6 When Is Fiscal Decentralization Good For Governance

OXFORD J.
OURNALS
O~F,ORD
I -
UNIVERSfl"Y PRESS
When Is Fiscal Decentralization Good for Governance?

Author(s): Charles R. Hankla
Source: Publius, Vol. 39, No. 4 (Fall, 2009), pp. 632-650
Published by: Oxford University Press
Stable URL: http://www.jstor.org/stable/40272230
Accessed: 04-01-2017 07:32 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
Oxford University Press is collaborating with JSTOR to digitize, preserve and extend access to Publius
I
This content downloaded from 103.231.241.233 on Wed, 04 Jan 2017 07:32:00 UTC
All use subject to http://about.jstor.org/terms
When is Fiscal Decentralization
Good for Governance?
CharlesR. Hankla*
Many developing countries are seeking to improve governance with fiscal decentralization. It is
therefore worth revisiting what we know about political and economic institutions to understand
how and under what circumstances decentralization can be beneficial. In an effort to do that, I
review past research on the governance implications of devolving power to subnational author-
ities. Based on this review, I find that the gains from decentralization depend sensitively on how
subnational authorities and intergovernmental relations are structured. I therefore conclude the
paper by drawing nine lessons from theory and experience to help improve the design of decen-
tralized institutions.
Even a cursory examination of the world's states reveals a dizzying array of

relationships between central and subnational governments. In some states, the
central government is clearly dominant and subnational authorities are merely
agents of its will. In others, strong regional and local leaders influence the crafting
of virtually all domestic policies. Many systems fall somewhere between these two
extremes, with central governments setting the direction of national policies and
subnational authorities exercising power on the margins. In some states, the
strength of subnational governments 1 lies in their influence over the national policy-
making process, whileOR in others it 2rests in their competence over specific policy
arenas set aside for them by a federal constitution.
This already diverse mixture of systems has been further enriched by a recent
global trend toward decentralization, most strikingly in the developing world.
International Monetary Fund data on twenty-nine countries, for example, show
that the percent of total public expenditure undertaken by regional and local
governments has increased from around 20 percent in the late 1970s to between 30
percent and 40 percent in the mid-1990s (Rodden 2006). There is reason to believe
that this trend is continuing and even accelerating as decentralization gains support
within the international development community. It has therefore become urgent
to understand the impact of decentralization on governance quality, especially in
*Georgia State University; chankla@gsu.edu
Publius:TheJournalof Federalismvolume39 number4, pp. 632-650

doi:10.1093/publius/pjn034
AdvanceAccesspublication6 November2008
© TheAuthor 2008. Publishedby Oxford UniversityPresson behalfof CSFAssociates:Publius,Inc.
All rights reserved.Forpermissions,pleaseemail:journals.permissions@oxfordjournals.org.
Fiscal Decentralization and Governance 633
poor countries where institutional changes bring the greatest potential for good
or ill.
In this article, I draw lessons from theory and experience to shed light on the
consequences of empowering subnational governments, especially in the developing
country context. My focus, unless otherwise indicated, is on the "region,"
"province," or "state" standing at the highest level of subnational government,
although many of the principles discussed will apply to local institutions as well. I
begin by mining current research in economics and political science for insights
into the potential benefits and pitfalls of decentralization. I then attempt to draw
general lessons for designing the most efficient system of intergovernmental
relations. In doing so, I seek the political and fiscal institutions best suited for
maximizing the benefits of decentralization and avoiding the pitfalls, with the goal
of providing the best governance outcomes possible. For the purposes of this
article, decentralization provides quality governance when (i) it1efficiently and
reliably provides citizens with the public goods and services that they desire, (ii) it
2ameliorates (or at least avoids exacerbating) regional inequality and separatism, and
(iii) it3protects macroeconomic stability. Throughout the article, I draw on the
experiences of a variety of states, centralized and decentralized, rich and poor, to
flesh out my broad points with real-world examples.
Is Decentralization Good for Developing Countries?

The State of the Debate
Do the benefits of decentralization for developing countries outweigh the potential
drawbacks? Most of the first-generation literature in economics views
decentralization in a very positive light, emphasizing, among other things, its
potential to improve efficiency (Oates 2005). More recently, a number of studies
have called into question the belief that decentralization is a uniformly good policy
choice. In this section, I will review the current state of the debate over
decentralization's impact on three policy outcomes: (i) the targeting and efficient
provision of public goods, (ii) inequality and regionalism, and (iii) macroeconomic
stability.
Decentralization,Targeting, and the Allocation of PublicGoods

Tiebout (1956) pioneered the idea that decentralization, especially fiscal
decentralization, can improve the efficiency of public goods provision by making
it less uniform. His argument was that, in countries where taxation and spending
powers are delegated to subnational governments, citizens can sort themselves
geographically according to their preferred combination of government services.
Two decades later, Oates (1972) took Tiebout's simple but powerful idea as a
starting point in developing a more comprehensive theory about the optimal level
634 C.R. Hankla
of decentralization (see Lockwood 2002). He identified a trade-off between the

1responsiveness of government to subnational preferences and 2its ability to
internalize externalities and achieve economies of scale. When regional
governments enjoy significant fiscal autonomy, the resulting diversity of tax and
spend policies enhances the efficiency and responsiveness of government. On the
other hand, Oates argued, it is inefficient for governments to provide public goods
whose effects may spillover into adjoining constituencies. Such public goods would
be underprovided because the regional government in question would only
consider the beneficial effects of the good in its own territory when deciding how
much to generate. As a result, Oates argued, there exists for every case an efficient
level of decentralization where the benefits of policy diversity are balanced by the
drawbacks of non-internalization.
A possible implication of Oates's argument is that, in most cases, public services
should be provided at a subnational level. National governments should concern
themselves with macroeconomic stabilization and with equalizing the regions and
encouraging spending that generates positive externalities (Bird et al. 2003). Not all
scholars agree, however, that decentralization is a necessary precondition for the
targeting of public goods. For example, some critics, pointing out that moving
from region to region tends to be costly, argue that citizens are unlikely to sort
themselves geographically by tax and spending preference (Manor 1999; Bardhan
2002). This relative immobility of citizens should be especially true, they contend,
in developing countries.
A second reservation that many scholars have about the Tiebout and Oates models
is that they assume that central governments are unable to target different bundles of
public goods to different regions (Breton 2002; Treisman 2007). If central governments
can in fact accommodate the heterogeneous preferences of their citizens with targeted
fiscal policies, subnational governments become superfluous.
While there is probably some truth to these critiques, a number of points can be
adduced to defend the first-generation theories. First, political impediments may
make it difficult for central governments to tax different regions or groups
differently or to provide more public goods in one place than another. Clever
approaches may be found to do so (such as imposing excise duties only on
products made in regions that prefer more government), but it is unlikely that
central authorities will be as efficient in targeting their fiscal behavior as sub-
national governments. Moreover, central governments, which respond to the
aggregate national preferences of voters, may lack the incentives to provide such
targeted policies. Finally, even if they are willing to vary their taxation and spending,
such governments are unlikely to be the most effective judges of regional preferences.
While it may be true that citizens rarely sort themselves geographically by fiscal policy
preference, subnational governments in a decentralized system can at least provide the
policies preferred by most current residents of their constituency.
1Targeting diverse regions with their preferred taxation and spending policies is
not the only important aspect of quality governance. Perhaps more fundamentally,
it is important that subnational 2governments have an incentive to provide public
over particularistic goods in the first place. Many scholars argue that decentral-
ization encourages such positive incentives by increasing the democratic account-
ability of the state. The basic intuition here is that government is more accountable
when it is closer to the people. First, citizens can better oversee the behavior of
public officials when they live in the same region than when the officials operate
from a distant national capital (see von Braun and Grote 2002). This improved
transparency results in better public service outcomes. Indeed, some studies have
found that political participation tends to improve with decentralization (Manor
1999; Crook 2003). Second, subnational government officials are in a better
position to judge what citizens want if they represent small constituencies.
A number of empirical studies have examined the relationship between
decentralization and governance (and more specifically public service delivery).
Excellent summaries of this literature can be found in Smoke (2006), Treisman
(1999), and Azfar et al. (2004). To take some examples, Lewis (1998) has found
that, in Kenya, local water services are superior to water services administered by
the central government. By contrast, Parry (1997) does not find clear evidence for
improvement in educational services after decentralization in Chile. More broadly,
Davoodi and Zou (1998) conclude that fiscally decentralized states tend to
experience slower economic growth, but Yilmaz (1999), criticizing their use of
subnational expenditure percentages to measure decentralization, finds evidence for
the opposite relationship. Von Braun and Grote (2002) conclude that
decentralization reduces poverty.
There is currently a debate in the literature between those who credit
decentralization with promoting transparency and accountability (i.e. Manor 1999;
Gurgur and Shah 2002; Crook 2003; Huther and Shah 1998) and those who believe
it exacerbates, or at best has no impact on, corruption (i.e. Treisman 2000, 2007;
Tanzi 2002). Experts in the first camp argue that citzens can better supervise
leaders at the regional level. Those in the second contend that more levels of
government provide more opportunities for abuse of office, particularly as
subnational government leaders may be less professional.
A related argument is that decentralization can place more responsibilities in the
hands of regional governments than they can effectively handle (Manor 1999; Shah
2003). Even if subnational authorities have sufficient revenue streams, they may not
possess the administrative resources necessary for effective governance. Many
regional governments lack sufficient numbers of trained civil servants to carry out
their policies, and this problem is especially acute in developing countries.
Finally, some scholars have highlighted the danger that decentralization, far from
enhancing accountability, may simply serve as a vehicle for strengthening the
636 C.R. Hankla
powers of regional elites (Smoke 2006; Azfar et al. 2004; Bardhan and Mookherjee
2001; von Braun and Grote 2002). Such individuals may highjack regional and
especially local governments to enhance their own positions. For subnational
democracy and accountability to exist care must be taken that lower-tier
governments truly represent all of their constituents.
Decentralization, Inequality, and Regionalism

A different strand of the literature, associated most closely with Barry Weingast (for
example, Weingast 1995), argues that decentralization is beneficial because of the
intergovernmental competition that it generates. Because, as Tiebout argued,
citizens, and especially investors, can easily move from one region to another if
they prefer lower taxes or better governance, decentralization puts subnational
authorities in direct competition with one another. In other words, it generates an
incentive to provide the maximum quality of government services at the lowest cost
(see Qian and Roland 1998). Regional leaders who fail to provide such high-quality
governance may find that their tax base shrivels and their economic development
slows. Decentralization also has the important benefit of preventing the central
government from engaging in harmful economic intervention. For all of these
reasons, according to Weingast and others working in this strand of the literature,
decentralization should discourage corruption, promote responsiveness, and check
the growth of government.
Other scholars question this optimistic view by pointing out that, under certain
circumstances, decentralization may weaken the economic and social cohesion of a
nation. If regional authorities fund themselves primarily with their own tax
revenues, it is inevitable that wealthy regions will enjoy more public services than
poorer regions, exacerbating inequality (Manor 1999; Prud'Homme 1995; and see
Aznar et al. 2004). West and Wong (1995), for example, have found that the
availability of education and medical services has grown more unequal in China
since decentralization. Such concerns are less relevant in systems that rely entirely
on central transfers to fund subnational governments.
Furthermore, some scholars have argued that subnational authorities, if not
carefully regulated by the central government, may enact policies that interfere with
the free flow of commerce within a country (Tanzi 2002). The United States
constitution provides the federal government with powers over interstate commerce
to prevent such an eventuality, and central governments around the world typically
work to protect internal markets. Nevertheless, in some decentralized countries
(such as India), regional governments have used their taxation and regulation
powers to interfere with cross-regional trade (Rao 2002).
A final concern is that, in highly divided societies, decentralization may
contribute to the strengthening of regional identities and even, in extreme cases,
the development of separatist movements. Research on the consequences of

decentralization for national cohesiveness is inconclusive and indeed contradictory.
Some scholars fear that decentralization will undermine the unity of diverse
countries (i.e. Treisman 1999; von Braun and Grote 2002), while others make the
opposite point, contending that decentralization can save divided states from
disintegration (i.e. Haysom 2003; Lijphart 1977). A number of states around the
world have relied on decentralization to maintain national cohesiveness, from the
creation of federal systems in Belgium and Switzerland to the recent devolution of
power in the United Kingdom to the current Iraqi constitution. Others, for
example France and the Soviet Union, have used a rigorously centralized state to
promote national unity. The specific impact of decentralization on cohesiveness is
probably only an issue in highly divided societies and is likely to vary on a case-by-
case basis (Crook 2003).
Decentralization and Macroeconomic Stability

A final strand of the literature critiques decentralization for the risks that it may
pose to macroeconomic stability (i.e. Prud'Homme 1995; Tanzi 2002; and see
Smoke 2006 and Wibbels 2000 for reviews of the literature). The danger is that
subnational governments may undermine the macroeconomic environment of
countries by engaging in profligate spending. Bearing out this fear empirically,
Treisman (2000) and Wibbels (2000) have found evidence linking federal systems
with higher inflation.
Other scholars have emphasized that decentralization need not pose significant
challenges for the macroeconomy if central governments either restrict the freedom
of subnational authorities to borrow on private capital markets or credibly commit
to a "no-bailout" policy (see Rodden 2006). If regional governments can borrow
beyond their capacity to pay and then turn to central authorities to finance their
debt, they have no incentive to behave responsibly. With the right mix of policy
tools, however, central governments can incentivize regional leaders to engage in
responsible financial practices.
Nine Lessonsfor the Designof DecentralizedInstitutions

What are we to conclude from these contrasting arguments about decentralization's
impact on governance? If there is an overarching theme in the literature, it is that
the impact of strengthening subnational institutions, whether positive or negative,
depends sensitively on case-specific details. Most scholars would probably agree
that decentralized institutions, when designed correctly, can improve governance
and avoid most of the potential drawbacks that plague poorly structured systems.
The consequences can be dire, however, when decentralization is done poorly. In
this section, I cull from theory and experience a series of nine lessons for designing
638 C. R. Hankla
well-functioning decentralized institutions. I group these lessons into three

categories-those that ensure that 1decentralization is genuine and effective, those
that2promote better subnational representation and governance practices, and those
that 3strengthen national public goods and political stability. Obviously, political
economy is by no means an exact science, but countries that take note of these
lessons are likely to experience smoother intergovernmental relations and better
performance outcomes.
Making Decentralization Genuine and Effective

1. Subnationalgovernmentsshould be delegatedspecificand important responsibilities.
Only when central authorities delegate significant responsibilities to regional
governments can the benefits of decentralization be realized (Manor 1999; Burki
et al. 1999). What sorts of responsibilities should they delegate? There is no
absolute rule here, and the answer will depend on a country's priorities, its
population size, and the administrative capacity of its subnational units. Generally
speaking, however, subnational governments should provide goods that generate
few cross-constituency spillover affects, especially when there may be significant
geographical differences in the quality of goods demanded. For example, the United
States constitution provides Congress with such powers as raising an army,
regulating commerce with foreign nations, setting naturalization policies, and
coining money, activities that would clearly be impossible for subnational
authorities to undertake efficiently. Residual powers are left to the states. In India,
perhaps the developing world's most established democratic federation, an annex to
the constitution assigns certain very specific powers to the center, others to the
states, and still others concurrently. The national list of powers includes everything
from foreign affairs to railways to the regulation of insurance and stock markets.
States, on the other hand, are responsible for public order, policing, fisheries,
agriculture, and many other items. Concurrent powers include education and
economic planning.
Specific powers may also be allocated to the local level under the same principle.
In a comparison of local governments in the established democracies of western
Europe, for example, Norton (1991) found that virtually all governments delegated
responsibilities such as water provision, tourism, roads, building permits, sewage,
trash, and libraries to the local level, and some went on to delegate electricity,
hospitals, and disaster response. It is also possible for certain policy areas to be
shared among different levels of government; in education policy, for example,
national or regional authorities often control curriculum and teacher certification,
and local authorities oversee school buildings. This more cooperative approach can
bring to bear the contrasting strengths of different levels of government on a single
public policy issue.
It is important to note, however, that many observers prefer the enumeration of

clear policy areas that are under the purview of subnational authorities to more
cooperative forms of intergovernmental relations. Clearly separated powers, they
argue, may avoid conflicts between levels of government and improve the ability of
voters to reward and punish politicians for their performance. Countries such as
the United States and India have rules that delineate the separate responsibilities of
national and subnational governments, although the reality is certainly much more
complicated. By contrast, France and Germany provide interesting examples of
countries where intergovernmental responsibilities are much more muddled. In
France, the regions have responsibility for economic development, infrastructure,
and transport, whereas the smaller departments exercise authority over social affairs
and roads, and the communes over planning, building permits, and the like.
However, the precise responsibilities and powers of each level of French
government remain somewhat contested in a number of policy areas, and policy-
making frequently involves multiple levels of authority (Cole 2006). In the more
decentralized case of Germany, states derive most of their authority from their
representation in the upper house of the national parliament, the Bundesrat, rather
than from independent powers over specific policy areas. Policy formation and
implementation (as well as taxation and spending) are shared, cooperative activities
between the center and the states. Rodden (2003) argues that this more cooperative
approach has impeded the ability of voters to impose fiscal discipline on the states
and has contributed to cases of dangerous fiscal indiscipline.
2. Subnational governments should enjoy sufficient expenditure freedom and
administrative capacity to perform their responsibilities.
For reasons already noted, achieving the benefits of decentralization is likely to
depend on the creation of genuinely independent subnational authorities. Along with
significant policy responsibility, the most important characteristic of such institutions
is that they have control over their expenditures. Only then can regional governments
adjust their behavior to the preferences of the citizens they represent. An alternative
approach is to provide subnational governments with an important voice in the
creation of the national budget, generally through an upper legislative chamber.
Of course, even responsive and fiscally independent subnational governments
1
will be unable to function properly if they lack administrative capacity (Manor
1999). A number of ideas have been floated for resolving this common problem,
including an2administrative sharing program with the central government (as in
India and South Africa) or among regional governments themselves (see Momoniat
2002). Private sector expertise can also be brought to bear in an effort to improve
subnational governance (Shah 2003). In any case, the building of strong
administrative structures is vital for decentralization to be effective.
3. Subnational governments should have the ability to tax or should receive steady and
unconditional transfers.
640 C.R. Hankla
Most experts emphasize the benefits of devolving taxation powers to regional

governments. Such an approach will maximize the accountability benefits of
decentralization by allowing subnational authorities to alter both the distribution of
government policies and the size of government itself. Further, efficiency is
maximized when governments raise their revenues from the same citizens who will
enjoy the benefits of the public goods they provide (see von Hagen 2003). In such
circumstances, regional authorities will not have the incentive to overspend that
they would if their funds came from central transfers made up primarily of revenue
generated outside their constituencies (Oates 1972; Hines and Thaler 1995;
Rodden 2002).
What resources should subnational governments tax? Economic theory points to
the importance of assigning regional authorities the right to tax relatively immobile
assets such as real estate, water and sewer services, etc. (von Hagen 2003; Shah
2003). Such a tax policy will reduce the danger that holders of mobile assets
will flee to the constituency with the lowest tax burden, leading to a race to the
bottom.
In the absence of significant regional taxing authority, steady and unconditional
central transfers are necessary to ensure that subnational governments can set their
own budgetary priorities (Manor 1999). There is a large literature on how best to
structure these transfers, but a common choice is to provide regional governments
with the portion of revenues collected in their constituencies (along with an
equalization grant). Whatever the approach, it is important that subnational
governments enjoy the fiscal resources necessary to carry out their policy
responsibilities.
India and Canada provide contrasting examples of sub-national finance. India
is currently struggling with the negative consequences of fiscal imbalance. While
India's states are allocated certain taxation powers, the most remunerative sources
of finance are reserved by the center. As a result, states collect only about 35
percent of public revenues, generating a center-state fiscal imbalance (McCarten
2003). This imbalance is corrected with sizable transfers from the central
government disbursed through the finance commission, the planning commission,
and the central ministries (see Hankla Forthcoming). The finance commission,
which is at least nominally independent, oversees tax sharing programs with the
states. Revenues are disbursed based on objective criteria, but the commission does
not use clear formulae, so some discretion is possible. Revenues from the planning
commission, which fund India's economic development plans, are determined by
bargaining among state and central officials. State officials are represented on a
powerful advisory board called the National Development Council, and there is
evidence that political considerations have entered into disbursement decisions, at
least in the past (see Hankla Forthcoming). Further, the fact that planning
commission funding dries up five years after a project begins poses serious
budgetary problems for a number of states, as they must budget for recurring costs
(McCarten 2003).
By contrast, Canada is characterized by ten powerful provinces which have
virtually complete revenue and borrowing powers (Simeon and Papillon 2006; Bird
and Tassonyi 2003). The provinces have virtually a free hand in imposing taxes,
and indeed they generate fully 83 percent of their financial resources from their
own revenues (Simeon and Papillon 2003). Interestingly, the provincial
governments tap many of the same revenue sources as the center, and they often
rely on federal authorities to collect their taxes (Bird and Tassonyi 2003). Other
sources of provincial finance include large unconditional block grants that Ottawa
distributes to correct the country's vertical fiscal imbalance.
Ensuring that Subnational Institutions are Representative and Efficient

4. Subnational institutions should be democratically elected.
Most of the benefits that accrue from decentralization are related in some way to
the improved accountability and transparency of regional government. Forms of
decentralization that eschew elections (such as a purely administrative reorganiza-
tion of the central government) may achieve modest improvements by locating
decision-makers closer to the citizenry. Without the incentives for responsiveness
that come with election, however, subnational leaders are less likely to provide
citizens with the quality and level of government that they desire (Manor 1999;
Bird and Vaillancourt 1998). Bearing out this argument empirically, Enikolopov
and Zhuravskaya (2007) have found no relationship between administrative
decentralization alone and improved service delivery or reduced corruption.
On a related issue, there is substantial evidence that increased electoral
competitiveness encourages quality governance and especially fiscal responsibility.
The reason is likely that, when elections are more strongly contested, incumbents
must worry more about the consequences at the ballot box of poor governance.
Hallerberg (2004), for example, has linked electoral competitiveness with fiscal
solvency in Western Europe, while Wibbels (2003) has made a similar argument for
US states. In a close examination of Mexico, Hecock (2006) has found that greater
competitiveness in subnational elections was associated with increased education
spending. It should be noted, however, that a stable party system at the subnational
level is also an important component of governance quality. A significant body of
research points to the negative impact of rapid legislative turnover on budget
deficits, arguing that it reduces the time horizons of elected officials (Roubini and
Sachs 1989; Grilli, Masciandaro, and Tabellini 1991; Franzese 2002).
India provides a useful example. In the past, the partisan make-up of
governments at both the state and federal levels in India was highly stable. As a
result, strong chief ministers dominated budgetary decisions at the state level, with
642 C. R. Hankla
legislatures largely quiescent due to strong party discipline. More recent years,
however, have brought a dizzying array of new parties to the fore, resulting in
cabinet instability at all levels of government. This cabinet instability has weakened
chief ministers and strengthened the role of minority parties in expenditure
decisions. There is also reason to believe that this growth in coalitions is increasing
the fiscal profligacy of state governments (McCarten 2003). After all, each
influential party will have supporters that it wants to reward using the public till.
Furthermore, the unstable support bases that plague many of India's key parties
(most notably the governing Congress Party) may discourage further deregulation
of the economy (see Hankla 2006).
5. Subnational political institutions should be designed to encourage the provision of
public goods.
An important factor that separates successful from unsuccessful examples of
decentralization is the quality of institutional design at the subnational level. The
literature provides us with a wide array of lessons for how to structure institutions
to maximize governance quality, most of which derive from research at the national
level. I will briefly summarize those lessons here; for a more in-depth discussion,
please see Hankla (2008).
Most researchers would agree on the value of combining a strong, unified
executive with robust legislative oversight to encourage the provision of public over
particularistic goods. Strong, unified executives, whether directly elected (as in
presidential systems) or chosen by the legislature (as in parliamentary systems), are
responsible to broader constituencies than any single legislator. Furthermore, they
are an easily identifiable focal point of responsibility for the consequences of policy
decisions. As a result, unified executives have a strong incentive to provide public
over particularistic goods and to ensure that government is run efficiently (e.g.
Mukherjee 2003; Blair 1991; Roubini and Sachs 1989; Baldez and Carey 1999;
Hallerberg and Marier 2004). To take an example, in Canada the strong provincial
premiers, who generally control legislative majorities in their provincial capitals, are
probably better able to resist the temptation to overspend than are weaker
executives in other systems. If a province gets embroiled in a debt crisis, voters
can easily blame the governing party and punish it in elections (Bird and
Tassonyi 2003}.
Of course, the importance of legislative oversight of executive power cannot be
overstated, and regional legislatures must possess the physical and administrative
resources necessary to ensure quality government (see, for example, Nijzink et al.
2006). A strong committee system and access to policy expertise and advice are
usually necessary for legislatures to function well.
Some observers recommend non-partisan elections at the regional and especially
local levels as protection from the animosities of national politics. Many municipal
elections in the United States, for example, are like this. Much research in political
science, however, emphasizes the benefits of partisan elections, especially when the
national party system is reproduced at the subnational level. The existence of
parties allows citizens to quickly identify the likely policy positions of any given
candidate and encourages issue driven rather than personality driven politics
(Manor 1999). Furthermore, some scholars have found that when subnational
leaders belong to the same political parties as the national leadership, they may be
less likely to engage in profligate spending and more likely to provide quality
governance (for example, Rodden 2006). The logic is that, if such individuals aspire
to a career in national politics, they must prove themselves competent at the
subnational level. It should be noted, however, that other researchers have begun to
question the link between co-partisanship and fiscal responsibility, arguing that it
reduces salutary fiscal competition among regions (see Khemani 2007).
Protecting National Priorities and Preserving National Unity

6. The central government should use transfers to ensure regional equality.
To promote national cohesion and prevent animosity among regions (in
addition to more normative reasons), central governments should consider using
transfer payments to ensure equality of resources among subnational governments
(Oates 1972; Bird et al. 2003). To take an example, the federal government of
Canada distributes equalization funds so that the revenues of the poorer provinces
are brought up to an average level based on their population size. In addition, a
special arrangement with Quebec has been made to allow that French-speaking
province, which is sensitive about its independence, even more autonomy over its
receipts (Bird and Tassonyi 2003).
7. The central government should use targeted and matching grants to promote core
national objectives and to encourage the provision of public goods with positive
spillovers at the subnational level.
Some observers are justifiably concerned that fiscally independent subnational
authorities could neglect to finance objectives seen as vital to the national
government (Bird et al. 2003). Central governments worried about this problem
frequently mandate that regional and local authorities must provide certain key
services at specific minimum levels. Such is the case, for example, with public
education in the United Kingdom (Davey 1983). Other central governments use
conditional or matching grants to encourage certain behaviors at the regional level
(Shah 2003). Medicaid, a medical insurance program for the poor in the United
States, works in this fashion. Likewise, the government of South Africa has used
matching grants to encourage the provision of public goods by subnational
authorities (Shah 2003). Among the most important uses for conditional and
matching grants is to encourage subnational authorities to provide public goods
whose benefits spillover outside their constituencies. Public goods of this sort might
644 C.R. Hankla
include the construction of hospitals and universities and the subsidization of

research and development. As noted above, such goods will be underprovided if the
central government does not take action.
8. Reformers should design subnational political institutions to maximize national
unity and protect minorities.
By carefully designing subnational political institutions, reformers can go a long
way toward realizing the benefits of decentralization for national unity and
minority protection and avoiding its pitfalls (see Hankla 2008 for a more detailed
treatment). Two institutional choices are particularly important here-electoral
system design and the role of national parties. The impact of electoral system
design on protecting minorities is the subject of a wide-ranging debate between
"consociationalists" and "centripetalists". Consociationalists (e.g. Lijphardt 1977)
advocate list proportional representation for its ability to ensure that all social
interests are represented by parties in government. Centripetalists (e.g. Horowitz
2003), by contrast, prefer the alternative vote and single transferable vote systems
for the incentives they give candidates to make broad appeals across social groups.
By allowing voters to rank-order their preferences, these systems reward moderate
candidates who win the second preferences of voters outside their social group. The
best approach is likely to depend on national context, with proportional
representation generally preferred when (i) society is divided into a small number
of groups, (ii) these group divisions are fixed, and (iii) elites are more moderate
than voters (see, for example, Reilly 2002). There is very little support in the
literature for traditional first-past-the-post electoral systems as used in the United
States, the United Kingdom, and many other countries.
One useful approach to reducing the risk of decentralization for national
political unity is to encourage the replication of the central party system at the
regional level (see Weaver 2002; Filippov et al. 2004). This can be done in two
ways-by holding regional elections at the same time as national elections and by
using the same electoral system at both levels. When national parties also compete
at the regional level, subnational politicians are much less likely to pursue separatist
agendas. The participation of national parties should reduce the chances that
regional parties with particularistic agendas will take power. Moreover, regional
politicians, when they owe their political fortunes to national parties, will need to
follow the national party line if they aspire to higher office in the future. Of course,
when national party leaders are too strong, there is a risk that regional variation
will be suppressed (Yilmaz et al. 2008), but the need to win subnational local
elections should serve as a corrective.
9. The central government should protect macroeconomic stability.
The central government, of course, is ultimately responsible for the stability of
the macroeconomy. Central governments must therefore use the policies under
their control to promote exchange rate stability, control inflation, and ensure a
common internal market. Most of the macroeconomic risks of decentralization,

however, can be overcome if central governments either limit borrowing by
subnational authorities or impose a credible no-bailout policy. It is probably
counterproductive for national leaders to ban all regional borrowing, especially
when the money is intended to finance investment, but controls on total debt are
probably wise (see Shah 2003). An alternative and perhaps superior approach is for
the central government to impose a credible no-bailout policy (Rodden 2006; Bird
et al. 2003). With such a policy, all debt that subnational governments acquire will
be solely their responsibility to pay down. As a result, regional leaders will have an
incentive to borrow only as much as they can repay, and private lenders will be
able to set interest rates according to the credit profiles of the subnational
governments. This approach maximizes the autonomy of subnational authorities
while minimizing the threat of macroeconomic difficulties.
There is much debate about when central governments can credibly follow a
policy of no-bailout, since it may be difficult to take no action to aid a subnational
government embroiled in a debt crisis. The evidence indicates that central
governments will be less able to commit to a no-bailout policy when they fund
subnational authorities with transfers (Rodden 2006).
French and Canadian fiscal relations provide an instructive example of how a
no-bailout policy can work in practice. All three levels of subnational government
in France have the freedom to borrow from the private market, although French
law stipulates that borrowed money should be used only to finance investment.
Furthermore, subnational governments are required to keep their accounts in
annual balance, and may be audited by the center if they fail to do so (Gilbert and
Guengant 2002). As a result, recent decentralization reforms in France have
increased the risk that subnational governments may default on their loans, but
only slightly. Continuing central oversight has helped to mitigate any
macroeconomic problems, and the fiscal performance of regional and local
governments in France is generally good (Gilbert and Guengant 2002).
Like France, Canada imposes few central regulations over provincial borrowing,
relying instead on an informal policy of no-bailout. This system may have contributed
to some indebtedness during the tight 1990s,but provincial finances are overall in very
good shape (Bird and Tassonyi 2003). As a result of the no bailout-policy, private
capital markets are free to adjust their lending practices based on the credit records of
individual provinces, providing an incentive for responsible behavior.
By contrast, efforts to limit subnational borrowing in India have been much less
successful than the no-bailout approach used in France and Canada. State
governments, under pressure to attract foreign direct investment and coping with
the declining importance of central planning transfers, have been more and more
attracted to borrowing as a means for financing their current accounts (McCarten
2003). They have been able to bypass central regulations in a number of ways,
646 C.R. Hankla
including by floating corporations and borrowing through them, borrowing from

public enterprises and special public accounts, and even over-drafting from the
Reserve Bank of India (Rajaraman 2000). Even worse, the central government
(including the finance commission) has a history of bailing-out profligate federal
states, reducing any incentive for fiscal discipline. As a result, several states have begun
to develop serious debt burdens (McCarten 2003). India, therefore, serves as a
cautionary tale for relying too heavily on central regulation to prevent overspending at
the state level. Subnational governments can often find ways around such regulations,
especially when central authorities cannot commit to a credible no-bailout policy.
Some FinalThoughts
More and more, national leaders in the developing world are seeking to improve
governance with decentralization. It is worth revisiting what we know about political
and economic institutions to understand how and under what circumstances
decentralization has been beneficial. This paper has sought to do that by reviewing past
research on the governance implications of devolving power to subnational authorities
and by drawing lessons from theory and experience for designing better decentralized
institutions. The nine lessons presented here, while by no means set in stone, are
intended to synthesize what is known about structuring intergovernmental relations,
with a focus on deriving practical lessons for those policy-makers in the developing
world considering decentralization for their citizens.
Acknowledgments
I would like to thank Jorge Martinez-Vazquez and Rich Engstrom for their
invaluable advice. I have also benefited from financial support from the US Agency
for International Development for work under contract to Georgia State University.
References
Aznar, Omar, Satu Kiihkonen, Anthony Lanyi, Patrick Meagher, and Diana Rutherford.
2004. Decentralization, governance and public services: The impact of institutional
arrangements. In Devolution and development, ed. Mwangi S. Kimenyi, and
Patrick Meagher. Burlington, VT: Ashgate.
Baldez, Lisa, and John M. Carey. 1999. Presidential agenda control and spending policy: Lessons
from General Pinochet's Constitution. American Journalof PoliticalScience43 (1): 29-55.
Bardhan, Pranab. 2002. Decentralization of governance and development. Journal of
EconomicPerspectives16 (4): 185-205.
Bardhan, P., and D. Mookherjee. 2001. Relative capture of local and central governments: An
essay in the political economy of decentralization. In Fiscal decentralization:Promisesand
pitfalls, ed. G. Eskeland, S. Devarajan, and H.F. Zhou. Washington, DC: The World Bank.
Bird, Richard, Bernard Dafflon, Claude Jeanrenaud, and Gebhard Kirchgassner. 2003.
Assignment of responsibilities and fiscal federalism. In Federalism in a changing world:
Learningfrom each other, ed. Raoul Blindenbacker, and Arnold Koller. Ithaca: McGill-
Queens University Press.
Bird, Richard M., and Almos Tassonyi. 2003. Constraining sub-national fiscal behavior in
Canada: Different approaches, similar results? In Fiscaldecentralizationand the challengeof
hard budget constraints,ed. Jonathan Rodden, Gunnar S. Eskeland, and Jennie Litvack.
Cambridge, MA: MIT Press.
Bird, Richard M., and Francois Vaillancourt. 1998. Fiscal decentralization in developing
countries: An overview. In Fiscal decentralization in developing countries, ed. Richard
M. Bird, and Francois Vaillancourt. New York: Cambridge University Press.
Blair, Philip. 1991. Trends in local autonomy and democracy: Reflections from a European
perspective. In Local government in Europe: Trends and developments,ed. Richard Batley,
and Gerry Stoker. London: McMillan Press.
Breton, Albert. 2002. An introduction to decentralization failure. In Managing fiscal
decentralization,ed. Ehtisham Ahmad, and Vito Tanzi. New York: Routledge.
Burki, Javed, Guillermo Perry, and William Dillinger. 1999. Beyond the center:Decentralizing
the State. Washington, DC: The World Bank.
Cole, Alistair. 2006. Decentralization in France: Central steering, capacity building, and
identity construction. FrenchPolitics4: 31-57.
Crook, Richard Charles. 2003. Decentralisation and Good Goverance. In Federalism in a
changing world: Learningfrom each other, ed. Raoul Blindenbacker, and Arnold Koller.
Ithaca: McGill-Queens University Press.
Davey, Kenneth. 1983. Financing regionalgovernment. New York: John Wiley and Sons.
Davoodi, Hamid, and Heng-fu Zou. 1998. Fiscal decentralization and economic growth: A
cross-country study. Journal of Urban Economics43: 244-257.
Enikolopov, Ruben, and Ekaterina Zhuravskaya. 2007. Decentralization and political
institutions. Journal of Public Economics91: 2261-2290.
Filippov, Mikhail, Peter C. Ordeshook, and Olga Shvetsova. 2004. Designingfederalism:A
theory of self-sustainablefederal institutions. New York: Cambridge University Press.
Franzese, Robert. 2002. Macroeconomicpolicies of developed democracies. Ann Arbor:
University of Michigan Press.
Gilbert, Guy, and Alain Guengant. 2002. The public debt of local governments in France. In
Local Public Finance in Europe, ed. Bernard Dafflon. Northampton, MA: Edwand Elgar.
Grilli, Vittorio, Donato Masciandaro, and Guido Tabellini. 1991. Political and monetary
institutions and public financial policies in the industrial democracies. EconomicPolicy6:
341-92.
Gurgur, Tugrul, and Anwar Shah. 2002. Localization and corruption: Panacea or pandora's
box? In Managing Fiscal Decentralization, ed. Ehtisham Ahmad, and Vito Tanzi.
New York: Routledge.
648 C.R. Hankla
Hallerberg, Mark. 2004. Domestic budgetsin a united Europe.Ithaca: Cornell University Press.
Hallerberg, Mark, and Patrik Marier. 2004. Executive authority, the personal vote, and
budget discipline in Latin American and Caribbean countries. American Journal of Political
Science48 (3): 571-587.
Hankla, Charles. 2006. Party linkages and economic policy: An examination of Indira
Gandhi's India. Business and Politics8 (3): 1-29.
---. 2008. Decentralization, governance, and the structure of local political institutions:
Assessing the evidence. Working Paper, Georgia State University.
---. Parties and patronage: Trade and industrial policy in India. Comparative Politics
(forthcoming).
Haysom, Nicholas R. L. 2003. Constitution making and nation building. In Federalismin a
Changing World: Learning From Each Other, ed. Raoul Blindenbacker, and Arnold Koller.
Ithaca: McGill-Queens University Press.
Hecock, R. Douglas. 2006. Electoral conmpetition, globalization, and subnational education
spending in Mexico, 1999-2004. American Journal of Political Science50 (4): 950-961.
Hines, James, and Richard Thaler. 1995. The flypaper effect. Journal of EconomicPerspectives
9: 217-226.
Horowitz, Donald L. 2003. Electoral systems: A primer for decision makers. Journal of
Democracy 14 (4): 115-127.
Huther, Jeff, and Anwar Shah. 1998. Applying a simple measure of good governance to the
debate on fiscal decentralization.Washington, DC: The World Bank.
Khemani, Stuti. 2007. Party politics and fiscal discipline in a federation. ComparativePolitical
Studies 40 (6): 691-712.
Lewis, Blaine. 1998. The impact of public infrastructure on municipal economic
development: Empirical results from Kenya. Review of Urban and Regional Development
Studies 10 (2): 142-456.
Lijphart, Arend. 1977. Democracyin plural societies:A comparativeexplanation.New Haven:
Yale University Press.
Lockwood, Ben. 2002. Distributive politics and the costs of centralization. Review of
Economic Studies 69 (2): 313-337.
Manor, James. 1999. The political economy of democratic decentralization.Washington, DC:
The World Bank.
McCarten, William J. 2003. The challenge of fiscal discipline in the Indian states. In Fiscal
Decentralization and the Challenge of Hard Budget Constraints, ed. Jonathan Rodden,
Gunnar S. Eskeland, and Jennie Litvack. Cambridge, MA: MIT Press.
Momoniat, Ismail. 2002. Fiscal decentralization in South Africa: A practitioner's perspective.
In Managing Fiscal Decentralization,ed. Ehtisham Ahmad, and Vito Tanzi. New York:
Routledge.
Mukherjee, Bumba. 2003. Political parties and the size of government in multiparty
legislatures: Examining cross-country and panel data evidence. Comparative Political
Studies 36 (6): 699-728.
Nijzink, Lia, Shaheen Mozaffar, and Elisabete Azevedo. 2006. Parliaments and the
enhancement of democracy on the African continent: An analysis of institutional capacity
and public perceptions. Journal of Legislative Studies 12 (3-4): 311-335.
Norton, Alan. 1991. Western European local government in comparative perspective. In
Local Government in Europe: Trends and Developments, ed. Richard Batley, and
Gerry Stoker. London: McMillan Press.
Oates, Wallace E. 1972. Fiscal federalism. New York: Harcourt, Brace, and Jovanovich.
---. 2005. Towards a second generation theory of fiscal federalism. International Tax and
Public Finance 12 (4): 349-373.
Parry, Taryn. 1997. Achieving balance in decentralization: A case study of education
decentralization in Chile. World Development 25 (2): 211-225.
Prud'Homme, Remy. 1995. The dangers of decentralization. World Bank Research Observer
10: 201-220.
Qian, Yingyi, and Gerard Roland. 1998. Federalism and the soft budget constraint. American
Economic Review 88 (5): 1143-1162.
Rajaraman, Indira. 2000. Fiscal features of local government in India. In Governance,
decentralization, and reform in China, India, and Russia, ed. Jean-Jacques Dethier. Boston:
Kluwer Academic Publishers.
Rao, M. Govinda. 2002. Fiscal decentralization in Indian federalism. In Managing fiscal
decentralization, ed. Ehtisham Ahmad, and Vito Tanzi. New York: Routledge.
Reilly, Benjamin. 2002. Electoral systems for divided societies. Journal of Democracy 13 (2):
156-170.
Rodden, Jonathan. 2002. The dilemma of fiscal federalism: Grants and fiscal performance
around the world. American Journal of Political Science 46 (3): 670-687.
---. 2003. Soft budget constraints and German federalism. In Fiscal Decentralization and
the Challenge of Hard Budget Constraints, ed. Jonathan Rodden, Gunnar S. Eskeland, and
Jennie Litvack. Cambridge, MA: MIT Press.
---. 2006. Hamilton's paradox: The promise and peril of fiscal federalism. New York:
Cambridge University Press.
Roubini, Nouriel, and Jeffrey Sachs. 1989. Government spending and budget deficits:
Industrial democracies. Working Paper 2919, National Bureau of Economic Research,
Cambridge, MA.
Shah, Anwar. 2003. Fiscal decentralization in transition economies and developing countries.
In Federalism in a Changing World: Learning From Each Other, ed. Raoul Blindenbacker,
and Arnold Koller. Ithaca: McGill-Queens University Press.
650 C.R. Hankla
Simeon, Richard, and Martin Papillon. 2006. Canada. In Distribution of powers and
responsibilities in federal countries, ed. Akhtar Majeed, Ronald L. Watts, and Douglas
M. Brown. Ithaca: McGill-Queens University Press.
Smoke, Paul. 2006. Fiscal decentralization in developing countries. In Public Sector Reform in
Developing Countries, ed. Yusuf Bangura, and George A. Larbi. New York: Palgrave
McMillan.
Tanzi, Vito. 2002. Pitfalls on the road to fiscal decentralization. In Managing fiscal
decentralization, ed. Ehtisham Ahmad, and Vito Tanzi. New York: Routledge.
Tiebout, C. M. 1956. A pure theory of local expenditure. Journal of Political Economy 64:
416-424.
Treisman, Daniel. 1999. Political decentralization and economic reform: A game theoretical
analysis. American Journal of Political Science 43 (2): 488-517.
---. 2000. The causes of corruption: A cross-national study. Journal of Public Economics
76 (3): 399-457.
---. 2007. The architecture of government: Rethinking political decentralization. New York:
Cambridge University Press.
Von Braun, Joachim Von, and Ulrike Grote. 2002. Does decentralization serve the poor? In
Managing fiscal decentralization, ed. Ehtisham Ahmad, and Vito Tanzi. New York:
Routledge.
Von Hagen, Jurgen Von. 2003. Fiscal federalism and political decision structures. In
Federalism in a changing world: Learning from each other, ed. Raoul Blindenbacker, and
Arnold Koller. Ithaca: McGill-Queens University Press.
Weaver, R. Kent. 2002. Electoral rules and governability. Journal of Democracy 13 (2):
111-125.
Weingast, Barry R. 1995. The economic role of political institutions: Market-preserving
federalism and economic growth. Journal of Law, Economics, and Organization 11: 1-31.
West, L., and C. Wong. 1995. Fiscal decentralization and growing regional disparities in rural
China. Oxford Review of Economic Policy 11 (4): 70-84.
Wibbels, Erik. 2000. Federalism and the politics of macroeconomic policy and performance.
American Journal of Political Science 44 (4): 687-702.
---. 2003. Bailouts, Budgets, and Leviathans: Comparative federalism and lessons from
the early United States. Comparative Political Studies 36 (5): 475-508.
Yilmaz, Serdar. 1999. The impact of fiscal decentralization on macroeconomic performance.
Paper prepared for the 92nd Annual Conference on Taxation, Atlanta, Georgia.
Yilmaz, Serdar, Yakup Beris, and Rodrigo Serrano-Berthet. 2008. Local government
discretion and accountability: A diagnostic framework for local governance. Paper No.
113, Social Development Working Papers, Local Governance and Accountability Series,
The World Bank, Washington, DC.

6 When Is Fiscal Decentralization Good For Governance

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

6 When Is Fiscal Decentralization Good For Governance

Uploaded by

Copyright:

Available Formats

OXFORD J.

When Is Fiscal Decentralization Good for Governance?

Even a cursory examination of the world's states reveals a dizzying array of

*Georgia State University; chankla@gsu.edu

Publius:TheJournalof Federalismvolume39 number4, pp. 632-650

Is Decentralization Good for Developing Countries?

Decentralization,Targeting, and the Allocation of PublicGoods

of decentralization (see Lockwood 2002). He identified a trade-off between the

Decentralization, Inequality, and Regionalism

the development of separatist movements. Research on the consequences of

Decentralization and Macroeconomic Stability

Nine Lessonsfor the Designof DecentralizedInstitutions

well-functioning decentralized institutions. I group these lessons into three

Making Decentralization Genuine and Effective

It is important to note, however, that many observers prefer the enumeration of

Most experts emphasize the benefits of devolving taxation powers to regional

Ensuring that Subnational Institutions are Representative and Efficient

Protecting National Priorities and Preserving National Unity

include the construction of hospitals and universities and the subsidization of

common internal market. Most of the macroeconomic risks of decentralization,

including by floating corporations and borrowing through them, borrowing from

You might also like