
[110]. This approach provides more in-depth arguments for policy
considerations rather than merely using popular jargons such as ‘cheap’
or ‘affordable’ as the basis of designing important energy policy.
An energy justice framework can be used to analyse the energy
trilemma in Indonesia, aimed at balancing energy security, energy
poverty and climate change mitigation. This concept is new and still
evolving and requires further research. The existing literature focuses
on policy interventions either to achieve energy security or to shift to
renewable energy. Particularly in the context of developing countries,
there is limited literature about the energy trilemma, especially the
tools used to analyse and understand the interactions of the trilemma
dimensions and to solve the problem.
This paper is a part of a broader research project examining the role
of the private sector in achieving Indonesia's renewable energy targets.
It serves as the literature review and provides the context of the re-
search. It also serves to understand the genesis of renewable energy
targets in Indonesia and to predict future trends that can be used as the
basis of policy and decision-making in renewable energy; as well as to
critically evaluate the feasibility of the targets and the measures re-
quired to achieve them. The paper uses a qualitative method which
does involve some subjectivity. Generalisation of the results may be
challenging due to the diversity of issues and stakeholders in Indonesia.
245
Renewable and Sustainable Energy Reviews 101 (2019) 231–247
Policy analysis in this area is always challenging to capture the com-
plexity of policy options. The changes in Indonesia's policy landscape
have been rapid, and the authors have tried to capture the dynamics
and revised the paper accordingly. Future research will include acces-
sing more experts’ views through one on one interviews. These experts,
where possible, have different stances regarding renewable energy
policy in Indonesia to give a wider and more balanced analysis around
the discourses surrounding energy policy in the country.
Acknowledgements
The authors would like to thank Australia Awards for the Ph.D.
scholarship; Peter Storer for language editing; Ministry of Energy and
Mineral Resources for the permission to reuse Fig. 1; Medrilzam from
the Ministry of National Development Planning for the permission to
reuse Figs. 3 and 4; Tobias Schmidt from ETH Zurich for the permission
to modify Figs. 12 and 13; and Aretha Aprilia for editing the revised
version of the paper.
Funding sources
The lead author receives scholarship from the Australia Awards.
This research did not receive any specific grant from funding agencies
in the public, commercial, or not-for profit sectors.
Conflicts of interest
The authors declare no conflict of interest.
References
[1] Government Regulation No 79 Year 2014 on National Energy Policy, 79 C.F.R. § II;
2014.
[2] International Energy Agency. Energy Policies Beyond IEA Countries: Indonesia
2015. Paris: Paris: IEA; 2015.
[3] International Energy Agency. Energy policy review of Indonesia. Paris: Paris:
OECD Publishing; 2008.
[4] Ministry of Energy and Mineral Resources. Concept note of energy resilience fund:
enhancement of renewable energy share in national energy mix and enhancement
of energy conservation aligned with Nawacita and RPJMN. Jakarta,Indonesia:
Ministry of Energy and Mineral Resources; 2016. p. 1–31.
[5] Ministry of Finance. State Budget 2017. In: Mininstry of Finance, editor. Jakarta,
Indonesia; 2017.
[6] World Bank. Indonesia: Data. [cited 5 February 2018]; September 2017. Available
from: 〈https://data.worldbank.org/country/indonesia〉.
[7] Statista. Indonesia: Share of economic sectors in the gross domestic product (GDP)
from 2006 to 2016; 13 February 2018. Available from: 〈https://www.statista.
com/statistics/319236/share-of-economic-sectors-in-the-gdp-in-indonesia/〉.
[8] Wicaksono A. Energy reform in Indonesia: One year after the new government. UQ
Energy Talk 2015〈https://www.scribd.com/document/380923039/ToSend-
UQEnergyTalk-AgungWicaksono-IndonesiaEnergyReform〉.
[9] Ardiansyah F, Gunningham N, Drahos P. An environmental perspective on energy
development in Indonesia. Chapter in book energy and non-traditional security
(NTS) in Asia. In: Caballero-Anthony M, Chang Y, Putra NA, editors. Energy and
non-traditional security (NTS) in Asia. Berlin, Heidelberg: Springer Berlin
Heidelberg, Berlin, Heidelberg; 2012. p. 121.
[10] Schmidt TS, Blum NU, Sryantoro Wakeling R. Attracting private investments into
rural electrification — a case study on renewable energy based village grids in
Indonesia. Energy Sustain Dev 2013;17(6):581–95. https://doi.org/10.1016/j.esd.
2013.10.001.
[11] Ministry of Energy and Mineral Resources. Rasio elektrifikasi 2017 melebih target.
(2017 electrification ratio exceeds target). [cited 8 February 2018]; 10 January
2018. Available from: 〈https://www.esdm.go.id/id/berita-unit/direktorat-
jenderal-ketenagalistrikan/rasio-elektrifikasi-2017-melebihi-target〉.
[12] Williams NJ, Jaramillo P, Taneja J, Ustun TS. Enabling private sector investment
in microgrid-based rural electrification in developing countries: a review. Renew
Sustain Energy Rev 2015;52:1268–81. https://doi.org/10.1016/j.rser.2015.07.
153.
[13] Ministry of Energy and Mineral Resources. Minerba dalam angka 2011–2016.
English translation: Mineral and coal in figures 2011–2016. Jakarta, Indonesia:
Directorate General of Mineral and Coal; 2017. p. 3.
[14] BP. BP Statistical Review of World Energy 2017. [cited 4 June 2018]; June 2017.
Available from: 〈https://www.bp.com/content/dam/bp/en/corporate/pdf/
energy-economics/statistical-review-2017/bp-statistical-review-of-world-energy-
2017-coal.pdf〉.
[15] Indonesian Coal Mining Association. Indonesian coal productions, exports,
M. Maulidia et al.
domestic sales and average of coal price based on ICPR/HBA Periods: 2009–2015;
2015 [cited 7 May 2017].
[16] Indonesia Investments. Indonesia investments: coal. [cited 5 February 2018];
2018. Available from: 〈https://www.indonesia-investments.com/id/bisnis/
komoditas/batu-bara/item236?〉.
[17] Qi Y, Stern N, Wu T, Lu J, Green F. China's post-coal growth. Nat Geosci
2016;9(8):564–6. https://doi.org/10.1038/ngeo2777.
[18] Wu Y, Zhang W. The driving factors behind coal demand in China from 1997 to
2012: an empirical study of input-output structural decomposition analysis.
Energy Policy 2016;95:126–34. https://doi.org/10.1016/j.enpol.2016.05.007.
[19] Garnaut R. Indonesia's resources boom in international perspective: policy di-
lemmas and options for continued strong growth. Bull Indones Econ Stud
2015;51(2):189–212. https://doi.org/10.1080/00074918.2015.1061910.
[20] International Energy Agency Coal information: overview. Statistics; 2017.
〈https://www.iea.org/publications/freepublications/publication/
CoalInformation2017Overview.pdf〉.
[21] Van Noorden R. The knock-on effects of Germany's nuclear phase-out. Nat News
2011. https://doi.org/10.1038/news.2011.348.
[22] Rehner R, McCauley D. Security, justice and the energy crossroads: assessing the
implications of the nuclear phase-out in Germany. Energy Policy 2016;88:289–98.
https://doi.org/10.1016/j.enpol.2015.10.038.
[23] Keppler JH. The economic costs of the nuclear phase-out in Germany. NEA News
2012;30(1):8–14.
[24] Walmsley MRW, Walmsley TG, Atkins MJ. Achieving 33% renewable electricity
generation by 2020 in California. Energy 2015;92:260–9. https://doi.org/10.
1016/j.energy.2015.05.087.
[25] Fauzi A. Revision of geothermal resources classification in Indonesia based on type
of potential power generation. In World Geothermal Congress 2015; 2015.
Melbourne, Australia: International Geothermal Association.
[26] Sukhyar R, Gurusinga CKK, Kasbani K, Widodo S, Munandar A, Dahlan D, et al.
Potential and development of geothermal resources in Indonesia. Jakarta,
Indonesia: Geological Agency, Ministry of Mineral Resources; 2014. p. 1–113.
[27] Sovacool BK, Mukherjee I. Conceptualizing and measuring energy security: a
synthesized approach. Energy 2011;36(8):5343–55. https://doi.org/10.1016/j.
energy.2011.06.043.
[28] Lontoh L, Beaton C, Clarke K. Beaton C, editor. Indonesia energy subsidy review: a
biannual survey of energy subsidy policies. Geneva, Switzerland: Global Subsidies
Initiative and International Institute for Sustainable Development; 2015.
[29] Cheon A, Urpelainen J, Lackner M. Why do governments subsidize gasoline con-
sumption? An empirical analysis of global gasoline prices, 2002–2009. Energy
Policy 2013;56:382–90.
[30] Ministry of Finance. Central Government financial report year 2014. In: Ministry
of Finance, editor. Jakarta, Indonesia: Ministry of Finance; 2015.
[31] Asian Development Bank. Fossil fuel subsidies in Indonesia: trends, impacts, and
reforms. Manila, The Philippines: Asian Development Bank (ADB); 2015.
[32] Durand-Lasserve O, Campagnolo L, Chateau J, Dellink R. Modelling of distribu-
tional impacts of energy subsidy reforms: an Illustration with Indonesia. In:
Durand-Lasserve O, editor. OECD Environment Working Papers; 2015. p. 1–51.
[33] Ministry of Finance. Central Government financial report year 2008. In: Ministry
of Finance, editor. Jakarta, Indonesia: Ministry of Finance; 2009.
[34] Ministry of Finance. Central Government financial report year 2009. In: Ministry
of Finance, editor. Jakarta, Indonesia: Ministry of Finance; 2010.
[35] Ministry of Finance. Central Government financial report year 2010. In: Ministry
of Finance, editor. Jakarta, Indonesia: Ministry of Finance; 2011.
[36] Ministry of Finance. Central Government financial report year 2011. In: Ministry
of Finance, editor. Jakarta, Indonesia: Ministry of Finance; 2012.
[37] Ministry of Finance. Central Government financial report year 2012. In: Ministry
of Finance, editor. Jakarta, Indonesia: Ministry of Finance; 2013.
[38] Ministry of Finance. Central Government financial report year 2013. In: Ministry
of Finance, editor. Jakarta, Indonesia: Ministry of Finance; 2014.
[39] Medrilzam M. Baseline and scenarios model formulation of greenhouse gas emis-
sions reduction policy for Intended Nationally Determined Contributions (INDC)
and the review of National Action Plan on GHG Emissions Reduction (RAN-GRK);
2015. p. 95 〈https://www.scribd.com/document/380925202/Indc-Bappenas-24-
Nov-2015-Final〉.
[40] Mujiyanto S, Tiess G. Secure energy supply in 2025: Indonesia's need for an energy
policy strategy. Energy Policy 2013;61:31–41.
[41] Barnes P. Indonesia, the political economy of energy/Philip Barnes. Oxford; New
York: Oxford; New York: Oxford University Press for the Oxford Institute of Energy
Studies; 1995.
[42] BP. BP Statistical Review of World Energy June 2009. 58th ed. London, UK: BP
PLC; 2009.
[43] BP. BP Statistical Review of World Energy June 2015. 64th ed. London, UK: BP
PLC; 2015.
[44] BP. BP Statistical Review of World Energy June 2016. 65 ed. London, UK: BP PLC;
2016.
[45] Maslanka PJ. Securing Indonesia's energy future. IAGS J Energy Secur 2015.
[46] Haryono E, Khalil S. An examination of government bureaucracy in facilitating
business: comparing Malaysia with Indonesia. J Gov Dev 2011(7):58–71.
[47] Dutu R. Challenges and policies in Indonesia's energy sector. Energy Policy
2016;98:513–9. https://doi.org/10.1016/j.enpol.2016.09.009.
[48] Barber G. Is this the end of the road for Indonesian oil? OilPricecom. UK: OilPrice;
2016.
[49] Government Regulation No. 27 Year 2017 on Amandment to Government
Regulation No. 79 Year 2010 on Recoverable Operational Costs and the Income
Tax Treatment in the Field Upstream of Oil and Gas Businesses, Government
246
Renewable and Sustainable Energy Reviews 101 (2019) 231–247
Regulation No. 27 Year 2017 C.F.R; 2017.
[50] Minister of Energy and Mineral Resources' Decree No. 5889 K/20/MEM/2016 on
Validation of PLN’s Electricity Supply Business Plan Year 2016 until Year 2025,
Minister of Energy and Mineral Resources' Decree No. 5889 K/20/MEM/2016 C.F.
R; 2016.
[51] Cornot-Gandolphe S. Indonesia's electricity demand and the coal sector. Oxford,
UK: The Oxford Institute for Energy Studies; 2017.
[52] Kim H-M, Yoo S-H. Coal consumption and economic growth in Indonesia. Energy
Sources Part B: Econ Plan Policy 2016;11(6):547–52. https://doi.org/10.1080/
15567249.2012.690503.
[53] Indonesian Coal Mining Association, PricewaterhouseCoopers. ICMA and PwC
joint study on Indonesia’s Coal 2016. Jakarta, Indonesia: Indonesian Coal Mining
Association and PricewaterhouseCoopers; 2016.
[54] Government Regulation No. 5 Year 2016 on National Energy Policy Presidential
Regulation No. 5 Year 2006 C.F.R; 2006.
[55] Amir S. Nuclear revival in post-Suharto Indonesia. Asian Surv 2010;50(2):265–86.
https://doi.org/10.1525/as.2010.50.2.265.
[56] National Nuclear Energy Agency Rencana pembangunan PLTN di Indonesia
(Development plan of nuclear power plants in Indonesia). Unknown. 〈http://
www.batan.go.id/index.php/id/infonuklir/nuklir-indonesia-infonuklir/program-
pltn/1810-rencana-pembangunan-pltn-di-indonesia〉.
[57] Ministry of Energy and Mineral Resources. Tarik ulur energi nuklir dalam program
energi nasional. English translation: pull and release of nuclear energy in the na-
tional energy program. [Online News] 15/05/2018 [cited 2018 15/05/2018];
2016. Available from: 〈http://ebtke.esdm.go.id/post/2016/09/14/1340/tarik.
ulur.energi.nuklir.dalam.program.energi.nasional〉.
[58] Siahaan M. Anggota DPR: Tidak seharusnya nuklir pilihan terakhir (Member of
parliament: Nuclear should not be the last option); 2016. 〈https://www.
antaranews.com/berita/568888/anggota-dpr-tidak-seharusnya-nuklir-pilihan-
terakhir〉.
[59] National Nuclear Energy Agency Komisi VII DPR-RI sepakat PLTN segera dibangun
(Commission VII of the Indonesian Parliament agrees nuclear power plant to be
built immediately); 2016. 〈http://www.batan.go.id/index.php/id/kedeputian/
manajemen/hhk/2657-komisi-vii-dpr-ri-sepakat-pltn-segera-dibangun〉.
[60] Effendi BS. Tujuh alasan mengapa Indonesia butuh PLTN dalam konteks pertum-
buhan ekonomi (7 reasons why Indonesia needs nuclear power facilities in the
context of economic growth). Focus group discussion on nuclear - ministry of re-
search, technology and higher education and national committee of economy and
industry. Bali, Indonesia: KEIN and BPPT; 2018.
[61] Ministry of Energy and Mineral Resources. Roadmap Pengembangan Energi Nuklir
2018–2025. English Translation: Roadmap of nuclear energy development
2018–2025. In: Directorate general of new andr enewable energy, editor. Jakarta,
Indonesia: MEMR; 2017. p. 78.
[62] Nuclear Energy Regulatory Agency of Indonesia DPR kembali tegaskan du-
kungannya terhadap pentingnya pembangunan PLTN (The parliament reaffirms its
support to the importance of developing nuclear power facilities); 2018.
[63] Amir S. Roundtable: Needed: The ability to manage nuclear power: When the
public doesn't trust you. Bulletin of the Atomic Scientists; 2014. 〈https://
thebulletin.org/needed-ability-manage-nuclear-power/when-public-doesn%E2%
80%99t-trust-you〉.
[64] Desker B. Southeast Asia going nuclear; 2013. 〈https://www.asianscientist.com/
2013/12/features/southeast-asia-going-nuclear-2013/〉;.
[65] Amir S, Roundtable: needed: ability to manage nuclear power: the transnational
dimensions of nuclear risk. Bulletin of the Atomic Scientists, 2014. 〈https://
thebulletin.org/needed-ability-manage-nuclear-power/transnational-dimensions-
nuclear-risk〉.
[66] Bappenas. Policy Paper: Keselarasan Kebijakan Energi Nasional (KEN) dengan
Rencana Umum Energi Nasional (RUEN) dan Rencana Umum Energi Daerah
(RUED). Laporan Akhir. (Policy Paper: Harmonising Nasional Energy Policy with
General Plan on National Energy and General Plan on Regional Energy. Final
Report). Indonesia; 2012.
[67] Law on Geothermal No. 21 Year 2014, Law No. 21 Year 2014 C.F.R; 2014.
[68] MEMR. Sharpening the focus towards energy sovereignty. final report of the task
force for accelerating the development of new renewable energy and energy
conservation. In: (P2EBT) TFfAtDoNREaEC, editor. Jakarta, Indonesia: MEMR;
2016. p. 69.
[69] The Jakarta Post. Renewable energy regulation repeats old mistakes: Association.
The Jakarta Post. Jakarta, Indonesia: The Jakarta Post; 2017.
[70] Agustinus M. Luhut bubarkan unit-unit khusus di kementerian ESDM (Luhut dis-
solves special units at the ministry of energy and mineral resources). Detikcom;
2016. 〈https://finance.detik.com/energi/d-3282274/luhut-bubarkan-unit-unit-
khusus-di-kementerian-esdm〉.
[71] Government of Indonesia. First Nationally Determined Contribution Republic of
Indonesia. Jakarta, Indonesia: Government of Indonesia; 2016.
[72] Tumiwa F, Imelda H. A brief analysis of Indonesia’s Intended Nationally
Determined Contribution (INDC). Jakarta, Indonesia: Institute for Essential
Services Reform (IESR); 2015.
[73] Institute of Essential Services Reform. Measuring the achievement of NDC in en-
ergy sector. Jakarta, Indonesia: Institute for Essential Services Reform; 2017.
[74] Presidential Regulation No. 22 Year 2017 on General Plan on National Energy
(RUEN), Presidential Regulation No 22 Year 2017 C.F.R; 2017.
[75] Ministry of Energy and Mineral Resources. Capaian bauran energi primer dan
prognosa capaian 2017 (Achievement of primary energy mix and prognosis of
2017 achievement). Jakarta, Indonesia: Ministry of Energy and Mineral Resources;
2017.
[76] Ministry of Energy and Mineral Resources Indonesia Peringkat 2 Produsen Listrik
M. Maulidia et al.
Panas Bumi Lampaui Filipina (Indonesia is the world's second biggest geothermal pro-
ducer, overtaking the Philippines); 2018. DOI: http://ebtke.esdm.go.id/post/2018/
04/28/1948/indonesia.peringkat.2.produsen.listrik.panas.bumi.
[77] Law on Energy No. 30 Year 2007, Law No. 30 Year 2007 C.F.R; 2007.
[78] Presidential Regulation No. 4 Year 2016 on Acceleration of Electricity
Infrastructure Development, Presidential Regulation No. 4 Year 2016 C.F.R; 2016.
[79] Minister of Energy and Mineral Resources' Regulation No. 50 Year 2017 on
Utilisation of Renewable Energy Resources for Electricity Provision, Minister of
Energy and Mineral Resources' Regulation No. 50 Year 2017 C.F.R; 2017.
[80] Dhani S. Assessing the current Indonesia's electricity market arrangements and the
opportunities to reform. Int J Renew Energy Dev 2014;3(1):55–64. https://doi.
org/10.14710/ijred.3.1.55-64.
[81] Jakob M, Steckel JC, Klasen S, Lay J, Grunewald N, Martínez-Zarzoso I, et al.
Feasible mitigation actions in developing countries. Nat Clim Change
2014;4(11):961. https://doi.org/10.1038/nclimate2370.
[82] Marquardt J. A struggle of multi-level governance: promoting renewable energy in
Indonesia. Energy Procedia 2014;58:87–94.
[83] Maulidia M, Halimanjaya A. The coordination of climate finance in Indonesia.
London, UK: The Overseas Development Institute (ODI); 2014. p. 1–11.
[84] Tumiwa F, Rambitan I. Scaling-up renewable energy investments: lessons from the
best practice models in Indonesia. Jakarta, Indonesia: Institute for Essential
Services Reform (IESR); 2013. p. 26.
[85] Erdogdu E. Investment, security of supply and sustainability in the aftermath of
three decades of power sector reform. Renew Sustain Energy Rev 2014;31:1–8.
https://doi.org/10.1016/j.rser.2013.11.014.
[86] Jakob M, Chen C, Fuss S, Marxen A, Edenhofer O. Development incentives for
fossil fuel subsidy reform. Nat Clim Change 2015;5(8):709. https://doi.org/10.
1038/nclimate2679.
[87] Agustinus M, Idris M. Ini alasan PLN enggan beli listrik dari mikrohidro (The
reasons why PLN is reluctant to buy electricity from micro-hydro). Detikcom;
2015. 〈https://finance.detik.com/energi/d-3080097/ini-alasan-pln-enggan-beli-
listrik-dari-mikro-hidro〉.
[88] Agustinus M. Sudah 5 kali Menteri ESDM tegur keras PLN (The minister of Energy
warns PLN for 5 times). Detikcom; 2016. 〈https://finance.detik.com/energi/d-
3259177/sudah-5-kali-menteri-esdm-tegur-keras-pln〉.
[89] Nakhooda S, Jha V. Getting it together: institutional arrangements for coordina-
tion and stakeholder engagement in climate finance. London, UK: The Overseas
Development Institute; 2014. p. 24.
[90] Nakhooda S. The effectiveness of international climate finance. London, UK:
Overseas Development Institute; 2013. p. 22.
[91] Sadie C. Financial incentives to enable clean energy deployment: policy overview
and good practices. Washington, D.C, USA: National Renewable Energy
Laboratory; 2016.
[92] Falconer A, Stadelmann M. What is climate finance? Definitions to improve
tracking and scale up climate finance - CPI Brief. Venice, Italy: Climate Policy
Initiative; 2014.
[93] Government of Indonesia. Third national communication under the united nations
framework convention on climate change. Jakarta, Indonesia: The Ministry of
Environment and Forestry; 2017.
[94] Halimanjaya A. Climate mitigation finance in leveraging private investments in
247
Renewable and Sustainable Energy Reviews 101 (2019) 231–247
Indonesia. J Sustain Financ Investig 2017;7(4):335–59. https://doi.org/10.1080/
20430795.2017.1318461.
[95] Joskow PL. Lessons learned from electricity market liberalization. Energy J
2008;29:9–41.
[96] Suprateka IM. PLN speeds up electrification of villages in Papua and West
Papua; 2017. 〈http://www.pln.co.id/media/siaran-pers/2017/12/pln-kebut-
listriki-desa-desa-di-papua-papua-barat〉.
[97] KataData. Papua’s electrification ratio is still below 50%. Katadata News and
Research. Jakarta, Indonesia: Katadata News and Research; 2017.
[98] PricewaterhouseCoopers. Power in Indonesia: investment and taxation guide
November 2016. 4th ed. Jakarta, Indonesia: PricewaterhouseCoopers; 2016.
[99] Nikomborirak D, Manachotphong W. Electricity reform in practice: the case of
Thailand, Malaysia, Indonesia and the Philippines. In: Intergovernmental group of
experts on competition law and policy. Geneva, Switzerland; 2007.
[100] Smith TB. Privatising electric power in Malaysia and Thailand: politics and in-
frastructure development policy. Public Adm Dev 2003;23(3):273–83. https://doi.
org/10.1002/pad.267.
[101] Kok Fong S. Market reforms and the performance of the Malaysian Electricity
Industry: The University of Queensland, School of Economics; 2011.
[102] Bakker K. Trickle Down? Private sector participation and the pro-poor water
supply debate in Jakarta, Indonesia. Geoforum 2007;38(5):855–68. https://doi.
org/10.1016/j.geoforum.2005.11.011.
[103] Santosa SP. The impact of privatization to the performance of the company: the
case of PT. Telkom. Bandung, Indonesia: University of Padjadjaran; 2008.
[104] Jamasb T, Pollitt MG. Electricity sector liberalisation and innovation: an analysis
of the UK's patenting activities. Res Policy 2011;40(2):309–24. https://doi.org/10.
1016/j.respol.2010.10.010.
[105] Bakhtyar B, Sopian K, Zaharim A, Salleh E, Lim CH. Potentials and challenges in
implementing feed-in tariff policy in Indonesia and the Philippines. Potentials
Chall Implement Feed- Tariff Policy Indones Philipp 2013;60:418–23.
[106] Devine L, Sastrawijaya KD, Kurniawan F. Indonesian government puts the squeeze
on renewable energy tariffs. Jakarta and Singapore: Baker McKenzie; 2017. p. 11.
[107] Blum NU, Wakeling RS, Schmidt TS. Rural electrification through village grids –
assessing the cost competitiveness of isolated renewable energy technologies in
Indonesia. Renew Sustain Energy Rev 2013;22:482–96.
[108] Heffron R, McCauley D, Sovacool B. Resolving society's energy trilemma through
the Energy Justice Metric. Energy Policy 2015;87:168–76.
[109] Gunningham N. Managing the energy trilemma: the case of Indonesia.(approach to
energy governance)(Case study). Energy Policy 2013;54:184–93.
[110] Labelle MC. In pursuit of energy justice. Energy Policy 2017;107:615–20. https://
doi.org/10.1016/j.enpol.2017.03.054.
[111] Cárdenas Rodríguez M, Haščič I, Johnstone N, Silva J, Ferey A. Renewable Energy
Policies and Private Sector Investment: Evidence from Financial Microdata. Off. J.
Eur. Assoc. Environ. Resour. Econ. 2015;62(1):163–88. https://doi.org/10.1007/
s10640-014-9820-x.
[112] Joskow PL. Lessons learned from electricity market liberalization. Ener. J.
2008;29:9–41.
[113] Jenkins K, McCauley D, Forman A. Energy justice: A policy approach. Ener. Policy
2017;105:631–4. https://doi.org/10.1016/j.enpol.2017.01.052.