Natural Resources: A Blessing or Curse? Factors Determining Economic Outcomes

Journal of Economic Literature 2011, 49:2, 366–420
http:www.aeaweb.org/articles.php?doi=10.1257/jel.49.2.366
Natural Resources: Curse or Blessing?

Frederick van der Ploeg*
Are natural resources a “curse” or a “blessing”? The empirical evidence suggests that
either outcome is possible. This paper surveys a variety of hypotheses and supporting
evidence for why some countries benefit and others lose from the presence of natu-
ral resources. These include that a resource bonanza induces appreciation of the real
exchange rate, deindustrialization, and bad growth prospects, and that these adverse
effects are more severe in volatile countries with bad institutions and lack of rule
of law, corruption, presidential democracies, and underdeveloped financial systems.
Another hypothesis is that a resource boom reinforces rent grabbing and civil con-
flict especially if institutions are bad, induces corruption especially in nondemocratic
countries, and keeps in place bad policies. Finally, resource rich developing economies
seem unable to successfully convert their depleting exhaustible resources into other
productive assets. The survey also offers some welfare-based fiscal rules for harness-
ing resource windfalls in developed and developing economies. (JEL O47, Q32, Q33)
1. Introduction such as Botswana or Norway, are more suc-

cessful while others perform badly despite
M any recognize the opportunities natu-

ral resources provide for economic
growth and development and thus the chal-
their immense natural wealth. Is it because
resource booms induce appreciation of the
real exchange rate and makes nonresource
lenge of ensuring that natural resource sectors less competitive (Dutch disease)?
wealth leads to sustained economic growth Are learning by doing and other spill-over
and development. Still, many countries are effects strong enough in those nonresource
cursed by natural resource wealth. The key traded sectors to warrant government inter-
question is why resource rich economies, vention? Or do the riches of a resource
* University of Oxford, VU University of Amsterdam,

Tinbergen Institute, CEPR and CESifo. I am grateful to 2–5 February 2009 and seminars at the Kiel Institute of
the editor Roger Gordon, two anonymous referees, Rabah World Economics, EUI, ISS, The Hague, Erasmus Uni-
Arezki, Maarten Bosker, Erwin Bulte, Paul Collier, David versity Rotterdam, Amsterdam, Tinbergen Institute,
Hendry, Torfinn Harding, Roland Hodler, Mansoob Mur- Tilburg, Pisa, Oxford University, Cambridge University,
shed, Steven Poelhekke, Radek Stefanski, Tony Venables, ETH Zurich, University of Birmingham and University of
David Vines, Klaus Wälde, Cees Withagen, Aart de Zeeuw Nottingham, and from participants in courses at Oxford
and participants of the CESifo Area Conference on Pub- University, the Tinbergen Institute and the International
lic Sector Economics, 21–23 April 2006, Munich, the 6th Monetary Fund for many helpful comments and sugges-
Annual Meeting of the EEFS, Sofia, 2007, the 10th Anni- tions. This work was supported by the BP funded Oxford
versary Conference of the Global Development Network, Centre for Analysis of Research Rich Economies.
366
van der Ploeg: Natural Resources 367
bonanza induce a shift from profit-making 2. Stylized Facts: Is the Natural Resource
entrepreneurship toward socially inefficient Curse Inevitable?
rent seeking? How much of this depends
on the quality of institutions, the rule of law Although some resource rich countries
and the degree of financial development? Is benefit from their natural wealth, others are
resource wealth plundered by corruption, in a terrible state. I discuss some well-known
rent grabbing, and civil war at the expense of examples of countries whose dependence on
widespread inequality and poverty? Does a natural resources have gone together with
resource boom maintain unsustainable, bad bad macroeconomic performance and grow-
policies for too long? Is depleting natural ing inequality among its citizens and contrast
wealth sufficiently reinvested in other pro- these with others that have benefited from
ductive assets? their natural resource wealth (section 2.1).
To shed light on these important ques- I also discuss historical evidence on how
tions, I first present the relevant stylized natural resources have led to establishment
facts (case studies, historical and statisti- of property rights and contributed to eco-
cal) on the heterogeneous experiences of nomic development (section 2.2). I then dis-
resource rich economies in section 2. I then cuss some cross-country stylized facts on the
put forward, in section 3, eight hypoth- effects of resources on economic and social
eses and offer supporting theory and the outcomes (section 2.3). Finally, I discuss sav-
best cross-country, panel-data, and quasi- ing statistics to see to what extent natural
experimental evidence that is available on resource wealth is converted into physical,
each hypothesis. What transpires is not only human, and other wealth (section 2.4). The
how much the experiences of resource rich main point of these stylized facts is to point
economies differ from other economies but out the enormous variety of experiences of
also the wide variety of experiences of dif- resource rich countries and the puzzles that
ferent resource rich economies. In section they suggest. I leave theories for the effects
4, I give detailed attention to the question of resources on growth and conflict and the
why so many resource rich developing econ- testing thereof to section 3.
omies deviate from the so-called Hartwick
2.1 Diverse Experiences of Illustrative
rule and do not fully reinvest their resource
Resource Rich Countries
rents in foreign assets or productive capi-
tal (e.g., buildings, roads, machines, human Accounts of the resource curse are avail-
capital, or health) even though saving is an able for many countries (e.g., Alan Gelb
essential part of economic development. 1988; Terry Lynn Karl 1997, 1999; Adrian J.
The puzzle is why observed and optimal B. Wood 1999; Richard M. Auty 2001b). The
saving rates do not seem to differ much in most dramatic example is perhaps Nigeria
nonresource economies but differ sharply (David Bevan, Paul Collier, and Jan Willem
in resource rich economies. I put forward Gunning 1999; Xavier Sala-i-Martin and
the “anticipation of better times” and the Arvind Subramanian 2003). Oil revenues
“voracious rent seeking” hypotheses to per capita in Nigeria increased from US$33
help explain this puzzle. Section 5 offers in 1965 to US$325 in 2000, but income per
welfare-based fiscal rules for harnessing capita has stagnated at around US$1,100 in
resource windfalls in developing economies PPP terms since its independence in 1960
paying special attention to capital scarcity, putting Nigeria among the fifteen poorest
absorption problems, and volatile revenue countries in the world. Between 1970 and
streams. Section 6 concludes. 2000, the part of the population that has to
368 Journal of Economic Literature, Vol. XLIX (June 2011)
survive on less than US$1 per day shot up widespread economic spill-over effects, and
from 26 to almost 70 percent. In 1970, the the financial opportunities that coca provided
top 2 percent had the same share of income have fueled violence and civilian conflict
as the bottom 17 percent but, in 2000, the especially outside the major cities (Joshua
same share as the bottom 55 percent. Clearly, D. Angrist and Adriana D. Kugler 2008).
huge oil exports have not benefited the aver- Greenland benefits from a large annual grant
age Nigerian. Although Nigeria has experi- from Denmark to ensure a similar GDP per
enced rapid growth of physical capital at 6.7 capita to the Danish one. As a result, it has
percent per year since independence, it has suffered from an appreciated real exchange
suffered a declining TFP of 1.2 percent per rate as well as rent seeking from a compre-
year. Capacity utilization in manufacturing hensive system of state firms and price regu-
hovers around a third. Two thirds of capacity, lations (Martin Paldam 1997).
often owned by the government, thus goes to Others discuss more positive experiences.
waste. Successive military dictatorships have Forty percent of Botswana’s GDP stems from
plundered oil wealth and Nigeria is known diamonds, but Botswana has managed to beat
for its anecdotes about transfers of large the resource curse. It has the second highest
amounts of undisclosed wealth. Oil wealth public expenditure on education as a fraction
has fundamentally altered politics and gov- of GNP, enjoys the world’s highest growth
ernance in Nigeria. It is hard to maintain that rate since 1965, and its GDP per capita is at
the standard Dutch disease story of worsen- least ten times that of Nigeria (Maria Sarraf
ing competitiveness of the non-oil-export and Moortaza Jiwanji 2001). The Botswana
sector fully explains its miserable economic experience is noteworthy since it started its
performance. Instead, exchange rate policy postcolonial experience with minimal invest-
seemed to be driven by rent and fiscal imper- ment and substantial inequality. Of sixty-five
atives and relative price movements were resource rich, developing countries, only
almost a by-product of the resource boom four managed to achieve long-term invest-
(Sala-i-Martin and Subramanian 2003). ment exceeding 25 percent of GDP and an
Other oil exporters (Iran, Venezuela, Libya, average GDP growth exceeding 4 percent—
Iraq, Kuwait, Qatar) experienced negative namely Botswana, Indonesia, Malaysia,
growth during the last few decades. OPEC and Thailand (Thorvaldur Gylfason 2001).
as a whole saw a decline in GNP per capita These three resource rich Asian countries
while other countries with comparable GNP have achieved this by economic diversifica-
per capita enjoyed growth. The deindustrial- tion and industrialization. Still, they fared
ization and disappointing growth experience less well than their neighbors Hong Kong,
of South Africa following the boom in gold Singapore, and South Korea with little raw
prices can be explained by the appreciation material wealth. Norway has shown remark-
of the real exchange rate in the 1970s fol- able growth of manufacturing and the rest of
lowed by gradual depreciations together with the economy compared with its neighbors
increased barriers to technological adoption despite phenomenal growth in oil exports
(Hildegunn E. Stokke 2008). The disruption since 1971 (Svein S. Andersen 1993; Erling
of the “air bridge” from 1994 onwards shifted Roed Larsen 2006). Norway is the world’s
the production of coca paste from Peru and third largest petroleum exporter after Saudi
Bolivia to Colombia and led to a huge boom Arabia and Russia, but is one of the least cor-
in the demand for Colombian coca leaf. This rupt countries in the world and enjoys well
has led to more self-employment and work developed institutions, far sighted manage-
for teenage boys in rural areas but not to ment and market friendly policies.
United Arab Emirates account for close to The positive experiences of the United
10 percent of the world’s crude oil and 4 per- States with its mineral abundance from the
cent of the world’s natural gas reserves but mid-nineteenth to the mid-twentieth cen-
has turned its resource curse into a blessing tury explain much of subsequent economic
(Ugo Fasano 2002). Its government debt is growth (H. J. Habakkuk 1962; Paul A. David
very small, inflation is low, and hydrocarbon and Gavin Wright 1997). It was a choice
wealth has been used to modernize infra- driven by collective learning and leading
structure, create jobs, and establish a gen- education in mining engineering and metal-
erous welfare system. Major strides in life lurgy, geological knowledge, transportation,
expectancy and literacy have been made increasing returns, and an accommodating
through universal and free access to educa- legal environment where the U.S. govern-
tion and health care. In anticipation of deple- ment claimed no ultimate title to the nation’s
tion of its natural resources, oil-rich Abu minerals (Wright and Jesse Czelusta 2002,
Dhabi has emphasized petrochemical and 2003, 2004). The main lesson is that one has
fertilizers, Dubai has diversified into light to learn to make the most of one’s resources
manufacturing, telecommunications, finance, (cf., Jean-Philippe C. Stijns 2005). The role
and tourism, and the other emirates have of private extraction and mining companies
focused on small-scale manufacturing, agri- was crucial in this learning process. The
culture, quarrying, cement, and shipping ser- United States was the world’s leading min-
vices. Many Latin American countries have eral economy in the very period that the
abandoned misguided state policies, encour- country became the world leader in manu-
aged foreign investment in mining, and facturing. Linkages and complementarities
increased the security of mining investment. of the nonresource sectors of the economy to
Since the 1990s, Latin America appears to be the private resource sectors were vital to the
the fastest growing mining region, well ahead American economic success. Governments
of Australia, Canada, Africa, and the United provided weak oversight. High wages may
States in terms of spending on exploitation. have contributed to returns being dispersed
Chile has recently achieved remarkable throughout the U.S. economy.
annual growth rates of 8.5 percent while the In 1913, the United States was the world’s
mining industry accounted for almost half of dominant producer of virtually every major
total exports. Peru ranks second in the world industrial mineral even though other coun-
in the production of silver and tin, fourth tries initially seemed to have more mineral
in zinc and lead, and eighth in gold and its reserves. New deposits were continuously
mineral sectors enjoy prospects for further discovered. The U.S. share of world mineral
growth. Another leader in this region is production in 1913 was far in excess of its
Brazil. Argentina seems to be moving ahead share of world reserves; mineral rich coun-
as well. tries like Brazil, Chile, Russia, Canada, and
Australia did much worse in developing new
2.2 Historical Evidence: Natural Resources,
reserves and cheaper techniques (David and
Evolution of Property Rights, and
Wright 1997). The U.S. experience suggests
Innovation
that impending scarcity of natural resources
Successful resource-based development can be compensated by technical progress in
does not primarily depend on geological exploration, extraction, and substitution and
endowment. The United States developed privatization of reserves. Many resource rich
its mineral potential ahead of other countries economies may have performed badly, not
and continents, including Latin America. because they relied too much on resources,
but because they failed in developing their and enforcing rights offset the higher costs
mineral potential through appropriate poli- of doing so (Harold Demsetz 1967). Private
cies. Investment in minerals-related knowl- mineral rights indeed became more explicit
edge seems a legitimate component of a as mine values increased. They evolved from
forward-looking development program. local property rules within the mining camps
Coal and iron ore deposits spurred indus- to formal territorial and state statutes and
trial development of Germany and the United judicial opinions as the extent and value of
Kingdom during the late nineteenth century. the deposits in the regions became more
South Korea and Japan have taken advantage of apparent (Libecap 1978). With increased
fallen transport costs and have become impor- competition for valuable resources, informal
tant steel producers despite relying on import rules were insufficient to reduce risk and
of iron ore. Still, history shows that good expe- support long-term investment to develop the
riences of resource rich economies are not mines. Making property rights more formal
always replicated. In the seventeenth century, boosted mining investment. However, in case
resource poor Netherlands outpaced Spain, of the Western timberlands, the transaction
even though the latter obtained much gold costs were more than the government price
and silver from its colonies in the New World. of land and timber depredations continued
More recently, resource poor Switzerland has (Libecap and Ronald N. Johnson 1979).
enjoyed an excellent economic performance These case studies suggest interesting
compared with resource rich Russia. In sum, hypotheses about transactions costs and the
the effects of natural resources on the econ- implications of property rights for turning the
omy vary from country to country and across resource curse into a blessing. For example,
different episodes in history. if transport costs are high relative to those
More work is required on the changing of manufactured goods, extra resources
role of natural resources throughout history. lower the domestic price for a key input to
The resource curse features especially dur- manufacturing giving domestic manufactur-
ing the last four decades, but before coun- ers a comparative advantage. For example,
tries such as the United States seemed to car producers in Detroit had cheap access
have harnessed resources for growth. Is this to iron ore. Another hypothesis is that those
because those countries that industrialized exploiting the natural resource can sell their
first also had good institutions and those rights and consume the entire present value
countries that remained underdeveloped of their reserves, thus causing an initial con-
had bad institutions and when resources sumption boom. Otherwise, their consump-
were exploited at a later stage they led to tion possibilities seem more limited leading
corruption, rent seeking, and strife? Key is to a higher saving rate. Another interesting
the contractual basis for exchange. Natural question is whether the more widespread
resources may be underproduced due to lack ownership of resources in the nineteenth
of effective property rights and high transac- century had something to do with a smaller
tion costs (Terry L. Anderson and Gary D. minimum efficient scale of production.
Libecap 2005). The Coase theorem says that
2.3 Cross-Country Correlations
with well-defined property rights private, vol-
untary negotiations yield efficient outcomes, Figure 1 indicates a negative correlation
but high transactions costs may preclude between growth performance and the share
such outcomes. More valuable resources of natural resources in merchandise exports,
tend to have more precise property rights but this does not tell us anything about cau-
because the larger benefits from defining sation. Natural resource d ependence may
China,P.R.: Mainland
Korea
China,P.R.:Hong
ThailandKong Malaysia
Portuga
Sri Poland Tunisia
Dominican Republic
Japa lAustri Lanka Indi
Finlan
Spai
Hungar
nIsrae Ital
Ugand Belgiu a United
Brazi Greec Canada
PakistaaGerman
dUnited
yn
Franc Netherland
y m Turke
Nepaln a yDenmar
Paragua
Sw ede
eStates Kingdom
l e Colombi
Mozamb iqu Morocc
Mexico
Burkina k yn Sudas
yl Banglades
Tanzani Slovak Philippine Bulgaria
ea
Keny Cameroon
Malaw
Faso Switzerlan
h Guatemal
Beni
Mal a Czech s
i a dnEl Hondura
Ethiopi Republic
Rwanda
sRepublic
Argentin Romani Belaru Senegal
Salvador South
Burundi Chad Ghana
s Africa
a Zimbabwe
Côte
Haiti d'Ivoire Georgia
Nicaragua
Sierra
Leone
Ukraine
Figure 1. Growth and Natural Resource Dependence

Source: World Development Indicators, 2004, World Bank
harm the economy through other variables GDP (a measure of financial development).
than lower growth (e.g., Gylfason, Tryggvi Furthermore, although there are excep-
Thor Herbertsson, and Gylfi Zoega 1999; tions such as Botswana, there is an inverse
Gylfason 2001, 2004). For example, partial correlation between resource dependence
cross-country correlations for oil exporters and school enrollment at all levels, expected
in the Arab world and elsewhere suggest years of schooling, and public spending on
that resource dependence is associated with education. This may matter as there is a
less nonresource exports and foreign direct positive correlation between education and
investment. Evidence of a sample of eighty- growth. Finally, empirically there is a posi-
seven countries suggest that resource tive correlation between natural resource
wealth is associated with less openness to dependence and macroeconomic volatil-
foreign trade and less openness to gross for- ity and a negative correlation between
eign direct investment, which in turn may macroeconomic volatility and growth
harm growth prospects. Also, in a sample (e.g., Frederick van der Ploeg and Steven
of eighty-five countries the share of natural Poelhekke 2009). These partial correlations
resource wealth in national capital is nega- are not inconsistent with the suggestion that
tively correlated with both gross domes- resource dependence crowds out foreign,
tic investment as percentage of GDP and social, human, real, and financial capital,
the average ratio of broad money (M2) to each effect tending to depress growth.
Table 1
Total, Natural, Produced and Intangible Capital, 2000
($ per Capita and Percentage Shares)
Natural Produced Intangible

Natural Produced Intangible Total capital capital capital
Income group capital capital capital wealth share share share
Low-income countries 1,925 1,174 4,434 7,532 26% 16% 59%

Middle-income countries 3,496 5,347 18,773 27,616 13% 19% 68%
High-income OECD countries 9,531 76,193 353,339 439,063 2% 17% 80%
World 4,011 16,850 74,998 95,860 4% 18% 78%
Note: All dollars at nominal exchange rates. Oil states excluded.

Source: World Bank 2006, table 2.1.
2.4 World Bank Data on Natural Capital for low-income to $16,430 for high-income
and Wealth of Nations countries. Tables 1 and 2 give a flavor of the
detailed results.1 Although global wealth
Various components of national wealth for per capita is $96,000, this masks huge vari-
the year 2000 (approximated by the pres- ety across countries. The share of produced
ent value of sustainable consumption dur- assets in total wealth is more or less the same
ing 2000–25 using a social discount rate of 4 irrespective of how poor or rich a country
percent) have been calculated for nearly 120 is. However, the share of natural capital in
countries in the world (World Bank 2006). total wealth is much higher in poorer coun-
Produced capital is estimated from historical tries while the share of intangible capital in
investment data with the perpetual inventory total wealth is substantially higher in richer
method. Natural capital consists of subsoil economies. Interestingly, richer countries
assets, timber resources, nontimber forest have a substantially higher value of natural
resources, protected areas, cropland, and capital per capita despite having lower shares
pastureland. Due to data problems, fisheries, of natural capital in total wealth. The results
subsoil water, and diamonds are excluded. confirm what we know from the literature
The explicit value of ecosystems is not evalu- on economic growth that intangible capital
ated either. The value of natural capital is is the main engine of growth and wealth.
estimated from world prices and local costs. Richer countries focus relatively more on
Intangible capital reflects the contribution of dynamic sectors such as manufacturing and
raw labor, human capital, R&D, social capi- services, whereas poorer countries specialize
tal, and other factors such as institutions and in the more static primary sectors.
rule of law. It is calculated residually as the
excess of total national wealth over the sum 1 One of the referees pointed out that these estimates
of produced and natural capital and is well of the share of resources in national wealth include human
explained by school years per capita, a rule wealth, so that countries with a high wage level such as
of law index, and remittances per capita. For Norway are measured as having a relatively small fraction
of their wealth in natural resources. Also, it is more dif-
example, an extra year of schooling yields ficult to control for initial conditions than with the ratio of
extra intangible capital varying from $840 resource exports to GDP.
Table 2
Components of Natural Capita, 2000 ($ per Capita)
Total
Subsoil Timber Protected natural
Income group assets resources NTFR areas Cropland Pastureland capital
Low-income countries 325 109 48 111 1,143 189 1,925
Middle-income countries 1,089 169 120 129 1,583 407 3,496
High-income countries (OECD) 3,825 747 183 1,215 2,008 1,552 9,531
World 1,302 252 104 322 1,496 536 4,011
Note: NTFR stands for non-timber forest resources. Oil states excluded.
Source: World Bank 2006, table 1.2.
Table 2 indicates that the poorer coun- to boost their economic performance and
tries rely relatively heavily on land resources others have done worse. Here we d iscuss
(more than two thirds of natural wealth in the theoretical support and evidence where
low-income countries). In the ten wealthi- available for a wide range of hypotheses
est countries, only Norway has a natural about the effects of natural resources on the
capital share of more than 3 percent (namely economy and society.2 Section 3.1 puts for-
12 percent). On the other hand, the bot- ward the hypothesis that a resource bonanza
tom ten countries all have shares of natural induces appreciation of the real exchange
capital in total wealth exceeding 30 percent. rate, contraction of the traded sector, and
Table 3 indicates that highly resource rich expansion of the nontraded sectors and offers
economies, such as the oil exporters Nigeria, some evidence for Brazil on this hypothesis.
Venezuela, and Algeria, sometimes even Section 3.2 shows that, if the traded sector
have negative shares of intangible capital in is the engine of growth, a resource bonanza
total wealth. This suggests that these coun- will lead to a temporary fall in growth. Early
tries have extremely low levels of GNI as cross-country evidence indeed indicates a
their returns on productive and intangible negative link between resources and growth.
capital are very low and possibly even nega- Subsequent panel-data and quasi-experimen-
tive. Consequently, they have very low total tal tests of this hypothesis are also discussed.
wealth and can sustain only very low levels Section 3.3 puts forward the hypothesis
of consumption per capita. This is another that the resource curse can be turned into
manifestation of the resource curse.
2 Earlier empirical work attempts to identify the poten-
tial channels of transmission for the resource curse by
3. Popular Explanations of the Natural regressing institutional quality, human capital, etc. on
Resource Curse natural resource dependence only and calculating the
indirect effects of resource dependence on growth from
The stylized facts discussed in section 2 the coefficients of these intermediate variables on growth
suggest that the experiences of resource rich (Elissaios Papyrakis and Reyer Gerlagh 2004; Jann Lay and
Toman Omar Mahmoud 2004), but this approach suffers
countries have been very heterogeneous. from potential omitted variable bias and other econometric
Some have harnessed their resource wealth problems.
Table 3
Intangible Capital and Wealth Composition in Highly Resource-Rich Countries
Percentage share of total wealth

Intangible capital Natural Produced Intangible
Country per capita ($) capital capital capital
Russian Federation 6,029 44 40 16
Guyana 2,176 65 21 14
Moldova 1,173 37 49 13
Venezuela 4,360 60 30 10
Gabon –3,215 66 41 –7
Syrian Arab Republic –1,598 84 32 –15
Algeria –3,418 71 47 –18
Nigeria –1,959 147 24 –71
Congo –12,158 265 180 –346
Source: World Bank 2006, p. 29.
a blessing for countries with good institu- countries experience negative genuine sav-
tions and provides some evidence in support ing. Of course, there may be other hypoth-
thereof. Section 3.4 discusses the hypotheses which we do not touch upon.3
esis that presidential democracies are more
3.1 Dutch Disease: Natural Resource
likely to suffer a negative effect of resources
Windfalls Cause Deindustrialization
on growth. Section 3.5 reviews economet-
ric and quasi-experimental evidence for the Early policy contributions highlight the
hypothesis that resource windfalls increase appreciation of the real exchange rate and
corruption, especially in countries with the resulting process of deindustrialization
nondemocratic regimes. Section 3.6 offers induced by the increase in oil exports in Britain
econometric support for the hypothesis that (Peter J. Forsyth and John A. Kay 1980, 1981).
volatility of resource windfalls is the quin- There has also been a relative decline of
tessence of the resource curse and also for Dutch manufacturing as a result of worsening
the hypothesis that the negative effect on of competitiveness a ssociated with the export
growth is less in countries with well devel-
oped financial systems. Section 3.7 puts for-
3 For example, resource dependence seems to be corre-
ward the hypothesis that resources induces
lated with a bigger Gini index of inequality and less politi-
voracious rent seeking and armed conflict, cal liberties, which in turn are correlated with lower growth
and examines cross-country, panel-data, and (Gylfason and Zoega 2003). Empirical evidence suggests
quasi-experimental evidence for this hypoth- that resources increase income inequality only in ethnically
polarized societies, after controlling for GDP, schooling,
esis. Section 3.8 discusses the hypothesis that and constraints on the executive (Ruikang Marcus Fum
resource windfalls encourage unsustainable and Roland Hodler 2010). Income inequality also reduces
and unwise policies. Section 4 is entirely immediately after an oil or mineral boom and increases
gradually thereafter; uncertainty about future commod-
devoted to two hypotheses that might ity export prices seems to increase long-run inequality
explain why many resource rich developing (Benedikt Goderis and Samuel W. Malone forthcoming).
of natural gas found in Slochteren (M. Ellman yields P ε = H[F(1 − LN) + E]/G(LN), where

1981). The idea behind this Dutch disease is H ≡ HT /HN is the productivity of the traded
that the extra wealth generated by the sale of and resource sectors relative to that of the
natural resources induces appreciation of the nontraded sector. This equation corresponds
real exchange rate and an ensuing contraction to the NTGME-locus in figure 2 and describes
of the traded sector (W. Max Corden and J. those combinations of the real exchange
Peter Neary 1982; Corden 1984). rate P and the share of labor employed in
We illustrate this with the Salter–Swan the nontraded sector LN that ensure clear-
model of a two-sector economy with a ing of the market for nontraded goods. The
resource windfall, abstracting from capital NTGME-locus slopes downwards, since a
accumulation, international investment, and higher P is associated with relatively lower
financial assets. Export of resources thus demand for nontraded goods and, thus,
equals net imports of traded goods, that with fewer workers employed in the non-
is HT E = CT − HT F(L T) where  denotes traded sector. Labor mobility between
the world price of natural resources, E the traded and nontraded sectors requires that
volume of exports of natural resources, CT labor is paid the same in each sector, so
consumption of traded goods, L T employ- that the value of the marginal product of
ment in the traded sector, HT productivity labor is equalized. This yields the LM-curve
in the traded and natural resource sectors, PG′(LN) = HF′(1 − LN), which gives those
and HT F(L T) output of the traded sector combinations of the real exchange rate P
(with F′ > 0, F′′ ≤ 0). Nontraded goods and the share of labor employed in the non-
market equilibrium requires CN = HN G(LN), traded sector LN that ensure labor market
where CN denotes consumption of nontraded equilibrium. The LM-curve slopes upward.
goods, LN employment in the nontraded A higher relative price of nontraded goods
sector, HN productivity in the nontraded sec- P pushes up the value of the marginal prod-
tor, and HN G(LN) output of the nontraded uct of employment in the nontraded sector,
sector (with G′ > 0, G′′ ≤ 0). With exogenous so employment in the traded sector must
labor supply of one unit and labor mobility decline in order to push up the marginal
between traded and nontraded sectors, labor product of labor in the traded sector.
market equilibrium requires L T + LN = 1. Higher natural resource revenue E
Households maximize utility U(CN, CT) sub- boosts national income and demand. Hence,
ject to the budget constraint PCN + CT = Y, the NTGME-locus shifts upwards, the
where P is the relative price of nontraded LM-locus is unaffected and equilibrium in
goods in terms of traded goods and national figure 2 shifts from A to A′. The short-run con-
income is defined by Y ≡ PHN G(LN) + sequences of higher resource revenues are
HT F(L T) + HT E. Optimality requires thus appreciation of the real exchange rate
UN/UT = P. With CES utility, we have (a higher relative price of nontraded goods
CN = Y/(1 + P ε−1)P, where ε is the elastic- P), decline of the traded sector and expan-
ity of substitution between traded and non- sion of the nontraded sector. Labor shifts
traded goods. The condition for equilibrium from the exposed to the sheltered sectors.
in the market for nontraded goods, This boosts both consumption and output of
nontraded goods. The rise in consumption of
HN G(LN) = CN = Y/(1 + P ε−1) traded goods and the contraction in the pro-
duction of traded goods is made possible by
[PHN G(LN) + HT F(L T) + HT E]
___ additional imports financed by the increase
=          ,
(P + P ε) in resource revenues. National income rises
Relative
price of LM
nontraded
LM
goods
A
A
A
NTGME
NTGME
NTGME
B
Figure 2. Natural Resource Dependence Reduces Competitiveness

Note: A resource boom shifts A to A′, so a shift from the traded to nontraded sector and real appreciation. With
time, relative productivity of the traded declines if the elasticity of substitution in demand goods is less than
unity. This shifts the equilibrium from A′ to A′′ and eventually to B. In the long run, there is real depreciation
and the allocation of labor is returned to its original level.
by more than natural resource revenues For the longer run effects one must allow
(d Y  =  H T d(E) +  C N dP > H T d(E)). capital and labor to be mobile across sectors
The natural resource bonanza thus increases and move beyond the specific factors frame-
welfare.4 The short-run effects of the Dutch work. In an open economy Heckscher–Ohlin
disease on unemployment are discussed in framework with competitive labor, capital,
appendix 1. and product markets, no resource use in pro-
duction and constant returns to scale in the
4 More elegant is to use duality (Neary 1988). Let
production of traded and nontraded goods,
Z( p) denote nonresource national income, so that Z′(P) a natural resource windfall induces a higher
equals nontraded output. Equilibrium in traded and (lower) wage-rental ratio if the nontraded
nontraded goods is given by Z(P) + E = e(P)U and sector is more (less) labor-intensive than the
Z′(P) = e′(P) U, respectively, where e(P) = Y/U indi-
cates the CPI and U denotes real consumption (utility). traded sector. In any case, there is a rise in
It follows that d U/d(E) = 1/e(P) > 0 and d P/d(E) the relative price of nontraded goods leading
= (PCN /Y)/(εS + εD), so that windfall revenue from abroad to an expansion of the nontraded sector and
boosts utility. It also leads to an appreciation of the real
exchange rate, especially if the share of nontraded goods a contraction of the traded sector. Labor and
in the consumption basket is large, the supply elastic- capital shift from the traded to the nontraded
ity εS ≡ P Y′′/CN > 0 is small, and the demand elasticity sectors. More interesting may be to study
εD ≡ −Pe′′ U/CN > 0 is small. If labor supply increases with
the real consumption wage (migrants, informal labor), the the effects of a resource boom in a dynamic
real exchange rate appreciates less. dependent economy with adjustment costs
for investment and allow for costly sectoral from the deindustrialization effects arising
reallocation of capital between nontraded from an appreciation of the real exchange
and traded sectors (A. K. M. Mahbub rate (Corden and Neary 1982). If the non-
Morshed and Stephen J. Turnovsky 2004). traded sector is more capital intensive, the
It is then more costly to transform one form real exchange rate depreciates if labor is
of existing capital into another, since this needed to secure the resource windfall; the
involves demolition. This way one has fac- Rybczinski theorem then says that the non-
tor specificity for each sector in the short run traded sector expands and the traded sector
and factor mobility across sectors in the long contracts. This increase in relative supply of
run. An advantage of this approach is that in nontraded goods fuels depreciation of the
the short and medium run the real exchange real exchange rate. Real exchange deprecia-
rate is no longer fully determined by the sup- tion may also result from a boost to natural
ply side and does not adjust instantaneously. resource exports if the traded sector is rela-
If a greater fraction of resource revenues tively capital intensive and capital is needed
is saved, the initial appreciation of the real for the exploitation of natural resources
exchange rate will be less and will eventually (Neary and Douglas Purvis 1982). Since less
be reversed (see appendix 4). One could also capital is available for the traded sector, less
use a model of endogenous growth in the labor is needed and thus more labor is avail-
dependent economy (e.g., Turnovsky 1996) able for the nontraded sector. This may lead
to explore the implications of a resource to a depreciation of the real exchange rate.
boom on economic growth. This also occurs if the income distribution is
What happens if the exploitation sector shifted to consumers with a low propensity
uses labor and capital as factor inputs? Apart to consume nontraded goods (Corden 1984).
from the hitherto discussed spending effects
3.1.1 Empirical Evidence for Dutch Disease
of a resource boom, there are also resource
Effects
movement effects (Corden and Neary 1982).
Deindustrialization occurs on account of Although early evidence for a shrinking
the usual appreciation of the real exchange manufacturing sector in response to terms
rate (the spending effect), but also due to of trade shocks and real appreciation has
the labor drawn out of both the nontraded been mixed (Sala-i-Martin and Subramanian
and traded sectors toward the resource sec- 2003), more recent evidence for 135 coun-
tor (the resource movement effect). Looking tries for the period 1975–2007 indicates that
at the longer run where both factors of the response to a resource windfall is to save
production (labor and capital) are mobile about 30 percent, decrease nonresource
between the traded and nontraded sectors exports by 35–70 percent, and increase non-
and the resource sector only uses labor, it resource imports by 0–35 percent (Torfinn
helps to consider a mini-Heckscher–Ohlin Harding and Anthony J. Venables 2010).
economy for the traded and nontraded sec- These findings hold in pure cross-sections of
tors. The Rybczinski theorem states that the countries (averages across one, two, three,
movement of labor out of the nonresource or four decades), in pooled panels of coun-
toward the resource sectors causes output tries, and in panel estimations including
of the capital-intensive nonresource sector dynamics and country fixed effects. Another
to expand. This may lead to the paradoxical study uses detailed, disaggregated sectoral
result of pro-industrialization if capital-inten- data for manufacturing and obtains simi-
sive manufacturing constitutes the traded lar results: a 10.0 percent oil windfall is on
sector, despite some offsetting effects arising average associated with a 3.4 percent fall in
value added across manufacturing, but less communities much. The evidence for Brazil
so in countries that have restrictions on capi- thus offers support for the Dutch disease
tal flows and for sectors that are more capital hypothesis, but also to waste in local govern-
intensive (Kareem Ismail 2010). Using as a ment and corruption (see section 3.3).
counterfactual the Chenery–Syrquin (1975)
3.2 Temporary Loss in Learning by Doing
norm for the size of tradables (manufac-
Curbs Economic Growth
turing and agriculture), countries in which
the resource sector accounts for more than A declining traded sector is the appropri-
30 percent of GDP have a tradables sector ate market response to a resource windfall.
15 percentage points lower than the norm In itself this does not justify government
(Milan Brahmbhatt, Otaviano Canuto, and intervention since it is optimal to special-
Ekaterina Vostroknutova 2010). The macro- ize in one’s comparative advantage. Why
economic and sectoral evidence thus seems are resource windfalls then perceived to
to offer support for Dutch disease effects. be a problem? One popular answer is that
Interestingly, macro cross-country and micro the traded sector is the engine of growth
U.S. county level evidence suggests that and benefits most from learning by doing
resource rich countries experience despe- and other positive externalities, hence non-
cialization as the least skilled employees resource export sectors temporarily hit by
move from manufacturing to the nontraded worsening competitiveness are unable to
sectors thus leading their traded sectors to fully recover when resources run out. This
be much more productive than resource can be demonstrated in a two-period, two-
poor countries (Karlygash Kuralbayeva and good Salter–Swan model where learning by
Radosław Stefánski 2010). doing is captured by future productivity of
Quasi-experimental, within-country evi- the traded sector increasing with current
dence on the Dutch disease for Brazil has production of traded goods (Sweder J. G.
recently also become available (Francesco van Wijnbergen 1984a) or with cumulative
Caselli and Guy Michaels 2009). This study experience (Paul Krugman 1987).5 If man-
exploits a dataset on oil dependence for ufacturing rather than agriculture enjoys
Brazilian municipalities, which is useful as learning by doing and the income elasticity
oil fields are highly concentrated geographi- of demand for agricultural goods is less than
cally and local resource dependence is more unity, shifting from manufacturing toward
likely to be exogenous as it is decided by the agriculture curbs growth in an open econ-
national oil company, Petrobras. It turns out omy (Kiminori Matsuyama 1992). Similarly,
that oil discoveries and exploitation do not if human capital spillover effects in produc-
affect non-oil GDP very much, albeit that tion are generated only by employment in
in line with the Dutch disease hypothesis the traded sector and induce endogenous
services expand and industry shrinks some- growth in both traded and nontraded sec-
what. But they do boost local public revenue, tors, natural resource exports lower employ-
20–25 percent (rather than 10 percent) going ment in the traded sector, hamper learning
to housing and urban development, 15 per- by doing, and thus stunt economic growth
cent to education, 10 percent to health, and (Jeffrey D. Sachs and Andrew M. Warner
5 percent on welfare. Interestingly, house-
hold income only rises by 10 percent, mostly
5 Similarly, giving aid to developing countries may lead
through higher government wages. The lack
to appreciation of the real exchange rate and decline of
of migration to oil-rich communities also manufacturing (Christopher S. Adam and Bevan 2006;
suggests that oil does not really benefit local Adam and Stephen O’Connell 2004).
1995; Gylfason, Herbertsson, and Zoega H ≡ HT /HN induces real depreciation (lower
1999). P) and, given ε < 1, a smaller nontraded sec-
With perfect international capital mobil- tor (lower LN). After an increase in E, the
ity and no specific factors of production, the economy gradually converges to the lower
wage, the relative price of nontraded goods, steady-state value of H, so over time produc-
and the capital intensities in the traded and tivity of the traded sector declines relative to
nontraded sectors are pinned down by the that of the nontraded sector.
world interest rate. Higher resource rev- We have already seen in section 3.1 that
enue then induces gradual movement of higher natural resource exports lead initially
labor from the traded to the nontraded sec- to real appreciation and expansion of the
tor. This reduces learning by doing and thus nontraded sector (the shift from A to A′ in
lowers the rate of labor-augmenting techni- figure 2). Over time, relative productivity of
cal progress so that the resource boom per- the traded relative to that of the nontraded
manently lowers the rate of growth. One can sector H declines gradually. This induces
show that nonresource GDP falls on impact gradual depreciations of the real exchange
after a resource discovery if the traded rate and falls in labor use in the nontraded
sector is capital-intensive (see appendix 2). sector, and corresponds to the movement
However, if production of traded goods from A′ to A′′ and eventually B in figure 2. In
requires natural resources as factor input, a the end, this completely chokes off the initial
higher world price of natural resources leads expansion of the nontraded sector and elimi-
to depreciation of the real exchange rate and nates the boom of the traded sector through
a lower capital intensity in the production of gradual depreciation of the real exchange
nontraded goods which accentuates the fall rate. The new steady-state level of produc-
in traded sector employment and throttles tion has also moved in favor of the nontraded
learning by doing and growth even more. sector, not due to reallocation of labor but
To illustrate how a resource boom affects due to the relative fall in the productivity of
relative productivity growth of the traded and the traded sector.
nontraded sector, the adverse effects of the
3.2.1 Empirical Evidence for Negative
Dutch disease on growth are illustrated with
Effect of Natural Resources on
a dynamic two-sector economy without capi-
Economic Growth
tal accumulation, absence of current account
dynamics and balanced trade (Ragnar Torvik The pioneering study on the empirical
2001). Both traded and nontraded sectors cross-country evidence shows that resource
contribute to learning. A foreign exchange rich countries indeed grow on average about
windfall arising from resource exports then one percentage point less during 1970–89
leads to appreciation of the real exchange even after controlling for initial income per
rate in the short run but real depreciation capita, investments during the period, open-
in the long run. To illustrate, allow produc- ness, and rule of law (Sachs and Warner
tivity growth in each sector to increase with 1995). The revised cross-country regressions
the number employed in that sector and explaining average growth in real GDP per
suppose that learning by doing is more sub- capita during 1970–90 are reported in the
stantial in the traded than nontraded sector. first regression of table 4. There is evidence
Suppose also that the elasticity of substitu- of conditional convergence since countries
tion between traded and nontraded goods with a low (log of the) level of initial real
in consumption ε is less than unity. A fall GDP per active member of the population
in relative productivity of the traded sector catch up and grow relatively fast. Countries
Table 4
Effects of Resource Dependence and Institutional Quality on Economic Growth
Annual growth in real Based on data in Sachs Mehlum, Moene,

GDP per capita Sachs and Warner (1997a) and Warner (1997b) and Torvik (2006b)
Initial income –1.76 (8.56) –1.28 (6.65) –1.26 (6.70)

Openness 1.33 (3.35) 1.45 (3.36) 1.66 (3.87)
Resource dependence –10.57 (7.01) –6.69 (5.43) –14.34 (4.21)
Rule of law 0.36 (3.54) — —
Institutional quality — 0.6 (0.64) –1.3 (1.13)
Investments 1.02 (3.45) 0.15 (6.73) 0.16 (7.15)
Interaction term — — 15.40 (2.40)
Number of countries 71 87 87
Adjusted R2 0.72 0.69 0.71
with a high log ratio of real public and private (dependence) is potentially endogenous and,
gross domestic investment to real GDP aver- if instrumented, it does not significantly affect
aged over 1970–89 grow faster. Countries growth whereas subsoil resource wealth
with a large number of years in which their (abundance) does have a significant positive
economy is rated as open and whose citizens effect on growth (Christa N. Brunnschweiler
accept the rule of law more easily (on a scale and Erwin H. Bulte 2008). However, natu-
from 1 to 6) grow faster. Even taking account ral resource wealth is also endogenous as it
of these traditional growth determinants, is calculated as the present value of natural
there is a strong negative effect of resource resource rents. If it is instrumented with the
dependence (measured by the share of more exogenous measure of economically
exports of primary products in GNP in 1970) recoverable reserves, there is no evidence for
on growth. This is what has become known either a curse or a blessing unless one allows
as the resource curse. This pioneering study for an indirect effect via volatility (van der
gives no role for institutions or bureaucratic Ploeg and Poelhekke 2010). Another issue
quality in explaining the curse. The second is the negative correlation between growth
regression reported in table 4 uses more performance and resource dependence,
countries, more years, and an index of insti- which may merely be picking up cross-
tutional quality (on a scale from 0 to 1). Using country variations in income per capita.
the starting year 1965 rather than 1970, it Alternatively, if the nonresource traded sec-
confirms that resource rich economies expe- tor declines and the wage premium for edu-
rience slower growth and that institutional cation falls, resource rich economies might
quality is not significant at the 5 percent level invest less in education and thus the growth
(see, however, section 3.3). rate falls. Hence, adding a control for edu-
These regressions are the cornerstone of cation implies that the negative coefficient
many discussions of the resource curse but on resource dependence should fall. Similar
can be criticized on econometric grounds. points apply to intermediate variables such
For example, the share of resources in GNP as wars or institutional quality, so one should
be careful about drawing inferences about whether it is openness to international trade,

the speed of convergence from the coeffi- quality of institutions, or financial devel-
cient on initial income. There may also be opment since these variables are highly
some omitted variable bias if a third factor correlated. The road forward might be to
say “underdevelopment” is driving income as exploit variation within a country where vari-
then countries with a low income potential ables that might confound the relationship
are measured as resource rich. between resources and macroeconomic out-
It is crucial to move from cross-country to comes do not vary and the danger of spuri-
panel data evidence to avoid omitted various correlation is minimized (cf., Caselli and
able bias arising from correlation between Michaels 2009).
initial income per capita and the omitted ini-
3.3 Turning the Resource Curse into a
tial level of productivity (Stephen L. Parente
Blessing: Good Institutions and No
and Edward C. Prescott 1994; Nazrul Islam
Corruption
1995). If resource dependence is expressed
as a fraction of national income, cross-coun- Increased corruption hampers economic
try regressions that do not control properly growth (Paolo Mauro 1995; Pranab Bardhan
for initial productivity underestimate the 1997; Carlos Leite and Jens Weidmann
speed of convergence and overestimate the 1999). Mineral wealth may prevent redis-
share of capital in value added. Even though tribution of political power toward the
this requires reliable data on changing quality middle classes and thus prevent adoption
of institutions, school attainment, resource of growth-promoting policies (Francois
dependence, etc., such problems need Bourguignon and Thierry Verdier 2000).
not arise with panel data regressions. One Resource wealth worsens quality of institu-
panel study investigating the link between tions since it allows governments to pacify
resources, institutional development and dissent, avoid accountability, and resist mod-
growth in ninety-one developing countries ernization (Isham et al. 2005). Corruption
during 1970–2000 finds that point-source and granting of import licenses and other
type natural resources (minerals, coffee, coca) privileges to cronies rather than Dutch dis-
retard democratic and institutional develop- ease seem to be why oil riches have ruined
ment, measured by the degree of democracy long-run performance of the Nigerian econ-
for each country over time, and this stunts omy (Sala-i-Martin and Subramanian 2003).
growth (George S. Mavrotas, S. Mansoob Resource wealth makes it easier for dictators
Murshed, and Sebastian Torres 2006; also to buy off political challengers as President
see Michael L. Ross 1999, 2001a). Another Mobuto has done in Congo with its wealth
panel data study finds that the impact of in copper, diamonds, zinc, gold, silver, and
resources on growth found in cross-country oil (Daron Acemoglu, James A. Robinson,
regressions disappears once one allows for and Verdier 2004). Resource riches raise the
fixed effects; resource dependence (primary value of being in power and induce politi-
exports as fraction of GNP) may be correlated cians to expand public sectors, bribe voters
with unobservable characteristics (Osmel by offering them well paid but unproductive
Manzano and Roberto Rigobon 2001). jobs and inefficient subsidies and tax hand-
Cross-country and panel-data results are outs, especially if accountability and state
sensitive to changing the sample period, the competence are lacking (Robinson, Torvik,
sample of countries, or the definition of vari- and Verdier 2006). Those profiting from
ous explanatory variables. The data may sim- the resource sector may bribe politicians to
ply not allow one to distinguish, for example, provide specific semi-public goods at the
expense of manufacturing, which curbs wel- e ntrepreneurship. Both are thus competing
fare if manufacturing enjoys returns to scale activities. If there are more productive entre-
(Bulte and Richard Damania 2008). Natural preneurs, demand in the economy and prof-
resources also make it attractive for political its of each entrepreneur increase provided
elites to block technological and institutional there are demand complementarities in pro-
improvements since this can weaken their duction (Kevin M. Murphy, Andrei Shleifer,
power (Acemoglu and Robinson 2006). and Robert W. Vishny 1989). In contrast, if
Depending on how resource rents affect a greater fraction of talented people is rent
the leader’s probability of survival, they can seeker (political insider, bureaucrat, oli-
induce a self-interested leader to invest more garch, war lord, etc.), the gain per rent seeker
or less in assets that favor growth such as declines. One can then distinguish two out-
rule of law or infrastructure, so the effects comes following a resource bonanza. If insti-
of resources on economic performance can tutions are strong and encourage productive
be highly non-monotonic (Caselli and Tom entrepreneurship, profits of entrepreneurs
Cunningham 2009). On the one hand, the increase. This means that, in equilibrium,
“busy” leader faces budget and time con- less people engage in rent seeking and more
straints. Hence, if a resource boom raises the in productive activities (see outcome A″ in
value of staying in office, he shifts from pro- figure 3). The rent of the resource bonanza is
ductive toward unproductive activities and more than dissipated. Examples of resource
patronage, contributing to a resource curse. rich countries with strong institutions are
On the other hand, the “strategic” leader uses Australia, Canada, the United States, New
the windfall to keep citizens happy and stay Zealand, Iceland, and Norway and also
longer in power, so the windfall becomes a Botswana (Acemoglu, Simon Johnson, and
blessing. A “fatalistic” leader realizes that a Robinson 2003; Acemoglu et al. 2003).
windfall boosts chances of rebellion and thus However, if institutions are weak, the legal
is more short-sighted and puts less effort system dysfunctions and transparency is low,
into developing the nonresource economy rent seeking has a higher return and unfair
and more into inefficient self-preservation. takeovers, shady dealings, corruption, crime,
However, if the leader responds by offer- etc. pay off. A natural resource bonanza thus
ing better and more outside opportunities elicits more rent seekers and there will be
to rebel groups, the windfall may become a less productive entrepreneurs. In equilib-
blessing. rium, profits fall and as a result the economy
A natural resource bonanza encourages is worse off (see outcome A′ in figure 3).
productive entrepreneurs to shift to rent Weak institutions may explain poor per-
seeking. With an aggregate demand exter- formance of oil-rich states such as Angola,
nality (and a constant tax rate and no exter- Nigeria, Sudan, and Venezuela, diamond-
nal trade), this lowers income by more than rich Sierra Leone, Liberia, and Congo, and
the extra income from the resource revenues drug states Colombia and Afghanistan. There
and thus lowers welfare (Torvik 2002). It institutions are often destroyed by civil wars
helps to make a difference between coun- over control of resources. Dependency on
tries with production-friendly institutions oil and other resources hinders democracy
and others with rent grabbing-friendly insti- and quality of governance (e.g., Ross 1999).
tutions (Halvor Mehlum, Karl Moene, and Also, timber booms have induced members
Torvik 2006a, 2006b). Suppose there is a of political elites to dissolve forestry manage-
fixed supply of people that can direct their ment and destroy institutions in Southeast
talent to either rent seeking or productive Asia (Ross 2001b).
Rents
Profits
Rents′ Rents
A′′
A′ A
Profits′
Profits
Entrepreneurs Rent seekers
Figure 3. Rent Grabbing and Producer Friendly Institutions

Note: A resource bonanza shifts equilibrium from A to A′′ if there are strong institutions, which means higher
profits and more entrepreneurs. In case of weak institutions, the equilibrium shifts from A to A′, so profits
decline and number of rent seekers increases.
Source: Mehlum, Moene, and Torvik 2006b.
3.3.1 Empirical Evidence on How no curse. This holds for fifteen out of the
Institutional Quality Transforms eighty-seven countries (including the United
Effect of Resources on Growth States, Canada, Norway, the Netherlands,
New Zealand, and Australia). Five coun-
The estimates reported in section 3.2 tries belong both to the top eight accord-
imply that the curse is cast in stone. But ing to natural resource wealth and to the
subsequent evidence offers support for the top fifteen according to per capita income.
hypothesis that with good institutions the Resource rich countries with bad institutions
curse can be turned into a blessing (Mehlum, typically are poor and remain poor. Related
Moene, and Torvik 2006a, 2006b). The third cross-country evidence strongly suggests
regression in table 4 indicates that countries that natural resources—oil and minerals in
with a high enough index of institutional particular—exert a negative and nonlinear
quality (> 14.34/15.4 = 0.93) experience impact on growth via their deleterious impact
Table 5
Marginal Effects of Different Resources on Growth for Varying Institutional Quality
Ores and metals Mineral Production of gold,

Primary exports exports as share production as silver and diamonds
share of GDP of GDP share of GNP as share of GDP
Worst institutions –0.548 –0.946 –1.127 –1.145

Average institutions –0.378 0.425 0.304 0.279
Average + one –0.288 1.152 1.062 1.183
s.d. institutions
Best institutions –0.228 1.629 1.560 1.776
Note: Institutional quality is an average of the indexes for bureaucracy, corruption, rule of law, risk of expro-
priation of private investment and repudiation of contracts by government.
Source: Boschini et. al. (2007).
on institutional quality6 rather than through within a short period and having control over
worsening of competitiveness of the non- resources. Two types can be distinguished
resource export sectors (Sala-i-Martin and (Boschini, Pettersson, and Roine 2007).
Subramanian 2003). The adverse effect of Institutional appropriability implies that
resource dependence on institutional qual- resource dependence only has an adverse
ity and growth is particularly strong for easily effect on economic development when insti-
appropriable “point-source” resources with tutions are poor. Technical appropriability
concentrated production and revenues and states that the impact of institutional qual-
massive rents such as oil, diamonds, miner- ity and resource dependence is more pro-
als, and plantation crops rather than agricul- nounced the more technically appropriable
ture (rice, wheat, and animals) whose rents the country’s resources are. Table 5 calculates
are more dispersed throughout the economy, the marginal effects of one standard devia-
and with easy appropriation of rents through tion change in various measures of resource
state institutions (Auty 1997, 2001b; Michael dependence that are increasingly technically
Woolcock, Lant Pritchett, and Jonathan appropriable on the average yearly growth
Isham 2001; Isham et al. 2005; Anne D. rate of GDP during 1975–88 for different
Boschini, Jan Pettersson, and Jesper Roine levels of institutional quality (from cross-
2007; Mavrotas, Murshed, and Torres 2006). country regressions with a sample of eighty
Appropriability matters since it indicates industrialized and developed countries, con-
the ease of realizing large financial gains trolling for trade openness, average share
of investment in GDP, and initial level of
income per capita). Going from top to bot-
6 This variable is instrumented by mortality rates of tom in table 5, we see that better institutions
colonial settlers (cf., Acemoglu, Johnson, and Robinson are conducive to growth indicating institu-
2001) and the fraction of the population speaking English
and European languages (cf., Robert E. Hall and Charles tional appropriability. Reading table 5 left
I. Jones 1999). to right, the importance of good institutions
increases in technical appropriability of Persson and Guido Tabellini (2003) and

resources confirming technical appropriabil- using a cross-country sample of ninety coun-
ity. The curse is thus not cast in stone. tries, estimates suggest that the resource
Bad institutions clearly have an adverse curse occurs in presidential, not parliamen-
effect on growth. They may also be more tary democracies (Jorgen Juel Andersen and
powerful explanations of cross-country varia- Silje Aslaksen 2008). Presidential systems
tions in income per capita than geography, are less accountable and less representa-
trade, or economic policies (Douglass C. tive and thus offer more scope for resource
North 1990; Hall and Jones 1999; Acemoglu, rent extraction. In contrast, parliamentary
Johnson, and Robinson 2001, 2003; Acemoglu systems seem better able at using resource
et al. 2003; Dani Rodrik, Subramanian, and revenues to promote growth. The nature of
Francesco Trebbi 2004; William Easterly the constitutional system is empirically more
and Ross Levine 2002), but not everybody important than democratic rule itself for
agrees fully (Edward L. Glaeser et al. 2004). the link between resource dependence and
Cross-country evidence also suggests a signifi- growth. The empirically observed resource
cant negative impact of natural resources on curse seems to be mostly driven by presiden-
income per capita after controlling for insti- tial countries and nondemocratic regimes.
tutional quality, trade openness, and geogra- The adverse effects of resource depen-
phy, and the curse seems particularly severe dence on growth survive controlling for geog-
in countries with bad institutions and low raphy such as kilometers to closest airport,
degrees of trade openness (Rabah Arezki and percentage land in tropics or incidence of
van der Ploeg forthcoming).7 Moving toward malaria (Sachs and Warner 2001). Natural
more trade openness and improving institu- resources can permanently boost income and
tional quality may thus turn the curse into a welfare through higher human capital, and
blessing. Cross-country evidence suggests that this can offset the direct negative effect of
resource dependence weakens institutions natural resources on the growth rate (Claudio
and thus leads to worse outcomes for indica- Bravo-Ortega and Jose de Gregorio 2005).8
tors of welfare such as the human develop- This may explain why Norway has fared bet-
ment index, availability of water, nourishment ter than most resource-dependent Latin
of the population, or life expectancy (Bulte, American countries. It is thus important to
Damania, and Robert T. Deacon 2005). ascertain whether a low growth rate with a
high level of income per capita is a normal
3.4 Natural Resource Curse Stronger in
state of affairs or induced by a resource curse.
Presidential Democracies
There is a host of further cross-country econo-
The average effect of natural resources on metric evidence on the curse (e.g., Leite and
growth across a sample of countries is thus Weidmann 1999; Gylfason, Herbertsson,
not very informative. Depending on qual- and Zoega 1999; Isham 2005). An influen-
ity of institutions and degree of openness, tial study states that primary commodities
there are huge variations. Following Torsten exports and fraction of GDP in mining belong
to the twenty-two most robust variables out
7 Gravity equations for bilateral trade flows are used
of a list of fifty-nine variables in explaining
as instruments for international trade (Jeffrey A. Frankel cross-country variations in economic growth
and David Romer 1999) and the fraction of the population (Sala-i-Martin 1997).
speaking English and Western European languages as the
first language (Hall and Jones 1999) and colonial origins
and settler mortality (Acemoglu, Johnson, and Robinson 8 Human capital does not appear in the growth regres-
2001) as instrument for institutional quality. sions but the interaction term with resources does.
3.5 Resource Windfalls Increase evidence for the effect of windfalls on cor-
Corruption, Especially in ruption can be found in quasi-experimental
Nondemocratic Regimes studies. One recent study compares changes
in perceived corruption in the island São
Resource dependence elicits corruption Tomé, which had a significant oil discovery
and rent seeking via protection, exclusive announcement in 1997–99, with the island
licenses to exploit and export resources by Cape Verde which did not find oil, both
the political elite, oligarchs and their cro- with similar histories, culture, and political
nies to capture wealth and political power. institutions, and uses a unique dataset of the
In a sample of fifty-five countries, resource characteristics of all scholarship applicants
dependence is indeed strongly associated during 1995–2005 and tailored household
with a worse corruption perceptions index surveys (Pedro C. Vicente 2010). It finds that
(from Transparency International, Berlin) corruption increased by close to 10 percent
which in turn is associated with lower growth after the announcements of the oil discovery
(Mauro 1995). Cross-country regressions but decreased slightly after 2004. Another
also suggest that natural resource wealth study uses data on Brazilian municipalities,
stimulates corruption among bureaucrats a political agency theory of career concerns
and politicians (Alberto Ades and Rafael Di with endogenous entry of candidates, and
Tella 1999). It also crowds out social capital, regression discontinuity design (Fernanda
erodes the legal system and elicits armed Brollo et al. 2010). It finds that a municipal
conflicts and civil wars (see section 3.7). windfall of 10 percent increases corruption
Panel evidence covering ninety-nine by 17–24 percent, raises the chances of the
countries during 1980–2004 suggests that incumbent holding on to office by 7 percent,
natural resources only induce corruption and shrinks the fraction of its opponents
in countries that have endured a nondemo- holding a college degree by 7 percent. Such
cratic regime for more than 60 percent of experimental studies pave the way for more
the years since 1956 controlling for income, convincing evidence on natural resources
time-varying common shocks, regional fixed and corruption.
effects, and some other covariates (Sambit
3.6 Volatility of World Resource Prices
Bhattacharyya and Hodler 2010). Effectively,
Harms Exports and Output Growth
“bad” politicians have a bigger incentive to
mimic “good” politicians in democracies. During the 1970s when commodity prices
Democratization may thus be a powerful were high, resource rich countries used them
instrument to curb corruption in resource as collateral for debt but during the 1980s
rich countries. Another study suggests that commodity prices fell significantly. Panel
the combination of high natural resource data estimation suggests that this has thrown
rents and open democratic systems retards many resource rich countries into debt cri-
growth unless there are sufficient checks ses. Indeed, if debt is also an explanatory
and balances which is not the case in many variable in the panel data estimation, the
new resource rich democracies (Collier and effect of resource dependence disappears.
Anke Hoeffler 2009).9 However, the best The empirical results suggest that the effect
9 However, longitudinally truncated, pooled cross- c ounterfactuals and suggests that oil and mineral depend-
sectional evidence may be misleading. Recent longi- ence may not be associated with undermining of democ-
tudinal evidence exploits within-country variations in racy or less complete transitions to democracy (Stephen
resource dependence and regime types to obtain explicit Haber and Victor Menaldo 2008).
of resource dependence is mainly driven by of growth that cannot be explained by the

boom–bust cycles induced by volatile com- conventional, relatively stable determinants
modity prices, debt overhang, and credit such as institutions, geography, and culture.
constraints, and much less by quality of Historical evidence for the period 1870–1939
bureaucracy (data from Stephen Knack and indeed suggests that volatility harms growth
Philip Keefer 1995) or degree of financial for the commodity-dependent “periphery”
development (Manzano and Rigobon 2001). nations rather than for Europe or the United
Changes in natural resource wealth are States (Christopher Blattman, Jason Hwang,
triggered by sudden changes in commod- and Jeffrey G. Williamson 2007). Resource
ity prices or resource discoveries, which can rich countries also suffer from poorly devel-
lead to boom and bust cycles. Resource rev- oped financial systems and from financial
enues are highly volatile (much more so than remoteness, so that they are likely to expe-
GDP) because their supply exhibits a low rience bigger macroeconomic volatility
price elasticity. Dutch disease can also induce (Andrew K. Rose and Mark M. Spiegel 2009).
real exchange rate volatility and thus to less Building on Aghion et al. (2009), van der
investment in physical capital and learn- Ploeg and Poelhekke (2009) show that with
ing, further contraction of the traded sector, commodity price volatility liquidity con-
and lower productivity growth (Gylfason, straints are more likely to bite and thus inno-
Herbertsson, and Zoega 1999). Cross- vation and growth will fall. Extending Garey
country evidence suggests that real exchange Ramey and Valerie A. Ramey (1995), they
rate volatility can seriously harm the long- offer evidence that the adverse growth effect
term productivity growth, especially in coun- of natural resources results mainly from
tries with low levels of financial development volatility of commodity prices, especially for
(Aghion et al. 2009). For a monetary growth point-based resources (oil, diamonds) and
model, it can be shown that real exchange in landlocked, ethnically polarized econo-
rate uncertainty can exacerbate the negative mies with weak financial institutions, cur-
investment effects of domestic credit market rent account restrictions, and high capital
constraints.10 Empirically, IMF data on forty- account mobility. Instrumenting resource
four commodities and national commodity exports with subsoil resource stocks, esti-
export shares and monthly indices on national mates suggest a strong negative and signifi-
commodity export prices for fifty-eight coun- cant effect of macroeconomic volatility on
tries during 1980–2002 suggest that there is a growth and a strong and positive effect of
long-run relationship between real commod- exports of especially point-source resources
ity prices and real exchange rates in about on macroeconomic volatility (van der Ploeg
one-third of these commodity-exporting and Poelhekke 2010).11 The indirect negative
countries (Paul Cashin, Luis F. Céspedes, and
Ratna Sahay 2004). However, many countries
11 The IV estimates yield an insignificant coefficient
with abundant natural resources are likely to
for the effect of point-source natural resources on mean
experience volatile real exchange rates that growth in GDP per capita, but a significant coefficient of
might explain observed volatile growth rates −0.394 at the 1 percent level for the effect of the standard
deviation of unanticipated growth in GDP per capita, and a
significant coefficient of 11.8 and 5.3 at the 1 percent level
10 With endogenous growth, if firms face tight credit for the effects of point-source and diffuse natural resource
constraints, long-term investment is pro-cyclical, amplifies dependence on the variance of unanticipated growth in
aggregate volatility and lowers mean growth for a given GDP per capita. The effects of financial development,
total investment rate (Philippe Aghion et al. 2005). Under openness, the distance to nearest coast or navigable river
complete financial markets, investment is countercyclical on the variance of unanticipated growth in GDP per capita
and mitigates volatility. are also significant at the 1 percent level.
effect of resource exports on growth via the and saving policies and improve efficiency
volatility channel outweighs any direct posi- of financial markets. It also helps to have a
tive effect of resources on growth. A nonlin- fully diversified economy since then shocks
ear specification suggests that the resource to nontraded demand can be accommodated
curse is operative only for countries with a through changes in structure of production
volatility of unanticipated growth exceeding rather than expenditure switching. This is
2.45 percent per annum. So it is operative important for inefficiently specialized coun-
for Bolivia but not for Norway (both have a tries such as Nigeria and Venezuela, but less
dependence of about 15 percent on point- so for diversified countries like Mexico or
source resource exports over the sample). Indonesia or naturally specialized countries
Volatility thus seems the quintessence of such as some Gulf States. Many resource
the resource curse, but is offset somewhat rich economies have highly specialized pro-
in countries with a high degree of financial duction structures and thus are very volatile.
development.
3.7 Natural Resource Wealth Induces
Volatile resource revenues hurt risk-averse
Voracious Rent Seeking12 and Armed
households, but welfare losses induced by
Conflict
consumption risk are tiny compared with
those from imperfect financial markets. If The political economy of massive resource
only debt contracts are available and bank- rents combined with badly defined property
ruptcy is costly, the economy and the real rights, imperfect markets, and poorly func-
exchange rate become more volatile when tioning legal systems provide ideal opportu-
there is specialization in traded goods and nities for rent seeking behavior of producers,
services and the nonresource traded sector thus diverting resources away from more
is small (Ricardo Hausmann and Rigobon productive activities (Gelb 1988; Auty 2001a,
2003). Effectively, shocks to demand for non- 2001b, 2004; Ross 2001a, 2001b). Economists
traded goods and services—driven by shocks demonstrate that resource revenues are
to resource income—are not accommodated prone to rent seeking and wastage. Indeed,
by movements in the allocation of labor but self-reinforcing effects of rent seeking if
by expenditure switching. This demands rent seekers compete and prey on produc-
much higher relative price movements. Due tive entrepreneurs can explain wide cross-
to bankruptcy costs, interest rates increase country differences in rent seeking (Murphy,
with relative price volatility. This causes spe- Shleifer, and Vishny 1993; Acemoglu 1995).
cialization away from nonresource traded More rent seekers lower returns to both rent
goods and services, which is inefficient. seeking and entrepreneurship with possibly
The less it produces of these traded goods large marginal effects on production. Since
and services, the more volatile the economy more entrepreneurs switch to rent seeking
becomes and the higher the interest rate has in times of a resource boom, multiple (good
to be. This causes the traded sector to shrink and bad) equilibrium outcomes arise. More
further until it vanishes.
Volatility is bad for growth but also for
12 Rent seeking is also relevant when countries receive
investment, income distribution, pov-
foreign aid (Jakob Svensson 2000). Aid can remove pres-
erty, and educational attainment (Joshua sure to reform, induce recipients to overstretch them-
Aizenman and Nancy Marion 1999; Karnit selves, cause a Samaritan’s dilemma with the donor
Flug, Antonio Spilimbergo, and Erik expected to bail out bad policies, siphon skilled workers
away from government and thus weaken institutions, and
Wachtenheim 1998). To get round these spark conflict over aid rents (Deborah A. Brautigam and
curses, one could resort to stabilization Knack 2004; Tim Harford and Michael Klein 2005).
rent seekers induce negative external effects production income makes warfare less
that depress profits for remaining entrepre- attractive and conflict less likely to occur,
neurs, which stimulate even more people to whereas higher resource income makes
shift from productive entrepreneurship to warfare more attractive as there is more to
wasteful rent seeking. Increased entrepre- fight over. Indeed, cross-country evidence
neurship can also crowd out rent seeking. suggests a negative relationship between
For example, private business can invent and shocks in the growth of production income
supply new substitutes for restricted imports and the risk of civil war (Collier and Hoeffler
and thus destroy the rents of quota licenses 2004; James D. Fearon and David D. Laitin
(Jean-Marie Baland and Patrick Francois 2003; Edward Miguel, Shanker Satyanath,
2000). and Ernest Sergenti 2004) and a positive
The “voracity effect” also causes a drag relationship between resource income and
on growth as seen after the oil windfalls in conflict (Collier and Hoeffler 2004; Fearon
Nigeria, Venezuela, and Mexico (Philip R. 2005). The export share of primary commod-
Lane and Aaron Tornell 1996; Tornell and ities is the largest single influence on the risk
Lane 1999). This effect implies that dysfunc- of conflict and the effect is nonlinear (Collier
tional institutions and poorly defined property and Hoeffler, 2004).13 For instance, a coun-
rights lead to a classical commons problem try with no resources has a probability of civil
whereby there is too much grabbing and conflict of merely 0.5 percent, but a coun-
rapacious rent seeking of natural resource try with a share of natural resources in GDP
revenues. It supposes a fixed number of rent of a quarter has a probability of 23 percent.
seekers. Capital can be allocated either to There is now a growing body of cross-country
a formal sector where rents derived from a evidence that rents on resources and primary
common-good stock may be appropriated or commodities, especially oil and other point-
to an informal sector with lower returns and source resources, increase chances of civil
no rent seeking. During a natural resource conflicts and wars especially in sub-Saharan
boom returns to capital investment in the Africa through weakening of the state or
formal sector rise, so rent seekers appropri- financing of rebels, sometimes by corpo-
ate proportionately more without destroying rations. Diamonds (Paivi Lujala 2010), oil
the incentive to invest in the formal sector. (Fearon and Laitin 2003; Ross 2004; Fearon
This happens if there is sectoral reallocation 2005; Macartan Humphreys 2005) and nar-
or if the elasticity of intertemporal substitu- cotics (Angrist and Kugler 2008) especially
tion is sufficiently high so that groups do not increase the risk of civil war onsets. Oil
refrain from excessively increasing appropri- increases the likelihood of conflict, especially
ation. Rapacious rent seeking in a Markov-
perfect equilibrium outcome of a differential 13 Katharina Wick and Bulte (2006) show analytically
game lowers the capital left for investment in the possibility of a nonmonotonic relationship between
the formal sector and thus curbs growth. The resources and conflict intensity. Point-based resources
can trigger intense contests but can also facilitate coordi-
higher profitability of investment is more nation on peaceful outcomes. They also demonstrate that
than undermined by each group of rent contesting resources through violent conflict may yield
seekers grabbing a greater share of national superior outcomes than contests through rent seeking.
Taking account of resource dependence being endogenous
wealth by demanding more transfers. As the to conflict seems to remove the statistical correlation
number of rent seeking groups increases, the between resource dependence and conflict onset, since
voracity effect dampens. historically conflict-torn societies become more dependent
on resources (Brunnschweiler and Bulte 2008). Resource
Production and resource income have dif- abundance (reserves under the ground) is associated with
ferential impact on armed conflict. Higher higher income and reduced chance of onset of war.
separatist conflict. Lootable resources such a Heckscher–Ohlin model of international

as gemstones and drug tend to prolong con- trade extended with an appropriation sector
flict but do not increase the chances of the (Ernesto Dal Bó and Pedro Dal Bó forthcom-
onset of conflict. There is no evidence for a ing). The empirical evidence indeed suggests
significant link between (legal) agricultural that the sharp fall in coffee prices in the 1990s
production and conflict. It is onshore rather has increased violence in regions growing cof-
than offshore oil that is more difficult to pro- fee by lowering wages and opportunity costs of
tect, encourages rebel groups and increases joining army groups while the sharp increase
the risk of violent conflict (Lujala 2010). in oil prices has fueled conflicts in oil regions
Some see conflict as reflecting limited by increasing municipal revenue through
capacity of poor countries to put rebellion rapacity (Oeindrila Dube and Juan F. Vargas
down (Fearon and Laitin 2003) and others 2008). Hence, conflict indeed intensifies if the
as lower opportunity cost of fighting (Collier price of labor-intensive commodities such as
and Hoeffler 2004). It matters whether coffee, sugar, banana, palm, and tobacco falls
civil strife and wars result from grievance, a but weakens if the price of capital-intensive
sense of injustice about how a social group commodities such as oil, coal, and gold falls.
is treated (e.g., systematic economic dis- The empirical evidence does not support the
crimination), or greed possibly induced by hypothesis that the state colludes with para-
massive rents of point-source resources as in military groups and protects oil. Also, satellite
Angola, Congo, and Sierra Leone (Murshed evidence does not support the hypothesis that
2002; Ola Olsson and Heather Congdon Fors the fall in coffee prices has induced substitu-
2004). Furthermore, feasibility is important tion toward coca that led to more violence in
if resources lead to ideological leaders being coffee regions; but violent deaths escalated
crowded out by opportunistic, rebel leaders differentially in coca regions during the 1990s
generating the worst civil wars (Jeremy M. (Angrist and Kugler 2008).
Weinstein 2005; Collier and Hoeffler 2005). Worrisome is that the estimated effects
However, cross-country evidence for the of natural resources on the outbreak and
effect of resources on conflict suffers from duration of war may be flawed since it fails
being confounded by the effects of quality of to take account of the potential impact of
institutions, rule of law, etc. on conflict. It is fighting and armaments accumulation on
more insightful to examine determinants of resource extraction itself. In the face of rebel
conflict at the subnational level, thus eliminat- attacks, rapacious depletion may be favored
ing such confounding influences. Exploiting by nationalized mining companies to reduce
variation across four types of violence (guer- the stake to be fought over despite its eco-
rilla attacks, paramilitary attacks, clashes, and nomic costs; furthermore, private mining
war-related casualties) in 900 municipalities companies invest less in unstable countries,
during 1988–2005 for Colombia and mak- especially if their mining investments are not
ing use of individual-level wage data from well protected and the government’s grip on
rural household surveys, a recent study tests office is weak; also commitment problems
the hypothesis that a higher price of capital- lead a government to underinvest in weap-
intensive commodities increases the return ons and mining companies to underinvest
on capital and lowers wages, so boosts conflict in mining equipment; and there may be an
over the ownership of resource production; incentive to bribe rebels to stave off war
conversely, a higher price of labor-intensive (van der Ploeg and Dominic Rohner 2010).
commodities boosts wages and reduces con- Without binding agreements and sufficient
flict. This hypothesis can be derived from military capacity of the resource-owning
faction, war can be avoided if resource rev- output can outweigh the increase in output
enue is transferred to resource-poor rebels due to the resource boom but not in homog-
(Carmen Beviá and Luis C. Corchón 2010). enous countries. There will thus be an ero-
State capacities should be modeled as for- sion of property rights and a resource curse
ward-looking investments by governments if the number of rival factions is large and
that are affected by the risk of external or natural resource revenues are substantial.
internal war, the degree of political instabil- Fractionalization and fighting can thus lead
ity, and dependence on natural resources; to overdissipation of resource rents. Here we
furthermore, repression and war may both show that the presence of natural resources
be driven by resources and peace, repression R can lead to erosion of property rights and
and war could be modeled as an ordered pro- a resource curse, especially if there are many
bit (Timothy Besley and Persson 2010). More rivaling factions (cf., Hodler 2006). Let
empirical work is needed on the relationship group i either work for productive purposes
between natural resources and conflict that li or fight fi. Group i obtains utility:
allows for endogeneity of mining investment
Ui = ϕH li = Ri with
( )[ ]
and resource extraction, development of state
capacity, and repression outcomes.
fi N
Ri =  _     R + (1 − ϕ) ∑     H lj
Especially point-source resource rents
may, by inducing conflict, put democratic ∑ j=1
N
    fj j=1
institutions to a survival test. Under democ-
racy, politicians are less able to appropriate N
resource rents for their own ends, but vio- and ϕ = 1 −  _
1    ∑ 
    f ,
F j=1 j
lent competition with other political fac-
tions is costly as armies need to be paid and
property may be destroyed. Theory suggests where H, N, ϕ, and F denote productivity,
that higher resource rents biases political the number of rivaling groups, the quality
choice from democracy toward violent con- of the legal system, and a measure of incor-
flict especially if politicians are short-sighted; ruptibility, respectively. The specification for
higher income induced by higher productiv- Ri indicates that group i appropriates more
ity makes democracy more likely (Aslaksen resources if they fight more than others, and
and Torvik 2006). the resource windfall and “stolen” resources
Governments of resource rich countries from productive activities are large. The
often seem unable to provide basic secu- specification of ϕ indicates that fighting
rity to their citizens since natural resource undermines effective property rights (cf.,
wealth elicits violence, theft, and looting Herschel Grossman 2001). The optimum
often financed by rebel groups and compet- outcome of this symmetric Nash game is:
(_ N ) ( F ) ( H )
ing war lords (e.g., Stergios Skaperdas 2002;
  1 −  _   _
Mehlum, Moene, and Torvik 2002). The
Nfi =   N − 1
   N − 1
   R    −1
effect of resources on incidence and dura-
tion of civil wars features strongly in political 3 5
rapacious erosion of
science (e.g., Ross 2004; Fearon and Laitin rent seeking property rights
( ) (_ HR   ).
2003; Collier, Hoeffler, and Mans Soderbom
2004). Rival groups fighting about the control ≥ _
  N − 1 
  
over natural resources may harm the qual- N
3
ity of the legal system and thus undermine rapacious
property rights. The resulting d estruction of rent seeking
More effort is devoted to fighting (rent Indonesia, Afghanistan, etc. and less violent
seeking, corruption or conflict) if resource resource conflicts in homogenous societies
revenues that are at stake R are high and the like Botswana. Religion or language is not
return on productive activities H is low. In such a good marker since people can easily
“incorruptible” countries (very large value acquire such characteristics. Caselli (2006)
of F, ϕ close to one), there can still be rapa- argues that resource dependence gener-
cious rent seeking, especially if the number ates power struggles and political instability,
of competing factions N is large, even though which increases the effective discount rate of
the erosion of property rights is insignificant. the governing group. Consequently, the elite
In “corruptible” countries (with low value invest less in long-run development.
of F), rent seeking is much more severe
3.8 Natural Resource Wealth Leads to
and may even erupt into corruption or out-
Unsustainable Government Policies
right conflict. Here fighting causes erosion
of property rights (higher 1 − ϕ = Nfi /F), Natural resource wealth may encourage
which in turn induces even more fighting, countries to engage in “excessive” borrow-
especially if “corruption culture” is strong ing, which harms the economy in the short
(low F). Both the “rapacious rent seeking” and long run (Arman Mansoorian 1991).
and the “erosion of property rights” effect is Heavy borrowing on the world market
stronger if there are more rival factions (high induces depreciation of the real exchange
N). Fighting implies that there are fewer rate in the long run. In an economy with
resources available for productive activities; overlapping generations of households with-
hence, utility of each group is lower. If the out a bequest motive, the generations alive at
country is homogenous (N = 1), there is no the time of the exploitation of the resource
fighting and no undermining of property borrow against future resource income and
rights so that a natural resource bonanza future generations bear the burden of servic-
always benefits consumption of its citizens. ing the debt. The consequent fall of aggre-
Indeed, empirical evidence suggests that the gate demand causes depreciation of the real
resource curse is more severe in countries exchange rate in the long run. Others also
that have many ethnic or religious factions find that resource rich countries have an
and many languages (Hodler 2006; van der incentive to borrow excessively (Manzano
Ploeg and Poelhekke 2009). and Rigobon 2001).
Caselli and Wilbur John Coleman (2006) In general, a sudden resource bonanza
provide a richer theory of coalitions formed tends to erode critical faculties of politicians
along ethnic lines competing for natural and induce a false sense of security. This
resources. In ethnically homogenous soci- encourages them to invest in projects that
eties, members of the losing coalition can are unnecessary, keep bad policies in force,
defect to winners at low cost, which rules and dress up the welfare state so that it is
out conflict as an equilibrium outcome. Of impossible to finance once natural resource
course, rent of each member of the win- revenues dry up. Politicians are likely to
ning coalition is diluted. In ethnically het- lose sight of growth-promoting policies, free
erogeneous societies, members of winning trade, and “value for money” management.
coalitions more easily recognize potential For example, after the discovery of natural
infiltrators by skin color or other physical gas in the Netherlands, the global oil price
characteristics and exclude them. We should shocks during the 1970s and 1980s and the
therefore see more conflict in ethnically het- consequent sharp rise in unemployment,
erogeneous societies such as Rwanda, Sudan, successive Dutch governments responded
irresponsibly. They expanded public employ- and deploy resources, enforce property
ment and consumption, made unemploy- rights, and resist demands of rent seekers.
ment and disability benefits more generous,
weakened eligibility conditions for benefits,
4. Why Do Many Resource Rich
raised the minimum wage, and implemented
Developing Countries Experience
protective labor market legislation (Neary
Negative Saving?
and van Wijnbergen 1986). Starting in 1989,
it has taken more than twenty years to put Section 3 has put forward eight important
the Dutch welfare state on a financially sus- hypotheses on how natural resources affect
tainable footing again. the economy, institutions, rent seeking, con-
Many developing countries erred by try- flict, and policy. Here I put forward two fur-
ing in vain to encourage industrialization ther hypotheses to explain the stylized fact
through prolonged import substitution using discussed in section 2.4 that many resource
tariffs, import quota, and subsidies for manu- rich developing countries are unable to fully
facturing. Neo-Marxist policymakers in these transform their large stocks of natural wealth
countries, but also many other economists into other forms of wealth. To set the scene,
during the 1970s and 1980s, found inspira- section 4.1 discusses the Hotelling rule for
tion from the Prebisch-Singer hypothesis, optimal intertemporal depletion of natural
namely the secular decline of world prices resources and the resulting utilitarian out-
of primary exports (David I. Harvey et al. come for transforming depleting exhaust-
2010), to attempt to avoid resource depen- ible natural resource assets into financial
dency through state-led industrialization and capital in a small open economy. I suppose
import substitution. These policies may also throughout that countries have some power
have been a reaction to the appreciation of on the market for natural resources but are
the real exchange rate and the decline of price takers in all other markets. Section 4.2
the traded manufacturing sectors caused adopts a Rawlsian max–min social welfare
by natural resource wealth. The substan- perspective to discuss the optimal level of
tial resource wealth in many of those coun- sustainable consumption and the Hartwick
tries may thus have prolonged bad policies. rule for reinvesting resource rents into dura-
Political scientists have advanced several ble, nonexhaustible assets. It also offers some
reasons why states have a proclivity to adopt evidence that many resource rich countries
and maintain suboptimal policies (e.g., Ross experience negative genuine saving. Section
1999). Cognitive theories blame policy fail- 4.3 then puts forward the “anticipation of
ures on short-sightedness of state actors who better times” hypothesis, which suggests
fail to take account of the adverse effects of that resource rich countries should borrow
their actions on generations that come after in anticipation of higher world prices for
the resource is exhausted, thus leading to resources and improvements in extraction
myopic sloth and exuberance. These cog- technology in the future. Section 4.4 puts
nitive theories also stress a get-rich-quick forward the “rapacious extraction” hypoth-
mentality among businessmen and a boom- esis to explain how, in absence of effective
and-bust psychology among policymakers. government intervention, conflict among
Political scientists point the finger at abuse rival factions induces excessive resource
of resource wealth by privileged classes, sec- extraction and investment and negative
tors, client networks, and interest groups. genuine saving when there is wasteful rent
They also emphasize the rentier state and seeking, investment in “white elephants” and
fault a state’s institutional weakness to extract short-sighted politicians.
4.1 Preamble: Optimal Conversion of depletion path of natural resources are set
Depleting Natural Resources into so that reserves are eventually completely
Foreign Assets exhausted. A resource discovery thus leads
to an immediate fall in the resource price
Most discussions of the resource curse and increase in the rate of resource deple-
and Dutch disease take the windfall as tion. Suppose world demand for resources
manna from heaven. Earlier literature, is given by E = E(), where  is the price
however, deals with the optimal intertem- of natural resources and ε ≡ − E′/E > 1
poral depletion of exhaustible resources the constant elasticity of demand. The social
(e.g., Partha Dasgupta and Geoffrey M. planner maximizes utilitarian social welfare,
Heal 1979). The main result is the Hotelling ∫
0∞
    U(C(t)) exp (−ρt) dt, subject to the
rule, which states that the rate of increase in equations describing natural resource
the marginal rent of resources must equal depletion, the dynamics of the current
the world interest rate (possibly including a account and the Cobb–Douglas production
risk premium). With no extraction costs and function, i.e.,
constant elasticity of demand for resources,
the Hotelling rule states that the capital  ·
  = −E − R,
  
S
gain on resources must equal the world
interest rate. This is based on the arbitrage  ·
  = r (A − K) + Y + E() − C and
  
A
principle, which says that one should be
indifferent between keeping the resource
under the ground (in which case the return Y = F(K, R) = KαRβ,
is the capital gain on reserves) and extract-
ing, selling, and getting a market return on where C, S, R, A, K, Y, r, and ρ denote con-
it. The rate of increase in marginal resource sumption, the resource stock, resource use
rents should thus equal the world interest in production, national assets, the capital
rate. Since marginal extraction costs differ stock, domestic production, the exogenous
widely across countries, optimal depletion world interest rate, and the subjective rate of
rates vary widely as well even if each coun- time preference, respectively. The produc-
try is a price taker. tion function has decreasing returns to scale
Consider the optimal conversion of deplet- with respect to K and R (0 < α + β < 1). It
ing exhaustible resources into foreign assets follows that:
for a small open economy that uses capi-
tal and resources in production, obtains an  ·
FK = r, FR = (1 − 1/ε),     / = r,
exogenous return on investment abroad, and
faces elastic demand for its resources on the  ·  ·
 E / E = −εr,  C / C = σ(r − ρ),
global market (Dasgupta, Robert Eastwood,
and Heal 1978). Maximizing social welfare
yields the Hotelling rule and the efficiency where σ is the elasticity of intertemporal
conditions that the marginal product of substitution. The first and second equation
capital must equal the world interest rate equate the marginal products of capital and
and that of resources the world price of resource to the interest rate and the mar-
resources. The optimal rate of resource ginal revenue of natural resources, the third
depletion thus equals the elasticity of world equation is the Hotelling rule that (given that
demand for its resources times the inter- demand for resources is iso-elastic) says that
est rate. The initial price and the resulting capital gains on natural resources must equal
the rate of interest r, and the fourth equa- national wealth constraint, one gets initial
tion is the Keynes–Ramsey rule. Effectively, consumption:
the first three equations result from maxi-
mizing the present value of natural resource C(0) = [(1 − σ) r + σρ]
( (  )(  )
and other income while the fourth equation
× A(0) + _
results from choosing the timing of consump- 1 − α − β _ Y(0)
   
 
r    − K(0)
tion to maximize utility. It follows that capital 1 − α
)
and natural resource use must decline over
+  _
time: Q(0) E(0)
εr    
,
    =  K / K = − _
 ·
 
/Y
Y
 ·
( β
    
1 − α − β )
r < 0, where Y(0), K(0), and (0) directly follow
from E(0) and R(0). Since Y(0) − rK(0)
(  )
equals W(0) + (1 − 1/ε)(0)E(0) and the
    = − _
 · 1 − α 
 
/R
R   
r < 0. wage W grows at the same rate as output,
1 − α − β
households consume a constant fraction of
the sum of financial, human, and natural
Substituting the rates of decline of E and resource wealth. Clearly, a higher σ or lower
R into the resource depletion equation, ρ boosts consumption growth but lowers
using the marginal factor productivity con- C(0). The expression for initial consumption
ditions, and integrating over time gives two holds for all instants of time. Natural resource
equations relating E(0) and R(0) to S(0): wealth (E/rε) declines over time at the rate
(ε − 1)r and human wealth declines at the
rate β r/(1 − α − β). Hence, supposing that
_ E(0) __
(1 − α − β)R(0)
 +    
  εr         = S(0) r = ρ, a constant level of consumption can
(1 − α) r be sustained by accumulating sufficient for-
eign assets, A, to compensate for the contin-
and R(0)−(1−α−β) ually declining levels of natural resource and
human wealth. It is easy to extend the results
(  ) [E ( )]
α to a return that declines with the level of for-
= _ (E(0)) 1 −  _
1−α
α   
r −1 1   
ε . eign investment, e.g, r = r (A − K), r′ < 0,
or to allow for some uncertain date in the
future at which prices fall due to invention of
some new alternative technology or source
Hence, a resource bonanza (higher S(0)) of resources (Dasgupta, Eastwood, and Heal
lifts the declining paths of resource use 1978).
and resource exports up (higher R(0) These pioneering insights on optimally
and E(0)), also lifts the declining paths of converting natural resources into financial
capital and production, and depresses the assets have not yet been extended to take
price trajectory (higher Q(0)). Since the account of resource windfalls harming com-
optimum production in this small open petitiveness, provoking corruption, rent seek-
economy depends only on world prices, ing, and other distortions.14 But analysis of a
the optimal trajectories of E, R, K, and Y
are independent of consumer preferences
(σ and ρ). Substituting these together with
14 Appendix 3 shows how to optimally convert deplet-
ing exhaustible resources into physical capital for a closed
the Euler equation into the present-value economy.
basic Dutch disease model without capital One feedback Nash outcome is an initial
accumulation albeit with learning by doing phase where the monopolist sets prices low
indicates that, after a windfall, one should enough to exhaust the fringe and a final
gradually adjust the optimal share of national phase where the monopolist enjoys higher
wealth consumed downward and accept monopoly profits; the price at the end of the
some adverse Dutch disease effects (Egil first phase is then not high enough to incite
Matsen and Torvik 2005). Lower growth in the fringe to postpone extraction (David
resource rich economies may thus be part M. G. Newbery 1981; Fons Groot, Cees
of an optimal growth path. The challenge is Withagen, and Aart de Zeeuw 2003).
to reinvestigate these issues for a dependent
4.2 Genuine Saving and the Wealth of
economy with capital accumulation, specific
Nations: A Pragmatic Guide
factors or intrasectoral adjustment costs, and
learning by doing, where wages and capital The Hartwick rule of investing all resource
intensities are not fixed by technology and rents in other forms of capital provides a
the world interest rate. pragmatic guide for sustainable develop-
Nonrenewable or exhaustible resources ment. Genuine saving is the traditional con-
typically imply steady declines in income cept of net saving, namely public and private
per capita. With environmental resources saving minus depreciation of public and pri-
being a production factor and production vate investment, plus current spending on
displaying constant returns to scale, capi- education to capture the change in intangi-
tal and labor run into jointly diminishing ble (human) wealth, minus the value of net
returns (William D. Nordhaus 1992). In the depletion of exhaustible natural resources
AK-model of endogenous growth with nat- and renewable resources (forests), minus
ural resources as production factor, a posi- damages of stock pollutants (carbon diox-
tive growth rate of consumption cannot be ide and particulate matter) (Kirk Hamilton
sustained forever either (Aghion and Peter and Michael Clemens 1999; Hamilton and
Howitt 1998, chapter 5). Faster popula- John M. Hartwick 2005). Alas, fisheries, dia-
tion growth increases pressure on the finite monds, subsoil water, and soil erosion are not
resource and thus reduces per capita growth. dealt with due to data problems. With posi-
However, resources such as fisheries, forests, tive genuine saving, a nation becomes richer
and agricultural land are renewable. This and social welfare increases, and with nega-
raises questions about how a limited renew- tive genuine saving, a nation loses wealth
able resource sector can coexist with a grow- and social welfare falls (Dasgupta and Karl-
ing sector in balanced growth equilibrium. Goran Mäler 2000). Wealth per capita is the
Typically, this requires technological progress correct measure of social welfare if the popu-
in use of the resource to be sufficiently faster lation growth rate is constant, per capita con-
than in use of other inputs. If proper account sumption is independent of population size,
is taken of renewable resources, ongoing production has constant returns to scale, and
growth is feasible (e.g., A. Lans Bovenberg current saving is the present value of future
and Sjak A. Smulders 1996; Ludvik Elíasson changes in consumption (Dasgupta 2001a).
and Turnovsky 2004). Genuine saving estimates calculated by
The literature on optimal oil exploitation World Bank (2006), based on Giles Atkinson
pays ample attention to the market structure and Hamilton (2003), presented in figure 4
of oil producers. Typically, the monopolist show an alarming picture. Countries with a
OPEC is considered together with a compet- large percentage of mineral and energy rents
itive fringe of price-following oil producers. of GNI typically have lower genuine saving
40 –
30 –
Genuine saving percent GNI
20 –
10 –
0 –
−10 –
−20 –
−30 –
−40 –
−50 –
–
–
–
–
0 10 20 30 40 50 60 70
Mineral and energy rents percent GNI
Figure 4. Genuine Saving and Exhaustible Resource Share

Source: World Bank 2006, figure 3.4.
rates. This means that many resource rich Figure 5 calculates by how much produc-
countries become poorer each year despite tive capital would increase by 2000 if coun-
the presence of large natural resources. tries would have invested their rents from
They do not fully reinvest their resources crude oil, natural gas, coal, bauxite, copper,
at the expense of future generations by not gold, iron, lead, nickel, phosphate, silver, and
investing in intangible or productive wealth. zinc in productive capital since 1970. The
For example, Venezuela combines negative calculations provide an upper bound since
economic growth with negative genuine they abstract from marginal extraction costs
saving while Botswana, Ghana, and China due to data problems. High resource depen-
with positive genuine rates enjoy substantial dence is defined as minimally a 5 percent
growth in the year 2003. Highly resource share of resource rents in GDP. We see that
dependent Nigeria and Angola have genu- resource rich countries with negative genu-
ine saving rates of minus 30 percent, which ine saving, such as Nigeria or Venezuela,
impoverishes future generations despite could have boosted their nonresource capi-
having some GDP growth. The oil/gas states tal stocks by a factor of five or four if the
of Azerbaijan, Kazakhstan, Uzbekistan, Hartwick rule would have been followed.
Turkmenistan, and the Russian Federation This is also true for oil/gas rich Trinidad
all have negative genuine saving rates; they and Tobago and copper rich Zambia. All the
seem to be consuming or wasting rather than countries in the top right quadrant (except
reinvesting their natural resource rents. Trinidad and Tobago) have experienced
400 –
Low High
350 – resource resource • Nigeria
dependence dependence
300 – • Zambia
Percent increase in produced capital if standard
• Venezuela
250 –
200 –
Hartwick rule followed
• Trinidad and Tobago

150 – • Guyana
100 –
• Bolivia • Mauritania
• Ecuador
• South Africa • Gabon
50 – Jamaica Algeria • • Congo Low
• Ghana • Peru capital
0 –
• Zimbabwe Mexico accum.
• Chile • Indonesia High
Egypt
−50 – • India capital
• Brazil Malaysia • China accum.
−100 –
• Thailand
Korea
−150 –
–
–
–
–
–
0 5 10 15 20 25 30 35
Percent share of resource rents in GDP (average 1970–2000)
Figure 5. Counterfactual Hartwick Rule

Source: World Bank 2006, figure 4.1.
declines in per capita income from 1970 their resource rents.15 Sub-Saharan Africa
to 2000. If the Hartwick rule would have has high population growth and shows sub-
been followed during the last few decades, stantial saving gaps typically of 10 to 50 per-
these economies would have been much less cent of GNP. For Congo and Nigeria, the
dependent on oil and other resources than saving gaps are as high as 110 percent and 71
they are. percent, respectively.
The Solow–Swan neoclassical model of Even countries that save a large part
economic growth predicts that countries of their natural resource wealth can fare
with high population growth have lower badly. An early influential study found that
capital intensities and thus lower income about half of the windfall income of six oil-
per capita. Similarly, in countries with high producing countries (Algeria, Ecuador,
population growth rates, genuine saving can Indonesia, Nigeria, Trinidad and Tobago,
be positive while wealth per capita declines
(World Bank 2006, table 5.2). Such countries
15 Positive population growth gives a negative shadow
are on a treadmill and need to create new
price of time and thus positive genuine saving; technical
wealth to maintain existing levels of wealth progress gives a positive shadow price of time and negative
per capita. They thus need to save more than genuine saving (Y. Hossein Farzin 2010).
and Venezuela) after the oil price hikes of equals E = E(/*) with ε ≡ − E′()/E
1973 and 1979 was invested domestically > 1, where * is the world price of oil sold
(overwhelmingly by the public sector), but by its competitors. Saving of the nation is
that nevertheless all countries experienced given by
prolonged periods of real exchange appre-  ·
ciation and negative growth (Gelb 1988).  A   = r (A − K) + [ E − TC(E)] + f (K) − C.
The roots of this puzzling feature may be
investment in socially undesirable public The initial stock of oil S0 defines the maxi-
investment projects or “white elephants” mum amount of oil that can be depleted:
∫ 
(Robinson and Torvik 2005) and high popu- ∞
lation growth. I show in the next section that,  ·
    = − E, S(0) = S0 or     E(t) dt = S0.
S
even without such “white elephants,” it may 0
not be optimal for such countries to save less
than their resource rents if world resource There are two efficiency conditions:
prices are expected to increase and improve-
ments in exploration technology are antici- f ′(K) = r and
pated in the future.
d[(1 − ε−1) − TC′(E)]/dt
 ___
         = r.
4.3 Anticipation of Better Times Can (1 − ε−1) − TC′(E)
Induce Negative Genuine Saving
The first one states that the marginal product
Consider a small resource-exporting econ- of capital is set to the interest charge. The
omy that takes the world interest and world second requires that the marginal resource
price for its final products as given but exerts rents must increase at a rate equal to the
some monopoly power on the world market world interest rate. An anticipated positive
for natural resources. I investigate max–min rate of increase in the world resource price
social welfare and investigate what needs or in the rate of technical progress in extrac-
to be done to sustain a constant level of per tion technology thus induces resource deple-
capita consumption.16 To do this, suppose tion to be postponed:
that there is no use of exhaustible resources
 ·
in production and no population growth. _          = [(1 + μ)π + μτ − r]/εE μ,
  E
The production function is f (K) with f ′ > 0
(   )
E
and f ′′ < 0, and there is no depreciation of  ·  ·
the capital stock. The cost of extracting E where τ ≡ − _ T      ≥ 0, π ≡  _
    
Q
   ,
units of resources is TC(E) with C′ > 0 and T Q
C′′ ≥ 0, where a fall in T indicates a boost to
extraction productivity. World demand for oil __ AC′(E)
    > 0
μ ≡    
(1 − ε−1) − AC′(E)
and εE ≡  _

EC′′(E)
 > 0.
 ·
16 Although the Keynes–Ramsey rule, C    / C  = σ(r − ρ),    
suggests that it is feasible to sustain a constant level of C′(E)
per capita consumption in the small resource-exporting
economy even if there is no max–min welfare (i.e., σ ≠ 0)
provided that r* = ρ, this is not the case for the closed With exogenous continual improvements
economy. The optimal path for the closed economy first in extraction technology (τ > 0), it pays
has per capita consumption rising and then falling and
vanishing asymptotically; the first phase may not occur
to delay depletion of reserves to reap the
(Dasgupta and Heal 1979, chapter 10 and appendix 3). benefits of technical progress. The rate of
 ·
increase in the price of resources and the  E / E  = −ε(r − π* )and the max–min saving
·
rate of change in resource depletion are then rule becomes A     = (1 − 1/ε)E − π* S.
reduced · even further. Now consider the case Saving marginal resource rents minus
π* ≡   *  /* > 0. This pushes up the rate of imputed interest on the value of natural
increase in the price of resources charged by resource reserves thus sustains a constant
the country and postpones resource deple- level of consumption. Countries with abun-
tion. I assume r − π* > μ(π* + τ), so π > π* dant reserves of exhaustible resources should
and reserves are not exhausted in finite thus run a current account deficit if resource
time. With a constant r* and π* and no costs rents fall short of the imputed rent on the
of extraction, one has E(0) = ε(π − π*)S0 value of resource reserves.  · Genuine sav-  ·
= ε(r − π*)S0. Reserves are exhausted rela- ing is thus negative, i.e.,  A   + (1 − ε−1)   
 
S
tively slowly if the world interest rate is low = −π* S < 0.
and the world rate of increase in the world Since the country saves less than its mar-
price is high. In general, this is also true if the ginal resource rents and postpones extrac-
rate of technical improvements in exploration of exhaustible resources if it expects
tion technology is high. Sustaining the max– extraction technology to continually improve
min level of constant consumption requires: or the price it can fetch for its resources
 · to continually increase in the future (cf.,
 A (  t) = [(t)(1 − ε−1) − A(t)C′(E(t))]E(t) Geir B. Asheim 1986; Jeffrey R. Vincent,
( ∫  )
Theodore Panayotou, and Hartwick 1997;
∫ 
∞ s van der Ploeg 2010b),17 it is a priori unclear
−       exp −      r (v) dv whether observed negative genuine saving
t t
for resource-rich economies are due to poor
× [τ (s)T(s)C(E(s)) + π (s)(s)E(s) institutions, badly functioning capital mar-
kets, corruption, or mismanagement or due
+   r ·  (s)(A(s) − K(s))] ds. to anticipation of better times. It is optimal
for a country with substantial oil reserves
to save less than a country with almost no
This saving rule extends the Hartwick rule reserves because it makes sense to sell more
to an open economy. The first term says that of its reserves in the future when the price of
the nation saves the marginal resource rents oil is higher.
valued at the world resource price minus mar-
4.3.1 The Hartwick Rule in the
ginal extraction costs, so depletion of natural
Global Economy
resource reserves must be compensated by
increases in foreign assets. The second term To examine the Hartwick rule for the
is the “anticipation of better times” term. It global economy, consider a world consisting
says that the nation saves less if it expects the of natural resource (say, oil) exporters and oil
world interest rate (provided A > K) or the importers. With free international trade in
price of its resources to increase in the future. oil and goods, perfect capital m obility, zero
The country then saves less and postpones
extraction. The nation also saves less if it 17 Similar arguments can be applied to deforestation in
expects positive technical progress in future a small open economy with a large endowment of forest
oil extraction technology. A special case arises land and small endowment of agricultural land (Hartwick,
if extraction costs are zero and the world Ngo Van Long, and Huilan Tian 2001). The early phases
of clearing forest land are then governed by the high price
price of resources follows the Hotelling rule of agriculture while later phases are driven by profits from
because then the depletion rate is given by marketing timber from cleared land.
labor mobility, no technical progress, no 4.4 Fractionalization, Voracious Depletion,

population growth, and identical technolo- Excessive Investment, and Genuine
gies for both blocks, the max–min egalitar- Saving
ian outcome can be characterized (Asheim
1986, 1996). Factor intensities are deter- Section 4.3 advanced the hypothesis that
mined by the world interest rate and price resource rich economies may not save all of
of oil. The ratios of output, capital, and their resource rents in anticipation of better
resource use in oil-exporting economies rel- times (e.g., higher rate of increase in the prices
ative to those in oil-importing economies are for its resource products or ongoing technical
then identical and equal to the ratio of the progress in resource extraction). An alterna-
labor force of oil exporters relative to that of tive hypothesis is that resource rich countries
oil importers. have to contend with rival factions competing
On the efficient max–min path, oil export- for natural resource rents. The modern polit-
ers consume the full marginal product of ical economy of macroeconomics literature,
their human capital plus their oil rents, but surveyed in Persson and Tabellini (2000),
consume only a fraction of the marginal abstracts from the intertemporal aspects of
product of physical capital and the remain- natural resource depletion but is of obvious
der is used to accumulate national wealth to relevance to the crucial question of why coun-
compensate for the decreasing rate of return tries seem to be impatient and do not reinvest
on capital. This fraction equals one minus all their resource rents. This literature high-
the ratio of the share of resource rents to the lights deficit biases in the absence of a strong
share of capital income in value added. Oil minister of finance due to government debt
importers consume less since they have no being a common pool (Andres Velasco 1999),
oil rents. If oil exporters owned all physical debt biases if political parties have partisan
capital, they would be investing all oil rents preferences over public goods and the prob-
in physical capital. They would then use all ability of removing the government from
natural resource rents for consumption and office is high (Alberto Alesina and Tabellini
run a foreign financial debt. Oil has no mar- 1990), and delayed stabilization resulting
ginal productivity as a stock but oil exporters from different groups in a “war of attrition”
can consume a fraction of the capital gains. attempting to shift the burden of higher taxes
Oil exporters can thus indefinitely sustain or spending cuts to other groups (Alesina and
positive consumption by consuming only a Allan Drazen 1991). A common insight of
fraction of their resource rents, especially this literature is that the rate of discount used
if these are large relative to capital income. by politicians may be higher than the rate of
Since the Hotelling rule implies that oil interest by, for example, the probability of
exporters enjoy a growing income from oil being removed from office. Indeed, if a fac-
revenues over time, they need to save less tion worries it may not be in office in the near
than the Hartwick rule to keep consumption future, it will extract natural resources much
constant. Conversely, oil importers need to faster than is socially optimal and will bor-
save more to afford the increasing cost of oil row against future resource income (or accu-
imports and sustain a constant level of con- mulate less assets than is socially optimal) in
sumption. Resource rich economies thus order to gain at the expense of future succes-
sustain consumption by consuming a frac- sors. This could show up as capital flight and
tion of their marginal resource rents. Alas, no higher private consumption for the faction in
empirical tests of this proposition are avail- power or higher public spending of the type
able yet. that primarily benefits those in power.
There are no studies available yet that Tornell (1996) and Tornell and Lane (1999),
attempt to apply these political economy I make a distinction between uncontested
insights to a formal model addressing the stocks of foreign assets (bonds and capital)
optimal depletion of natural resources. This and contested stocks of natural resources.20
is an interesting area for further research and Furthermore, extraction costs are zero, the
some of these issues are discussed in section production function of each group is given
5 that deals with optimal harnessing of given by f (Ki) = K  αi  (1/N)1−α, Q(·) = E−1(·), and
natural resource windfalls. Here I offer, as a the
 · saving equation ofN each faction equals
first step, a simple analysis of how common-  A  i = r (Ai − Ki) + Q(∑ j=1     Ej) Ei + f (Ki) − Ci.
pool problems induce competing factions to Resources of each competing faction are
use a discount rate greater than the inter- perfect substitutes in demand. A homog-
est rate in the Hotelling rule which in turn enous society with perfect property  · rights
leads to voracious natural resource deple- has the usual Hotelling rule     / = r.
tion, excessive investment rates, less build- The Hotelling rule  · under fractionalization
up of foreign assets, and lower consumption (N > 1) becomes    / = r + ξ(N − 1) > r,
than is socially optimal for the small open so resource prices rise faster than the rate
economy of section 4.3.18,19 This analysis also of interest if there are factions contest-
allows me to illustrate how natural resources ing resource rents, seepage is strong, or
are gradually transformed into foreign assets. property rights imperfect. As a conse-
Although I do not offer a full political econ- quence of the higher discount rate used
omy analysis, I do clarify some conceptual by competing factions, resource extrac-
issues to do with measuring genuine saving tion is more voracious and the rate of
in noncompetitive environments and suggest decline of natural resource revenues,
·
that World Bank figures may underestimate ( E)/(E)  = −(ε − 1)[r  + ξ(N  − 1)],
genuine saving. is higher in more fractionalized societ-
The dynamics of the stock of natural ies. Although fractionalized societies save
resources
 · owned by each faction i is given a greater fraction
 · of their natural resource
by  S i = −Ei + ∑ j≠i
    ξ(Ej − Ei), Si(0) = Si0.    / E = [1 + (ξ(N − 1)/(εr +

revenues, A
Here ξ > 0 indicates the speed by which (ε − 1)× ξ(N − 1)))] (1 − (1/ε)) ≥ 1 − (1/ε)
oil, gas or water seeps from one field to they end up with less wealth in the long run.
another or the degree of imperfection of Hence, SWF ≡ limt→∞ A(t) − A0 = [1 +
property rights on natural resources (van (ξ(N − 1)/(εr + (ε − 1)ξ(N − 1)))] Q(ε(r +
der Ploeg 2010c). No seepage (as is the ξ(N − 1))) S0 < Q(εr) S0 if N > 1, espe-
case for gold, silver, or diamonds) or per- cially if N is large and world demand for
fect property rights corresponds to ξ = 0. resources is more elastic. The sustainable
In general, we have ξ > 0. As in Lane and level of consumption equals interest on ini-
tial foreign assets plus wage income plus
interest on accumulated wealth, C = rA0 +
(1 − α)(α/r)α/(1−α) + rSWF, and is thus
18 With a max–min specification of social welfare
(σ = 0), consumption of each faction will be constant
over time. With a positive elasticity of intertemporal sub- lower in fractionalized societies where
stitution (σ > 0), short-sighted factions induce excessive resources suffer from weak property rights
resource extraction and less accumulation of foreign assets
that will lead to a bias toward higher consumption in the
short run and lower consumption in the long run.  ·
19 This can also be shown for a closed economy with 20 The Keynes–Ramsey rule is again C   / C  = σ(r − ρ)so
capital accumulation where the subgame-perfect Nash that the rate of growth in consumption is not affected by
equilibrium yields a suboptimally low level of sustainable conflict among factions. The analysis focuses on Rawlsian
consumption (van der Ploeg 2010b). max–min outcomes (σ = 0).
and seepage. It is thus optimal to gradually is higher and is closer to the world price
transform natural resource reserves into of resources. Estimates of genuine saving
interest-earning foreign assets. The wealth should use accounting prices; if they use
of the state, i.e., sovereign wealth, gradually marginal revenues, they yield a too optimis-
grows from A0 to A0 + SWF. The final level tic estimate, and if they use market prices of
of accumulated foreign assets in a fraction- resources, they yield a too pessimistic esti-
alized society is less than in a homogenous mate of genuine saving in fractionalized soci-
society despite the lower initial price of natu- eties. Using true · accounting prices,
 · genuine
 ·
ral resources. Also, the speed of transforma- saving is zero,  A (  0) + A(0) S (0) =  A ( 0) −
tion is faster in a fractionalized society. It is A(0)E(0) = 0, even though the struggle
the interest earned on sovereign wealth that over resources depresses consumption and
makes up for dissipating resource revenues welfare. Effectively, both resource extraction
and thus makes it possible to sustain con- and investment in foreign assets occur at a
stant consumption as resources are depleted. rate that is from a social perspective too high,
Comparing the market with the socially opti- thereby leaving genuine saving unaffected.
mal outcome suggests that benevolent gov- Interestingly, rapacious rent seeking in
ernments redistribute from the lucky cohort itself does not explain the observed nega-
that discovers resources to later cohorts by tive genuine saving rates of many developing
bequeathing them a large stock of foreign resource rich countries (unless erroneously
assets. market rather than accounting prices are
If there is an imperfect mechanism used to calculated genuine saving—a data
for resource allocation, one must use the artifact). The “anticipation of better times”
true accounting prices A when calculat- hypothesis (see section 4.4) helps to explain
ing genuine saving (Dasgupta and Mäler observed negative genuine saving but a
2000; Dasgupta 2001b; Kenneth J. Arrow, deeper analysis of the political distortions of
Dasgupta, and Mäler 2003). These are the rapacious rent seeking should offer a better
effect of a marginal increase in the initial explanation. Countries with a lot of fighting
stock of resources on social welfare divided about natural resources suffer from corrup-
by the effect of a marginal increase in initial tion and erosion of the quality of the legal
foreign assets on social welfare. In the pres- system, thus discouraging saving and invest-
ent context, this amounts to: ment in productive capital (see section 3.7),
(1 −  _1ε   ) (0) ≤  (0) ≡  _
may overinvest in public investment projects
∂ C/∂ S
0
  
  as an inefficient form of distribution to the
∂ C/∂ A
A
0 own group members as they are not so obvi-
[  ]
ε[r + ξ(N − 1)] ously corrupt within the context of a dynamic
∂ SWF
=  _  
   = __
   
   citizen–candidate model for a representative
∂ S0 εr + (ε − 1) ξ(N − 1)
democracy (cf., Besley and Stephen Coate
( )
1997, 1998), may overinvest in public invest-
× 1 −  _
1    (0) ≤ (0).
ε ment projects with negative social surplus
(“white elephants”) as a form of credible
redistribution as all politicians can commit to
In societies that are homogenous or have socially efficient public investment projects
perfect property rights, the accounting (cf., Robinson and Torvik 2005), and often
price equals marginal resource revenue. In attract short-sighted politicians. If added to
fractionalized societies with insecure prop- my explanation of voracious resource deple-
erty rights, however, the accounting price tion and excessive investment, these features
∆C – Revenue flow
– Bird-in-hand
–
–
–
PIH
–
–
–
–
–
Developing
–
–
–
–
T0 T1 t
Figure 6. Alternative Prescriptions for Harnessing Natural Resource Windfalls

Note: The incremental consumption path indicated by “Developing” is the optimal path obtained by maximiz-
ing social welfare for a developing economy which suffers capital scarcity and has to pay an interest premium
on its outstanding foreign debt.
Source: Collier et al. 2010.
should give a realistic explanation of the neg- benchmark for harnessing such a windfall
ative genuine saving rates observed in many is based on the permanent income hypoth-
developing resource rich economies. esis, which says that countries should bor-
row ahead of the windfall, pay back incurred
debt, and build up sovereign wealth dur-
5. Harnessing Natural Resource Windfalls
ing the windfall and finance the permanent
in Developing Economies
increase in consumption out of the interest
Despite the normative and political analy- on the accumulated sovereign wealth after
ses of converting depleting natural resources the windfall has ceased. Indeed, the IMF has
into productive assets discussed in section 4, often recommended resource rich countries
there are good technical reasons to pump oil to put their windfalls in a sovereign wealth
as fast as possible out of the ground once a fund (e.g., Jeffrey Davis et al. 2001). Figure
field has been opened. So it may be better 6 shows how the permanent income hypoth-
to focus at the optimal way of harnessing a esis and the consequent building up of such
given windfall (e.g., Collier et al. 2010). Such a fund are used to optimally harness unantic-
windfalls are typically anticipated (five years ipated windfalls. In practice countries such
or so) and temporary (say, twenty years). The as Norway prefer to restrict incremental
consumption to interest earned on the fund saving should be directed toward accumu-
and not to use the windfall until it is banked, lating of domestic private and public capital
which gives the conservative bird-in-hand and cutting debt rather than accumulating
rule. Estimation of fiscal reaction functions foreign assets (van der Ploeg and Venables
for non-hydrocarbon tax and public spend- 2010). The resulting optimal micro-founded
ing using official projections for hydrocar- path for incremental consumption is given in
bon revenues and the pension burden for figure 6. Effectively, the windfall brings for-
Norway suggests that fiscal reactions have ward the development path of the economy.
been partially forward-looking with respect Although the hypothesis of learning-by-
to the pension bill, but indeed not with doing in the traded sector may be relevant for
respect to hydrocarbon revenues (Harding advanced industrialized economies, devel-
and van der Ploeg 2009). The primary non- oping economies are more likely to suffer
hydrocarbon deficit should according to the from absorption constraints in the nontraded
permanent income hypothesis react only to sector especially as it is unlikely that capital
permanent oil/gas revenues, but in practice in the traded sector can easily be unbolted
it also reacts to current revenues. This sug- and shunted to the nontraded sector. This
gests that Norway has used the bird-in-hand cuts the other way, since it is then optimal
rule rather than the permanent income rule. to temporarily park some of the windfall in
VAR analysis of a DSGE model of oil-rich a sovereign wealth fund until the nontraded
economies with a traded and nontraded sec- sector has produced enough home-grown
tor suggests that the fiscal rules of Mexico capital (infrastructure, teachers, nurses, etc.)
and Norway with respectively a small and big to alleviate absorption bottlenecks and allow
emphasis on saving windfalls can explain the a gradual rise in consumption (see appen-
Mexican hump-shaped impulse responses dix 4). The economy experiences temporary
for output, the real exchange rate, and pri- appreciation of the real exchange rate and
vate consumption and the flat responses for other Dutch disease symptoms. However,
Norway (Anamaría Pieschacón 2009). More these are reversed as home-grown capital is
DSGE work is needed on resource rich accumulated.
economies, also paying attention to monetary There are many other resource manage-
policy rules, sterilization of foreign exchange ment issues. First, governments should real-
windfalls, and unemployment in the light of ize that, if imports are mostly financed by an
natural resource windfalls. exogenous stream of foreign exchange com-
One must take account of the special fea- ing from resource rents, revenue generated
tures of resource rich developing countries. by tariffs is illusory as the increase in tariff
Many of them are converging on a develop- revenue is offset by reducing real resource
ment path, suffer capital scarcity and high revenue (Collier and Venables forthcom-
interest rates resulting from premium on ing). Tariffs effectively reduce the domestic
high levels of foreign debt, and households purchasing power of the windfall of foreign
do not have access to perfect capital mar- exchange. Second, tax capacity typically
kets. In that case, the permanent income erodes quickly during windfalls. Since legal
hypothesis is inappropriate. In contrast to and fiscal capacity are likely to be comple-
transferring much of the increment to future ments (Besley and Persson 2009), this leads
generations (as with the permanent-income to grave concerns about the adequate sup-
and bird-in-hand rules), the optimal time ply of common-interest public goods such as
path for incremental consumption should fighting external wars or inclusive political
be skewed toward present generations and institutions. Third, the political economy of
windfalls dictates that incumbents may avoid more panel-data and quasi-experimental
putting resource revenues in a liquid sover- studies are needed to shed light on this key
eign wealth fund that can be easily raided by issue. The best available empirical evidence
political rivals. There is thus a bias to exces- suggests that countries with a large share of
sive investment in illiquid, partisan projects, primary exports in GNP have bad growth
especially if the probability of being kicked records and high inequality, especially if
out of office is high (Collier et al. 2010). quality of institutions, rule of law, and cor-
There may also be a tendency to overinvest ruption are bad. This potential curse is par-
in partisan projects with negative social sur- ticularly severe for point-source resources
plus (“white elephants”) if politicians find it such as diamonds and precious metals. The
hard to credibly commit to socially efficient resource curse is, however, not cast in stone.
projects (Robinson and Torvik 2005). Fourth, Resource rich countries with good institu-
harnessing windfalls in face of the notorious tions, trade openness, and high investments
volatility of commodity prices implies that in exploration technology seem to enjoy the
governments build precautionary and liquid- fruits of their natural resource wealth. On
ity buffers (by postponing spending and the other hand, the curse seems more severe
bringing taxes forward) and extract natural in presidential democracies. Resource rich
resources excessively fast (compared with countries are also vulnerable to the notorious
the certainty-equivalent Hotelling rule) to volatility of commodity prices, especially if
minimize the commodity price risk of future their financial system is not well developed.
remaining reserves, especially if the degree Recent research, taking account of the endo-
of prudence is high and commodity price geneity of resource dependence, suggests
shocks are persistent and have high vari- that volatility may be the quintessence of the
ance (van der Ploeg 2010a). Future work resource curse. Of course, there is also cross-
needs to extend existing results on uncer- country and panel-data econometric evi-
tainty about future demand for the resource dence that natural resource dependence may
and about exploration and reserves that will undermine the quality of institutions. And
ultimately be available for exploitation (e.g., there is an interesting quasi-experimental
Robert S. Pindyck 1980) to a setting where study on São Tomé, using Cape Verde as con-
governments must decide on their intra- and trol, which suggests that announcements of
intertemporal allocation of public goods and oil discoveries lead to corruption. Resource
setting of tax rates. It is also necessary to bonanzas also reinforce rent grabbing, espe-
investigate how options and other financial cially if institutions are bad, and keep in
instruments can be used to shield economies place bad policies (debt overhang, building
from commodity price volatility and what a too generous welfare state, etc.). There
political constraints prevent these instru- is also evidence that dependence of point-
ments from being used in practice. source resources makes countries prone to
civil conflict and war, although these results
fail to convincingly take account of the effect
6. Concluding Remarks
of conflict on natural resource production.
A quasi-experimental within-country study A recent quasi-experimental study on the
of the districts of Brazil suggests that the districts of Colombia offer evidence that
economic argument that a resource bonanza capital-intensive resources such as oil are
induces appreciation of the real exchange much more prone to civil conflict than labor-
rate and a decline of nonresource export intensive resources such as coffee, rice, or
sectors may have some relevance, but much bananas.
Although, from a normative perspective, finance its spending programs. The advan-
countries should invest their natural resource tage is that the burden of proof for spending
rents into reproducible assets such as physi- resource revenues is with the government.
cal capital, human capital, infrastructure, or Most important is for countries to learn from
foreign assets, World Bank data suggest that the U.S. history and adopt an optimistic,
resource rich economies do not fully reinvest forward-looking approach to technological
their resource wealth and therefore have innovation in resource exploration and the
negative genuine saving rates. But resource search for new reserves. Predatory govern-
rich countries may grow less simply because ments induce mining companies to be less
they save less than other countries. However, transparent about their natural resource rev-
if these countries anticipate a positive rate of enues and become less efficient.21
increases in future resource prices or contin- The analysis of resource rich countries
ual improvements in exploration technology, draws on macroeconomics, public finance,
it may make sense for them to borrow. Rival public policy, international economics,
factions competing for control of resources resource economics, economic history, and
will speed up extraction and may well lead applied econometrics. It also benefits from
to overinvestment. To explain negative gen- collaboration with political scientists and his-
uine saving, more is needed; for example, torians. More research needs to be directed
rapacious resource extraction being associ- at the changing role of institutions through-
ated with erosion of the legal system, inef- out history and in particular to understand
ficient rent seeking, investment in “white why the resource curse seems to be some-
elephants,” and short-sighted politicians. In thing of the last four or five decades whereas
well developed economies, it may be optimal before natural resources were harnessed to
to put natural resource revenues in a sov- promote growth. Also, future work should
ereign wealth fund. In contrast, developing apply the insights from contract theory to
countries often face capital scarcity in which design good incentive-compatible contracts
case it is more appropriate to use the wind- between governments and exploration
fall to pay off debt and lower interest rates to companies. Future research should also be
boost private and domestic capital accumula- directed at appropriate design of auctioning
tion and speed up the process of economic mineral rights. Work is also needed on the
development. Many countries find it hard to question of whether resource rich countries
absorb a substantial and prolonged windfall have different saving patterns, e.g., in world
of foreign exchange since it takes time for financial markets (“petrodollars”) rather
the nontraded sectors to accumulate “home- than in domestic productive capital, and on
grown” capital. Whilst these Dutch disease how this might affect their rate of economic
bottlenecks are being resolved, it is optimal growth if reserves are privately owned. The
to park the windfall revenue abroad until
there is enough capacity to sensibly invest 21 Using a panel of seventy-two industries from fifty-
in the domestic economy. However, fear of one countries over sixteen years, the negative effect of
the fund being raided by political rivals can expropriation risk on corporate transparency appears to
induce a suboptimal political bias toward too be strongest for industries whose profits are highly corre-
lated with oil prices and transparency is lower if oil prices
much partisan, illiquid investment. are high and property rights are bad (Artyom Durnev and
An interesting option is to change the Sergei Guriev 2007). Lack of transparency may lead oil-
constitution to guarantee that resource rev- rich countries to overreport reserves to raise expected
future supply, discourage rival development of oil substi-
enues are handed to the public. The govern- tutes, and thus improve future market conditions (Philip
ment has to subsequently tax its citizens to Sauré 2010).
answers should be contrasted with the situa- rigidities, significant effects on unemploy-
tion where reserves are publicly owned and ment and inflation (Eastwood and Venables
managed by politicians who may be voted 1982; Willem H. Buiter and Purvis 1983; van
out of office soon. The answers will undoubt- Wijnbergen 1984b). Although a higher oil
edly depend on whether there is presidential price boosts demand for the domestic manu-
or a parliamentary system. Future research facturing good, that effect may be swamped
should tackle these questions with rich polit- by the real appreciation created by increased
ical economy models. demand for the home currency. The result
The wide diversity in experiences of may be a decline in domestic manufactur-
countries with substantial natural resources ing output and higher unemployment as
means that comparative analysis and well as a temporary rise in inflation. The oil
exchange of experiences of managing price shock has elements of both a demand
resource rich economies could be very fruit- and supply shock but an increase in resource
ful and that real progress can be made in reserves is mainly a demand shock. Natural
advancing the plight of poor countries with resource discoveries generate permanent
abundant natural resources. Future empiri- income effects well beyond the productive
cal work should move from cross-section to life of the new natural resource reserve.
panel-data regressions to overcome prob- The initial increase in income above its
lems of omitted variable bias and to allow permanent level leads to a current account
for the changing quality of institutions (see surplus but is reversed when reserves run
International Monetary Fund 2005). At the out. Natural resource windfalls do not nec-
same time, detailed country studies and essarily imply a shrinking of manufacturing
quasi-experimental studies are necessary as exports or output and an increase in unem-
often the devil is in the detail and results are ployment but, if a windfall is anticipated,
often clouded by confounding factors. The the real exchange will appreciate and unem-
discovery of natural resources has often been ployment will rise ahead of the windfall.
associated with devastating conflicts and Other simulations of Dutch disease effects
disastrous economic performance. Future and unemployment use perfect-foresight,
research should thus extend the normative intertemporal general equilibrium models
theories of optimally converting depleting with temporary real wage rigidity, short-run
natural resources into productive assets to capital specificity, long-run capital mobility
allow for rent seeking, corruption, and con- between sectors, international capital mobil-
flict. More generally, more work is needed on ity, intermediate inputs, adjustment costs
how to manage natural resource revenues in of investment, dynamics of capital accumu-
a way that promotes sustainable growth, alle- lation, government debt, current account
viates poverty, and avoids conflict. This chal- imbalances, and far-sighted behavior of
lenge is particularly relevant for the resource firms and households (Michael Bruno and
rich, volatile, and conflict prone economies Sachs 1982). With overlapping generations
of Africa with their high population growth or household liquidity constraints, it matters
rates and poor institutions. whether the government uses the resource
windfall to cut public debt or increase trans-
fers. Oil price shocks then induce real appre-
Appendix 1:
ciation and transient unemployment. It is
Unemployment and Dutch Disease
worthwhile to investigate further the effects
A higher world price of natural resources of resource dependence on wage formation
has, in the presence of short-run nominal in competitive and noncompetitive labor
markets (Monojit Chatterji and Simon Price of the world interest rate r and the world
1988; Rolf Jens Brunstad and Jan Morten resource price . Capital market equilibrium
Dyrstad 1997). Capital market imperfections demands Pg(kN) = f (kT, rT) = r and gives,
may also generate adverse growth effects of together with the condition fr(kT, rT) = ,
resource booms. For example, if resource kN, kT and rT in terms of r and P (or ). I
income cannot be invested in international obtain (suppressing r) that rT = rT(),
capital markets, resource rich economies rT′ < 0 and:
may experience slower steady-state growth as
people live beyond their means and are over- W = W(), P = P(), kN = kN () with
shooting their steady-state levels (Francisco
Rodriguez and Sachs 1999). P′ = dW W′ = −cQ /cW < 0 and
k  ′N    = −g′ PQ /g′′ < 0.

Appendix 2:
Endogenous Growth and Dutch Disease
Along the factor price frontier, the wage
I extend Sachs and Warner (1995) to allow and the price of nontraded goods decrease if
for natural resource use in production of the world price of natural resources increases.
traded goods, RT. The traded and nontraded The latter induces a fall in capital intensity
sectors have the same labor-augmenting proof the nontraded sector. Overlapping house-
ductivity growth, fully determined by the holds with logarithmic utility and discount
share of employment in the traded sector factor 1/(1 + ρ) < 1 enjoy wage w when
LT. The production functions of the two sec- young and receive a natural resource divi-
tors in extensive and intensive form are thus dend per effective worker of e. It follows that
given by: aggregate consumption per effective young
worker is given by:
(  )(  )
XT = F(L T H, KT, RT) and μ 1+ρ
cNt = _     _    
P(t) 2 + ρ
XN = G(LN H, KN) with
Ht = (1 + θL Tt−1)Ht−1, θ > 0 and [

× Wt + t et + __
  
   ( 
(1 + rt−1)
(1 + ρ)(1 + LTt−1)
  )
]
xT = XT/L T H = F(1, kT, rT)
× (Wt−1 + t−1 et−1)
≡ f (kT, rT) and
xN = XN/L N H = G(1, kN) ≡ g( kN). = (1 − L Tt)g(kN(t)),
The zero profit conditions are 1 = where μ indicates the relative utility weight
cW(W, r, )W + cr(W, r, )r + cQ(W, r, ) (and budget share) of nontraded consump-
and dW(W, r)W + dr(W, r)r = P, where W tion. The factor (1 + LNt−1) is necessary to
indicates the wage, r the exogenous world convert from old to young workers, the fac-
interest rate, and c(·) and d(·) are the unit- tor (1 – LNt) is to convert output per worker
cost functions homogenous of degree one to output per young worker in the nontraded
associated with the CRTS production func- sector, and the labor market equilibrium con-
tions G(·) and F(·). They give the price of dition LTt + LNt =  1 has been used. This condi-
nontraded goods P and the wage W in terms tion for nontraded goods market equilibrium
can be written as a stable difference equa- max–min egalitarianism which leads to a

tion LTt = Ω(L Tt−1, et, et−1, t, t−1) with constant level of per capita consumption.
0 < Ω1 < 1, Ωi < 0, i = 2, 3, 4, 5. An in- Nondecreasing per capita consumption is
crease in resource dividend induces a grad- infeasible under exponential population
ual shift of employment from the nontraded growth if resources are essential inputs in
to the traded sector (LT falls), so there is less production and there is no technical progress
learning by doing and the growth rate is per- (Dasgupta and Heal 1974; Robert M. Solow
manently lowered ((Ht – Ht−1)/Ht−1 falls). 1974; Joseph E. Stiglitz 1974), but feasible
In this setup, the resource dividend cannot with quasi-arithmetic population growth
affect relative productivity. If this dividend is (Tapan Mitra 1983; Asheim et al. 2007). The
driven by a higher world price of resources, so-called Hartwick rule states that natural
depreciation of the real exchange rate and resource rents should be fully reinvested in
the lower capital intensity in production of reproducible capital under max–min social
nontraded goods lead to even bigger falls welfare. This entails in the absence of popu-
in traded sector employment, learning by lation growth a constant savings rate equal
doing, and the rate of growth. GDP is given to the constant functional share of resource
by e + WH + r (KT + KN) = E + (W + r) inputs (Hartwick 1977). With no popula-
× H[kN + L T(kT – kN)]. Hence, GDP grows tion growth and no technical progress, the
at the rate ξθLT where the nonresource share economy features constant consumption and
of GDP is ξ. Nonresource GDP falls on is thus a max–min optimum. If there is posi-
impact after a shock in e1 if the traded sec- tive population growth, a max–min optimum
tor is capital-intensive, that is requires constant consumption per head. If
consumption per head were rising (falling)
∂ GDP/∂ (e1) = 1 over time, welfare could be raised if earlier
(later) generations saved and invested less
+ (W + r)H1(∂ L T1/∂ (e1))(kT – kN) < 1) or consumed capital at the expense of later
(earlier) generations. A max–min optimum
as ∂ L T1/∂(e1) < 0. then requires that investment in reproduc-
ible capital exceeds natural resource rents.
Consider a closed economy with resource
depletion ∫
Appendix 3:
0∞
    R(t) dt = S0 zero deprecia-
Hartwick Rule for Reinvesting Natural
tion, savings rate s ≡  K / Y, Cobb–Douglas
Resource Rents in a Closed Economy
production Y = F(K, R) = K αRβL1−α−β, and
Does exhaustibility of natural resources population growth rate equal to η. Firms
constrain the growth potential if resources set marginal products to factor prices, that
are essential in production? The answer is FR = Q and FK = r. The Hotelling  · rule
 · in
depends on the ease with which reproduc-  ·absence of extraction costs is  
F  
  /F
R R =     −
/Y
Y
 
ible inputs can be substituted for exhaustible      = FK = αY/K. The following saving rate
/R
R
natural resources. Utilitarian social welfare sustains a stable income per capita:
implies that consumption first rises and then
vanishes in the long run (e.g., Dasgupta and  ·  ·
     − η = sr − αη + β( Y /  Y − η − r)
/ Y
Y
Heal 1979). It is difficult to defend from an
ethical point of view that the consumption
=  __
(s − β)r − αη
level of future generations vanishes asymp-        = 0
1 − β
totically. Hence, the normative focus in the
literature on natural resources has been on ⇒ s = β + (α/r)η ≡ s*.
If there is no population growth (η = 0), the Divisia consumption price index rather
all resource rents must be invested in capi- than the price index of output. Aggregation
tal to sustain a constant income per capita across multiple infinitely lived households
(i.e., R = sY or s = β). This is the well- with heterogeneous consumption prefer-
known Hartwick rule and holds for general ences is feasible under constant returns to
production functions.22 It corresponds to scale. The return on the increasingly scarce
a max–min optimum, since it sustains con- natural resource increases at the expense
stant consumption per capita. With popula- of the increasingly abundant other factors
tion growth (η > 0), the country must invest of production. Capital gains then represent
more than the resource rents to sustain capitalization of those future changes in fac-
constant income and constant consumption tor prices and are effectively a transfer from
per capita (s* > β). The interest rate then one factor to another rather than a change in
declines while the capital–output ratio rises resources available to the whole economy. As
with time, so the saving rate rises over time. a result, in a closed economy where all fac-
The steady-state depletion rate is r − η, tors are entirely owned by households, the
so societies with fast growing populations net gains are zero and capital gains should
should deplete their resources less rapidly. not be included in real income.
Without population growth and technical
progress, the Hartwick rule also results in a
max–min optimum in economies with many Appendix 4:
consumption goods, heterogeneous capital Absorption Constraints and
goods and endogenous labor supplies pro- Dutch Disease Dynamics
vided there is free disposal and stock rever-
sal (Avinash Dixit, Peter Hammond, and Assume a small open dependent econ-
Michael Hoel 1980). The conditions under omy with perfect access to the interna-
which a max–min optimum implies adher- tional capital market. The traded good is the
ence to the Hartwick rule are also known numeraire. Production in the traded sector
(e.g., Withagen and Asheim 1998; Mitra only used labor, so normalizing productiv-
2002). ity at one we have YT = L T and W = 1. The
The Hartwick rule is related to the nontraded sector has a Cobb–Douglas pro-
Hicksian definition of real income, that is duction function, YN = Kα L  1−α
N  , 0 < α < 1.
“the maximum amount a man can spend and Profit maximization yields the demand
still be as well off at the end of the week as for labor in the nontraded sector, LN
at the beginning.” The general equilibrium = K[(1 − α)P]1/α, where P is the relative
features of such a Hicksian definition of real price of nontraded goods. Labor market equi-
income, defined as zero change in the present librium then gives L T = 1 − K[(1 − α)P]1/α.
discounted value of current and future util- Output of nontraded goods is given by YN
ity, are well understood (Asheim and Martin = K[(1 − α)P](1−α)/α. Denoting the unit-
L. Weitzman 1991; J. A. Sefton and M. R. cost function for producing capital goods by
Weale 2006). In contrast to national account- c(P) = Pγ with 0 < γ < 1 the share of non-
ing practice, income must be deflated with traded goods in the production of home-
grown capital, profit maximization requires
 ·
that the marginal product of capital, r(P)
22 Differentiating K      = F(K, R, L) −  C = FRR and
· = α[(1 − α)P](1−α)/α, must equal the rental
   = F RR yields
using
  ··
   = F
K
the
 · Hotelling

K
 
     + F
 ·  

R
 
    −  
C
and Hartwick
·
  
=
 ·
( 
F  
  /F ) F
rules, K
R + F
   

R
 
·
    − C
·  
    = K
··
    − C, change, r*, plus the depreciation charge, δ,
 · K R R R R R   · (P)/c(P).
so C    = 0. minus the expected capital gains, c
Preferences are homothetic and e(P) = Pβ, after downward and permanent jump in λ
0 < β <  1, denotes the unit-expenditure induced by a windfall of foreign exchange.
function, hence consumption in nontraded Note that as the share of traded goods in cap-
goods is given by CN = e′(P)U, where U ital goods vanishes, γ → 0, the capital stock
denotes real consumption (or utility). adjusts immediately to a natural resource
Equilibrium on the market for nontraded windfall. As a result of the downward jump
goods  · is given by CN + c′(P)I = YN, where in λ, there is an immediate and permanent
I=      + δK denotes gross investment. The

K upward jump in K and there is no need for
representative consumer maximizes utility, the real exchange rate to appreciate whatso-
∫0∞    ln(U) exp(−ρt) dt, subject to the con- ever. However, much capital (think of nurses
straint, ∫0∞    [e(P)U + c(P)I] exp (−r*t) dt and teachers as well as infrastructure) must
≤ F0 + V0 + ∫0∞     (YT + PYN) exp (−r*t) dt, be homegrown and cannot be imported.
where F indicates foreign assets (bonds) and Consequently, γ is closer to one and absorp-
V the present value of natural resource rev- tion constraints will manifest themselves.
enues (i.e., natural resource wealth). The This may be seen from the saddle-path dia-
budget constraint states that the present gram given in figure 7. The optimal response
value of the stream of current and future to a windfall is for the real exchange to
consumption and investment spending on appreciate on impact signaling labor to shift
traded and nontraded goods cannot exceed from the traded to the nontraded sector and
initial foreign assets plus initial resource shifting demand from nontraded to traded
wealth plus the present value of current and goods. Over time, investment induces a
future traded and nontraded production. If gradual expansion in homegrown capital
we suppose that r* = ρ, the optimality condi- that permits a gradual reversal of the initial
tion for the consumer is 1/U = λ e(P), where appreciation of the real exchange rate. The
the marginal utility of wealth λ has to be resulting temporary boost to the return on
constant over time. At the time the resource capital in the nontraded sector r(P) is in line
windfall becomes known (upward jump in with the anticipated capital losses on those
V0), λ jumps down and stays at this lower capital goods (as over time the relative price
value forever after. A resource windfall thus of investment goods c(P) will fall and return
corresponds to an unanticipated, permanent to its original level). The windfall results in
fall in the marginal utility of wealth λ. an immediate and permanent increase in
The adjustment path follows from the the consumption of traded goods, but con-
system of differential equations describing, sumption of nontraded goods increases
respectively, equilibrium in the market for on impact and subsequently continues to
nontraded goods and equity arbitrage: increase toward its new steady-state level.
[ ]
Homegrown capital also jumps up on impact
_ β P1−γ
 K    = K((1 − α)P) α   −  _    _
 ·   
  1−α  
γ   
 
  − δK,

and then continues to rise to its new steady-
λP state level. Due to the gradual increase
K(0) = K0, in consumption as supply constraints are
_
gradually relaxed, the total stock of assets
      = r* + δ − α((1 − α)P) α   _
[ ]
 ·   
  1−α  
γ  , 
P
P increases by more than the windfall. Hence,
there is initial saving (parking funds abroad)
P(0) free. relative to the permanent income hypoth-
esis. Van der Ploeg and Venables (2010) pro-
The steady-state value of P is independent vide a much more general analysis allowing
of λ but the steady-state value of K increases for capital accumulation in the traded sector
∙
P=0
∙
K= 0
Figure 7. Absorption Constraints and Dutch Disease Dynamics
as well and highlighting the impossibility of Causes, Macroeconomic Symptoms: Volatility, Crises
and Growth.” Journal of Monetary Economics, 50(1):
shifting capital between the two sectors once 49–123.
it has been installed. Acemoglu, Daron, and James A. Robinson. 2006. “Eco-
nomic Backwardness in Political Perspective.” Amer-
References ican Political Science Review, 100(1): 115–31.
Acemoglu, Daron, James A. Robinson, and Thierry
Acemoglu, Daron. 1995. “Reward Structures and the Verdier. 2004. “Kleptocracy and Divide-and-Rule:
Allocation of Talent.” European Economic Review, A Model of Personal Rule.” Journal of the European
39(1): 17–33. Economic Association, 2(2–3): 162–92.
Acemoglu, Daron, Simon Johnson, and James A. Rob- Adam, Christopher S., and David L. Bevan. 2006. “Aid
inson. 2001. “The Colonial Origins of Comparative and the Supply Side: Public Investment, Export Per-
Development: An Empirical Investigation.” Ameri- formance, and Dutch Disease in Low-Income Coun-
can Economic Review, 91(5): 1369–1401. tries.” World Bank Economic Review, 20(2): 261–90.
Acemoglu, Daron, Simon Johnson, and James A. Rob- Adam, Christopher S., and Stephen A. O’Connell.
inson. 2003. “An African Success Story: Botswana.” 1999. “Aid, Taxation and Development in Sub-Saha-
In In Search of Prosperity: Analytic Narratives on ran Africa.” Economics and Politics, 11(3): 225–53.
Economic Growth, ed. Dani Rodrik, 80–119. Prince Adam, Christopher S., and Stephen A. O’Connell.
ton and Oxford: Princeton University Press. 2004. “Aid versus Trade Revisited: Donor and Recip-
Acemoglu, Daron, Simon Johnson, James A. Robin- ient Policies in the Presence of Learning-by-Doing.”
son, and Yunyong Thaicharoen. 2003. “Institutional Economic Journal, 114(492): 150–73.
Ades, Alberto, and Rafael Di Tella. 1999. “Rents, Aslaksen, Silje, and Ragnar Torvik. 2006. “A Theory
Competition, and Corruption.” American Economic of Civil Conflict and Democracy in Rentier States.”
Review, 89(4): 982–93. Scandinavian Journal of Economics, 108(4): 571–85.
Aghion, Philippe, George-Marios Angeletos, Abhi- Atkinson, Giles, and Kirk Hamilton. 2003. “Savings,
jit Banerjee, and Kalina Manova. 2005. “Volatility Growth and the Resource Curse Hypothesis.” World
and Growth: Credit Constraints and Productivity- Development, 31(11): 1793–1807.
Enhancing Investment.” National Bureau of Eco- Auty, Richard M. 1997. “Natural Resource Endow-
nomic Research Working Paper 11349. ment, the State and Development Strategy.” Journal
Aghion, Philippe, Philippe Bacchetta, Romain Ran- of International Development, 9(4): 651–63.
cière, and Kenneth Rogoff. 2009. “Exchange Rate Auty, Richard M. 2001a. “The Political Economy of
Volatility and Productivity Growth: The Role of Resource-Driven Growth.” European Economic
Financial Development.” Journal of Monetary Eco- Review, 45(4–6): 839–46.
nomics, 56(4): 494–513. Auty, Richard M., ed. 2001b. Resource Abundance
Aghion, Philippe, and Peter Howitt. 1998. Endogenous and Economic Development. Oxford and New York:
Growth Theory. Cambridge and London: MIT Press. Oxford University Press.
Aizenman, Joshua, and Nancy Marion. 1999. “Volatility Auty, Richard M. 2004. “Economic and Political Reform
and Investment: Interpreting Evidence from Devel- of Distorted Oil-Exporting Economies.” http://www.
oping Countries.” Economica, 66(262): 157–79. earth.columbia.edu/cgsd/events/resource_curse_
Alesina, Alberto, and Allan Drazen. 1991. “Why timetable.html.
Are Stabilizations Delayed?” American Economic Baland, Jean-Marie, and Patrick Francois. 2000. “Rent-
Review, 81(5): 1170–88. Seeking and Resource Booms.” Journal of Develop-
Alesina, Alberto, and Guido Tabellini. 1990. “A Positive ment Economics, 61(2): 527–42.
Theory of Fiscal Deficits and Government Debt.” Bardhan, Pranab. 1997. “Corruption and Develop-
Review of Economic Studies, 57(3): 403–14. ment: A Review of Issues.” Journal of Economic Lit-
Andersen, Jorgen Juel, and Silje Aslaksen. 2008. “Con- erature, 35(3): 1320–46.
stitutions and the Resource Curse.” Journal of Devel- Besley, Timothy, and Stephen Coate. 1997. “An Eco-
opment Economics, 87(2): 227–46. nomic Model of Representative Democracy.” Quar-
Andersen, Svein S. 1993. The Struggle over North Sea terly Journal of Economics, 112(1): 85–114.
Oil and Gas: Government Strategies in Denmark, Besley, Timothy, and Stephen Coate. 1998. “Sources
Britain and Norway. New York: Oxford University of Inefficiency in a Representative Democracy: A
Press; Oslo: Scandinavian University Press. Dynamic Analysis.” American Economic Review,
Anderson, Terry L., and Gary D. Libecap. 2005. “Forg- 88(1): 139–56.
ing a New Environmental and Resource Economics Besley, Timothy, and Torsten Persson. 2009. “The Ori-
Paradigm: The Contractual Bases for Exchange.” gins of State Capacity: Property Rights, Taxation,
Unpublished. and Politics.” American Economic Review, 99(4):
Angrist, Joshua D., and Adriana D. Kugler. 2008. “Rural 1218–44.
Windfall or a New Resource Curse? Coca, Income, Besley, Timothy, and Torsten Persson. 2010. “State
and Civil Conflict in Colombia.” Review of Econom- Capacity, Conflict and Development.” Econometrica,
ics and Statistics, 90(2): 191–215. 78(1): 1–34.
Arezki, Rabah, and Frederick van der Ploeg. Forth- Bevan, David L., Paul Collier, and Jan Willem Gunning.
coming. “Do Natural Resources Depress Income Per 1999. The Political Economy of Poverty, Equity, and
Capita?” Review of Development Economics. Growth: Nigeria and Indonesia. Oxford and New
Arrow, Kenneth J., Partha Dasgupta, and Karl-Goran York: Oxford University Press.
Mäler. 2003. “The Genuine Savings Criterion and Beviá, Carmen, and Luis C. Corchón. 2010. “Peace
the Value of Population.” Economic Theory, 21(2–3): Agreements without Commitment.” Games and Eco-
217–25. nomic Behavior, 68(2): 469–87.
Asheim, Geir B. 1986. “Hartwick’s Rule in Open Bhattacharyya, Sambit, and Roland Hodler. 2010. “Nat-
Economies.” Canadian Journal of Economics, 19(3): ural Resources, Democracy and Corruption.” Euro-
395–402. pean Economic Review, 54(4): 608–21.
Asheim, Geir B. 1996. “Capital Gains and Net National Blattman, Christopher, Jason Hwang, and Jeffrey
Product in Open Economies.” Journal of Public Eco- G. Williamson. 2007. “Winners and Losers in the
nomics, 59(3): 419–34. Commodity Lottery: The Impact of Terms of Trade
Asheim, Geir B., Wolfgang Buchholz, John M. Hart- Growth and Volatility in the Periphery 1870–1939.”
wick, Tapan Mitra, and Cees Withagen. 2007. “Con- Journal of Development Economics, 82(1): 156–79.
stant Savings Rates and Quasi-arithmetic Population Boschini, Anne D., Jan Pettersson, and Jesper Roine.
Growth under Exhaustible Resource Constraints.” 2007. “Resource Curse or Not: A Question of Appro-
Journal of Environmental Economics and Manage- priability.” Scandinavian Journal of Economics,
ment, 53(2): 213–29. 109(3): 593–617.
Asheim, Geir B., and Martin L. Weitzman. 2001. “Does Bourguignon, Francois, and Thierry Verdier. 2000.
NNP Growth Indicate Welfare Improvement?” Eco- “Oligarchy, Democracy, Inequality and Growth.”
nomics Letters, 73(2): 233–39. Journal of Development Economics, 62(2): 285–313.
Bovenberg, A. Lans, and Sjak A. Smulders. 1996. Dutch Disease and Unemployment.” Oxford Eco-
“Transitional Impacts of Environmental Policy in nomic Papers, 40(2): 302–21.
an Endogenous Growth Model.” International Eco- Chenery, Hollis, and Moises Syrquin. 1975. Patterns
nomic Review, 37(4): 861–93. of Development, 1950–1970. Oxford and New York:
Brahmbhatt, Milan, Otaviano Canuto, and Ekaterina Oxford University Press.
Vostroknutova. 2010. “Dealing with Dutch Disease.” Collier, Paul, and Anke Hoeffler. 2004. “Greed and
Economic Premise 16. Washington, D.C.: World Grievance in Civil War.” Oxford Economic Papers,
Bank. 56(4): 563–95.
Brautigam, Deborah A., and Stephen Knack. 2004. Collier, Paul, and Anke Hoeffler. 2005. “Resource
“Foreign Aid, Institutions, and Governance in Sub- Rents, Governance, and Conflict.” Journal of Conflict
Saharan Africa.” Economic Development and Cul- Resolution, 49(4): 625–33.
tural Change, 52(2): 255–85. Collier, Paul, and Anke Hoeffler. 2009. “Testing the
Bravo-Ortega, Claudio, and Jose de Gregorio. 2005. Neocon Agenda: Democracy in Resource-Rich Soci-
“The Relative Richness of the Poor? Natural eties.” European Economic Review, 53(3): 293–308.
Resources, Human Capital, and Economic Growth.” Collier, Paul, Anke Hoeffler, and Mans Soderbom.
World Bank Policy Research Working Paper 3484. 2004. “On the Duration of Civil War.” Journal of
Brollo, Fernanda, Tommaso Nannicini, Roberto Peace Research, 41(3): 253–73.
Perotti, and Guido Tabellini. 2010. “The Politi- Collier, Paul, Frederick van der Ploeg, Michael Spence,
cal Resource Curse.” Centre for Economic Policy and Anthony J. Venables. 2010. “Managing Resource
Research Discussion Paper 7672. Revenues in Developing Economies.” IMF Staff
Brunnschweiler, Christa N., and Erwin H. Bulte. 2008. Papers, 57(1): 84–118.
“The Resource Curse Revisited and Revised: A Tale Collier, Paul, and Anthony J. Venables. Forthcoming.
of Paradoxes and Red Herrings.” Journal of Environ- “Illusory Revenues: Import Tariffs in Resource-Rich
mental Economics and Management, 55(3): 248–64. and Aid-Rich Economies.” Journal of Development
Bruno, Michael, and Jeffrey Sachs. 1982. “Energy Economics.
and Resource Allocation: A Dynamic Model of the Corden, W. Max. 1984. “Booming Sector and Dutch
‘Dutch Disease.’” Review of Economic Studies, 49(5): Disease Economics: Survey and Consolidation.”
845–59. Oxford Economic Papers, 36(3): 359–80.
Brunstad, Rolf Jens, and Jan Morten Dyrstad. 1997. Corden, W. Max, and J. Peter Neary. 1982. “Boom-
“Booming Sector and Wage Effects: An Empiri- ing Sector and De-industrialisation in a Small Open
cal Analysis on Norwegian Data.” Oxford Economic Economy.” Economic Journal, 92(368): 825–48.
Papers, 49(1): 89–103. Dal Bó, Ernesto, and Pedro Dal Bó. Forthcoming.
Buiter, Willem H., and Douglas D. Purvis. 1983. “Oil, “Workers, Warriors, and Criminals: Social Conflict in
Disinflation, and Export Competitiveness: A Model General Equilibrium.” Journal of the European Eco-
of the ‘Dutch Disease.’” In Economic Interdepen- nomic Association.
dence and Flexible Exchange Rates, ed. Jagdeep Dasgupta, Partha. 2001a. Human Well-Being and the
Bhandari, Bluford H. Putnam, and Jay H. Levin, Natural Environment. Oxford and New York: Oxford
221–48. Cambridge and London: MIT Press. University Press.
Bulte, Erwin H., and Richard Damania. 2008. Dasgupta, Partha. 2001b. “Valuing Objects and Evalu-
“Resources for Sale: Corruption, Democracy and the ating Policies in Imperfect Economies.” Economic
Natural Resource Curse.” B.E. Journal of Economic Journal, 111(471): C1–29.
Analysis and Policy, 8(1). Dasgupta, Partha, Robert Eastwood, and Geoffrey M.
Bulte, Erwin H., Richard Damania, and Robert T. Heal. 1978. “Resource Management in a Trading
Deacon. 2005. “Resource Intensity, Institutions, and Economy.” Quarterly Journal of Economics, 92(2):
Development.” World Development, 33(7): 1029–44. 297–306.
Caselli, Francesco. 2006. “Power Struggles and the Dasgupta, Partha, and Geoffrey M. Heal. 1974. “The
Natural Resource Curse.” Unpublished. Optimal Depletion of Exhaustible Resources.”
Caselli, Francesco, and Wilbur John Coleman. 2006. Review of Economic Studies, 41: 3–28.
“On the Theory of Ethnic Conflict.” Centre for Eco- Dasgupta, Partha, and Geoffrey M. Heal. 1979. Eco-
nomic Policy Research Discussion Paper 5622. nomic Theory and Exhaustible Resources. Cam-
Caselli, Francesco, and Tom Cunningham. 2009. bridge and New York: Cambridge University Press.
“Leader Behaviour and the Natural Resource Curse.” Dasgupta, Partha, and Karl-Goran Mäler. 2000. “Net
Oxford Economic Papers, 61(4): 628–50. National Product, Wealth, and Social Well-Being.”
Caselli, Francesco, and Guy Michaels. 2009. “Do Oil Environment and Development Economics, 5(1–2):
Windfalls Improve Living Standards? Evidence from 69–93.
Brazil.” Unpublished. David, Paul A., and Gavin Wright. 1997. “Increas-
Cashin, Paul, Luis F. Céspedes, and Ratna Sahay. 2004. ing Returns and the Genesis of American Resource
“Commodity Currencies and the Real Exchange Abundance.” Industrial and Corporate Change, 6(2):
Rate.” Journal of Development Economics, 75(1): 203–45.
239–68. Davis, Jeffrey, Rolando Ossowski, James Daniel, and
Chatterji, Monojit, and Simon Price. 1988. “Unions, Steven Barnett. 2001. “Stabilization and Savings
Funds for Nonrenewable Resources.” International 89(3): 379–99.

Monetary Fund Occasional Paper 205. Fum, Ruikang Marcus, and Roland Hodler. 2010. “Nat-
Demsetz, Harold. 1967. “Toward a Theory of Property ural Resources and Income Inequality: The Role
Rights.” American Economic Review, 57(2): 347–59. of Ethnic Divisions.” Economics Letters, 107(3):
Dixit, Avinash, Peter Hammond, and Michael Hoel. 360–63.
1980. “On Hartwick’s Rule for Regular Maximin Gallup, John Luke, Jeffrey D. Sachs, and Andrew D.
Paths of Capital Accumulation and Resource Deple- Mellinger. 1999. “Geography and Economic Devel-
tion.” Review of Economic Studies, 47(3): 551–56. opment.” International Regional Science Review,
Dube, Oeindrila, and Juan F. Vargas. 2008. “Commod- 22(2): 179–232.
ity Price Shocks and Civil Conflict: Evidence from Gelb, Alan, ed. 1988. Oil Windfalls: Blessing or Curse?
Colombia.” Unpublished. Oxford and New York: Oxford University Press.
Durnev, Artyom, and Sergei Guriev. 2007. “The Glaeser, Edward L., Rafael La Porta, Florencio Lopez-
Resource Curse: A Corporate Transparency Chan- de-Silanes, and Andrei Shleifer. 2004. “Do Institu-
nel.” Centre for Economic Policy Research Discus- tions Cause Growth?” Journal of Economic Growth,
sion Paper 6547. 9(3): 271–303.
Easterly, William, and Ross Levine. 2002. “Tropics, Goderis, Benedikt, and Samuel W. Malone. Forth-
Germs, and Crops: How Endowments Influence coming. “Natural Resource Booms and Inequality:
Economic Development.” National Bureau of Eco- Theory and Evidence.” Scandinavian Journal of
nomic Research Working Paper 9106. Economics.
Eastwood, Robert, and Anthony J. Venables. 1982. Groot, Fons, Cees Withagen, and Aart de Zeeuw. 2003.
“The Macroeconomic Implications of a Resource “Strong Time-Consistency in the Cartel-versus-
Discovery in an Open Economy.” Economic Journal, Fringe Model.” Journal of Economic Dynamics and
92(366): 285–99. Control, 28(2): 287–306.
Elíasson, Ludvik, and Stephen J. Turnovsky. 2004. Grossman, Herschel I. 2001. “The Creation of Effec-
“Renewable Resources in an Endogenously Grow- tive Property Rights.” American Economic Review,
ing Economy: Balanced Growth and Transitional 91(2): 347–52.
Dynamics.” Journal of Environmental Economics Gylfason, Thorvaldur. 2001. “Natural Resources, Edu-
and Management, 48(3): 1018–49. cation, and Economic Development.” European
Ellman, M. 1981. “Natural Gas, Restructuring and Re- Economic Review, 45(4–6): 847–59.
industrialisation: The Dutch Experience of Industrial Gylfason, Thorvaldur. 2004. “Natural Resources and
Policy.” In Oil or Industry? Energy Industrialisation Economic Growth: From Dependence to Diversifi-
and Economic Policy: Issues in Canada, Mexico, Nor- cation.” Centre for Economic Policy Research Dis-
way, and the United Kingdom, ed. Terry Barker and cussion Paper 4804.
Vladimir Brailovsky. London: Academic Press. Gylfason, Thorvaldur, Tryggvi Thor Herbertsson,
Enders, Klaus, and Horst Herberg. 1983. “The Dutch and Gylfi Zoega. 1999. “A Mixed Blessing: Natural
Disease: Causes, Consequences, Cures and Calm- Resources and Economic Growth.” Macroeconomic
atives.” Weltwirtschaftliches Archiv, 119(3): 473–97. Dynamics, 3(2): 204–25.
Farzin, Y. Hossein. 2010. “Sustainability, Optimality, Gylfason, Thorvaldur, and Gylfi Zoega. 2003. “Inequal-
and Development Policy.” Review of Development ity and Economic Growth: Do Natural Resources
Economics, 14(2): 262–81. Matter?” In Inequality and Growth: Theory and
Fasano, Ugo. 2002. “With Open Economy and Sound Policy Implications, ed. Theo S. Eicher and Stephen
Policies, U.A.E. Has Turned Oil ‘Curse’ into a Bless- J. Turnovsky, 255–92. Cambridge and London: MIT
ing.” IMFSurvey, October 21, 330–32. Press, 255–92.
Fearon, James D. 2005. “Primary Commodity Exports Habakkuk, H. J. 1962. American and British Tech-
and Civil War.” Journal of Conflict Resolution, 49(4): nology in the Nineteenth Century: The Search for
483–507. Labour-Saving Inventions. Cambridge and New
Fearon, James D., and David D. Laitin. 2003. “Ethnic- York: Cambridge University Press.
ity, Insurgency, and Civil War.” American Political Haber, Stephen, and Victor Menaldo. 2008. “Do Natu-
Science Review, 97(1): 75–90. ral Resources Fuel Authoritarianism? A Reappraisal
Flug, Karnit, Antonio Spilimbergo, and Erik Wachten- of the Resource Curse.” Unpublished.
heim. 1998. “Investment in Education: Do Economic Hall, Robert E., and Charles I. Jones. 1999. “Why Do
Volatility and Credit Constraints Matter?” Journal of Some Countries Produce So Much More Output per
Development Economics, 55(2): 465–81. Worker than Others?” Quarterly Journal of Econom-
Forsyth, Peter J., and John A. Kay. 1980. “The Eco- ics, 114(1): 83–116.
nomic Implications of North Sea Oil Revenues.” Fis- Hamilton, Kirk, and Michael Clemens. 1999. “Genuine
cal Studies, 1(3): 1–28. Savings Rates in Developing Countries.” World Bank
Forsyth, Peter J., and John A. Kay. 1981. “Oil Revenues Economic Review, 13(2): 333–56.
and Manufacturing Output.” Fiscal Studies, 2(2): Hamilton, Kirk, and John M. Hartwick. 2005. “Invest-
9–17. ing Exhaustible Resource Rents and the Path of Con-
Frankel, Jeffrey A., and David Romer. 1999. “Does sumption.” Canadian Journal of Economics, 38(2):
Trade Cause Growth?” American Economic Review, 615–21.
Hannesson, Rognvaldur. 2001. Investing for Sustain- Oxford Economic Papers, 44(4): 641–71.
ability: The Management of Mineral Wealth. Boston; Knack, Stephen, and Philip Keefer. 1995. “Institutions
Dordrecht and London: Kluwer Academic. and Economic Performance: Cross-Country Tests
Harding, Torfinn, and Frederick van der Ploeg. 2009. Using Alternative Institutional Measures.” Econom-
“Fiscal Reactions to Anticipated Hydrocarbon Wind- ics and Politics, 7(3): 207–27.
falls and Pension Burden: Is Norway’s Stabilization Krugman, Paul. 1987. “The Narrow Moving Band, the
Fund Prudent Enough?” OxCarre Research Paper Dutch Disease, and the Competitive Consequences
27. of Mrs. Thatcher: Notes on Trade in the Presence of
Harding, Torfinn, and Anthony J. Venables. 2010. Dynamic Scale Economies.” Journal of Development
“Exports, Imports and Foreign Exchange Windfalls.” Economics, 27(1–2): 41–55.
Unpublished. Kuralbayeva, Karlygash, and Radosław Stefański. 2010.
Harford, Tim, and Michael Klein. 2005. “Aid and the “Windfalls, Structural Transformation and Special-
Resource Curse: How Can Aid Be Designed to Pre- ization.” OxCarre Research Paper 54.
serve Institutions?” Public Policy Journal, 291: 1–4. Lane, Philip R., and Aaron Tornell. 1996. “Power,
Hartwick, John M. 1977. “Intergenerational Equity and Growth, and the Voracity Effect.” Journal of Eco-
the Investing of Rents from Exhaustible Resources.” nomic Growth, 1(2): 213-41.
American Economic Review, 67(5): 972–74. Larsen, Erling Roed. 2006. “Escaping the Resource
Hartwick, John M., Ngo Van Long, and Huilan Tian. Curse and the Dutch Disease? When and Why Nor-
2001. “Deforestation and Development in a Small way Caught Up with and Forged Ahead of Its Neigh-
Open Economy.” Journal of Environmental Econom- bors.” American Journal of Economics and Sociology,
ics and Management, 41(3): 235–51. 65(3): 605–40.
Harvey, David I., Neil M. Kellard, Jakob B. Madsen, Lay, Jann, and Toman Omar Mahmoud. 2004.
and Mark E. Wohar. 2010. “The Prebisch–Singer “Bananas, Oil, and Development: Examining the
Hypothesis: Four Centuries of Evidence.” Review of Resource Curse and Its Transmission Channels by
Economics and Statistics, 92(2): 367–77. Resource Type.” Kiel Working Paper 1218.
Hausmann, Ricardo, and Roberto Rigobon. 2003. “An Leite, Carlos, and Jens Weidmann. 1999. “Does
Alternative Interpretation of the ‘Resource Curse’: Mother Nature Corrupt—Natural Resources, Cor-
Theory and Policy Implications.” National Bureau of ruption, and Economic Growth.” International Mon-
Economic Research Working Paper 9424. etary Fund Working Paper 99/85.
Hodler, Roland. 2006. “The Curse of Natural Resources Libecap, Gary D. 1978. “Economic Variables and the
in Fractionalized Countries.” European Economic Development of the Law: The Case of Western Min-
Review, 50(6): 1367–86. eral Rights.” Journal of Economic History, 38(2):
Humphreys, Macartan. 2005. “Natural Resources, 338–62.
Conflict, and Conflict Resolution: Uncovering the Libecap, Gary D., and Ronald N. Johnson. 1979. “Prop-
Mechanisms.” Journal of Conflict Resolution, 49(4): erty Rights, Nineteenth-Century Federal Timber
508–37. Policy, and the Conservation Movement.” Journal of
International Monetary Fund. 2005. “Building Institu- Economic History, 39(1): 129–42.
tions.” In World Economic Outlook: Building Insti- Lujala, Paivi. 2010. “The Spoils of Nature: Armed Civil
tutions, 125–60. Washington, D.C.: International Conflict and Rebel Access to Natural Resources.”
Monetary Fund. Journal of Peace Research, 47(1): 15–28.
Isham, Jonathan, Michael Woolcock, Lant Pritchett, Mahbub Morshed, A. K. M., and Stephen J. Turn-
and Gwen Busby. 2005. “The Varieties of Resource ovsky. 2004. “Sectoral Adjustment Costs and Real
Experience: Natural Resource Export Structures and Exchange Rate Dynamics in a Two-Sector Depen-
the Political Economy of Economic Growth.” World dent Economy.” Journal of International Economics,
Bank Economic Review, 19(2): 141–74. 63(1): 147–77.
Islam, Nazrul. 1995. “Growth Empirics: A Panel Data Mansoorian, Arman. 1991. “Resource Discoveries and
Approach.” Quarterly Journal of Economics, 110(4): ‘Excessive’ External Borrowing.” Economic Journal,
1127–70. 101(409): 1497–509.
Ismail, Kareem. 2010. “The Structural Manifestation Manzano, Osmel, and Roberto Rigobon. 2001.
of the ‘Dutch Disease’: The Case of Oil Exporting “Resource Curse or Debt Overhang?” National
Countries.” International Monetary Fund Working Bureau of Economic Research Working Paper 8390.
Paper 10/103. Matsen, Egil, and Ragnar Torvik. 2005. “Optimal
Karl, Terry Lynn. 1997. The Paradox of Plenty: Oil Dutch Disease.” Journal of Development Economics,
Booms and Petro-States. Berkeley and London: Uni- 78(2): 494–515.
versity of California Press. Matsuyama, Kiminori. 1992. “Agricultural Productivity,
Karl, Terry Lynn. 1999. “The Perils of the Petro-State: Comparative Advantage, and Economic Growth.”
Reflections on the Paradox of Plenty.” Journal of Journal of Economic Theory, 58(2): 317–34.
International Affairs, 53(1): 31–48. Mauro, Paolo. 1995. “Corruption and Growth.” Quar-
Kletzer, Ken, David M. G. Newbery, and Brian D. terly Journal of Economics, 110(3): 681–712.
Wright. 1992. “Smoothing Primary Exporters’ Price Mavrotas, George, S. Mansoob Murshed, and Sebastian
Risks: Bonds, Futures, Options and Insurance.” Torres. 2006. “Natural Resource Endowment and
Recent Economic Performance.” Unpublished. Resource Curse Hypothesis and Its Transmission
McPherson, Charles. 2004. “Managing Petroleum Rev- Channels.” Journal of Comparative Economics,
enues in Developing Countries.” Presented at the 32(1): 181–93.
Workshop Escaping the Resource Curse, Columbia Parente, Stephen L., and Edward C. Prescott. 1994.
University Workshop, 26/2. “Barriers to Technology Adoption and Devel-
Mehlum, Halvor, Karl Moene, and Ragnar Torvik. opment.” Journal of Political Economy, 102(2):
2002. “Plunder & Protection Inc.” Journal of Peace 298–321.
Research, 39(4): 447–59. Persson, Torsten, and Guido Tabellini. 2000. Political
Mehlum, Halvor, Karl Moene, and Ragnar Torvik. Economics: Explaining Economic Policy. Cambridge
2006a. “Cursed by Resources or Institutions?” World and London: MIT Press.
Economy, 29(8): 1117–31. Persson, Torsten, and Guido Tabellini. 2003. The Eco-
Mehlum, Halvor, Karl Moene, and Ragnar Torvik. nomic Effects of Constitutions. Cambridge and Lon-
2006b. “Institutions and the Resource Curse.” Eco- don: MIT Press.
nomic Journal, 116(508): 1–20. Pieschacón, Anamaría. 2009. “Oil Booms and Their
Miguel, Edward, Shanker Satyanath, and Ernest Ser- Impact through Fiscal Policy.” Unpublished.
genti. 2004. “Economic Shocks and Civil Conflict: An Pindyck, Robert S. 1980. “Uncertainty and Exhaustible
Instrumental Variables Approach.” Journal of Politi- Resource Markets.” Journal of Political Economy,
cal Economy, 112(4): 725–53. 88(6): 1203–25.
Mitra, Tapan. 1983. “Limits on Population Growth van der Ploeg, Frederick. 2010a. “Aggressive Oil
under Exhaustible Resource Constraints.” Interna- Extraction and Precautionary Saving: Coping with
tional Economic Review, 24(1): 155–68. Volatility.” Journal of Public Economics, 94(5–6):
Mitra, Tapan. 2002. “Intertemporal Equity and Effi- 421–33.
cient Allocation of Resources.” Journal of Economic van der Ploeg, Frederick. 2010b. “Voracious Transfor-
Theory, 107(2): 356–76. mation of a Common Natural Resource into Produc-
Murphy, Kevin M., Andrei Shleifer, and Robert W. tive Capital.” International Economic Review, 51(2):
Vishny. 1989. “Industrialization and the Big Push.” 365–81.
Journal of Political Economy, 97(5): 1003–26. van der Ploeg, Frederick. 2010c. “Why Do Many
Murphy, Kevin M., Andrei Shleifer, and Robert W. Resource-Rich Countries Have Negative Genuine
Vishny. 1993. “Why Is Rent-Seeking So Costly Saving? Anticipation of Better Times or Rapacious
to Growth?” American Economic Review, 83(2): Rent Seeking.” Resource and Energy Economics,
409–14. 32(1): 28–44.
Murshed, S. Mansoob. 2002. “Conflict, Civil War and van der Ploeg, Frederick, and Steven Poelhekke. 2009.
Underdevelopment: An Introduction.” Journal of “Volatility and the Natural Resource Curse.” Oxford
Peace Research, 39(4): 387–93. Economic Papers, 61(4): 727–60.
Neary, J. Peter. 1988. “Determinants of the Equilib- van der Ploeg, Frederick, and Steven Poelhekke. 2010.
rium Real Exchange Rate.” American Economic “The Pungent Smell of ‘Red Herrings’: Subsoil
Review, 78(1): 210–15. Assets, Rents, Volatility and the Resource Curse.”
Neary, J. Peter, and Douglas D. Purvis. 1982. “Sectoral Journal of Environmental Economics and Manage-
Shocks in a Dependent Economy: Long-Run Adjust- ment, 60(1): 44–55.
ment and Short-Run Accommodation.” Scandina- van der Ploeg, Frederick, and Dominic Rohner. 2010.
vian Journal of Economics, 84(2): 229–53. “War and Natural Resource Exploitation.” OxCarre
Neary, J. Peter, and Sweder J. G. van Wijnbergen, eds. Research Paper 42.
1986. Natural Resources and the Macroeconomy. van der Ploeg, Frederick, and Anthony J. Venables.
Cambridge, Mass.: MIT Press; Oxford: Blackwell. 2010. “Absorbing a Windfall of Foreign Exchange:
Newbery, David M. G. 1981. “Oil Prices, Cartels, and Dutch Disease Dynamics.” OxCarre Research Paper
the Problem of Dynamic Inconsistency.” Economic 52.
Journal, 91(363): 617–46. van der Ploeg, Frederick, and Anthony J. Venables.
Nordhaus, William D. 1992. “Lethal Model 2: The 2011. “Harnessing Windfall Revenues: Optimal
Limits to Growth Revisited.” Brookings Papers on Policies for Resource-Rich Developing Economies.”
Economic Activity, 2: 1–43. Economic Journal, 121(551): 1–30.
North, Douglass C. 1990. Institutions, Institutional Ramey, Garey, and Valerie A. Ramey. 1995. “Cross-
Change and Economic Performance. Cambridge; Country Evidence on the Link between Volatility
New York and Melbourne: Cambridge University and Growth.” American Economic Review, 85(5):
Press. 1138–51.
Olsson, Ola, and Heather Congdon Fors. 2004. “Congo: Robinson, James A., and Ragnar Torvik. 2005. “White
The Prize of Predation.” Journal of Peace Research, Elephants.” Journal of Public Economics, 89(2–3):
41(3): 321–36. 197–210.
Paldam, Martin. 1997. “Dutch Disease and Rent Seek- Robinson, James A., Ragnar Torvik, and Thierry
ing: The Greenland Model.” European Journal of Verdier. 2006. “Political Foundations of the Resource
Political Economy, 13(3): 591–614. Curse.” Journal of Development Economics, 79(2):
Papyrakis, Elissaios, and Reyer Gerlagh. 2004. “The 447–68.
Rodriguez, Francisco, and Jeffrey D. Sachs. 1999. Income in a General Equilibrium.” Review of Eco-
“Why Do Resource-Abundant Economies Grow nomic Studies, 73(1): 219–49.
More Slowly?” Journal of Economic Growth, 4(3): Skaperdas, Stergios. 2002. “Warlord Competition.”
277–303. Journal of Peace Research, 39(4): 435–46.
Rodrik, Dani, Arvind Subramanian, and Francesco Solow, Robert M. 1974. “Intergenerational Equity and
Trebbi. 2004. “Institutions Rule: The Primacy of Exhaustible Resources.” Review of Economic Stud-
Institutions over Geography and Integration in Eco- ies, 41: 29–45.
nomic Development.” Journal of Economic Growth, Stiglitz, Joseph E. 1974. “Growth with Exhaustible
9(2): 131–65. Natural Resources: Efficient and Optimal Growth
Rose, Andrew K., and Mark M. Spiegel. 2009. “Inter- Paths.” Review of Economic Studies, 41: 123–37.
national Financial Remoteness and Macroeconomic Stijns, Jean-Philippe C. 2005. “Natural Resource Abun-
Volatility.” Journal of Development Economics, 89(2): dance and Economic Growth Revisited.” Resources
250–57. Policy, 30(2): 107–30.
Ross, Michael L. 1999. “The Political Economy of the Stokke, Hildegunn E. 2008. “Resource Boom, Produc-
Resource Curse.” World Politics, 51(2): 297–322. tivity Growth and Real Exchange Rate Dynamics—
Ross, Michael L. 2001a. “Does Oil Hinder Democ- A Dynamic General Equilibrium Analysis of South
racy?” World Politics, 53(3): 325–61. Africa.” Economic Modelling, 25(1): 148–60.
Ross, Michael L. 2001b. Timber Booms and Institu- Svensson, Jakob. 2000. “Foreign Aid and Rent-
tional Breakdown in Southeast Asia. Cambridge; New Seeking.” Journal of International Economics, 51(2):
York and Melbourne: Cambridge University Press. 437–61.
Ross, Michael L. 2004. “What Do We Know about Tornell, Aaron, and Philip R. Lane. 1999. “The Voracity
Natural Resources and Civil War?” Journal of Peace Effect.” American Economic Review, 89(1): 22–46.
Research, 41(3): 337–56. Torvik, Ragnar. 2001. “Learning by Doing and the
Sachs, Jeffrey D., and Andrew M. Warner. 1995. “Nat- Dutch Disease.” European Economic Review, 45(2):
ural Resource Abundance and Economic Growth.” 285–306.
National Bureau of Economic Research Working Torvik, Ragnar. 2002. “Natural Resources, Rent Seek-
Paper 5398. ing and Welfare.” Journal of Development Econom-
Sachs, Jeffrey D., and Andrew M. Warner. 1997a. “Nat- ics, 67(2): 455–70.
ural Resource Abundance and Economic Growth.” Turnovsky, Stephen J. 1996. “Endogenous Growth in
In Leading Issues in Economic Development, First a Dependent Economy with Traded and Nontraded
edition, ed. Gerald M. Meier and James E. Rauch, Capital.” Review of International Economics, 4(3):
161–67. Oxford and New York: Oxford University 300–321.
Press. Velasco, Andres. 1999. “A Model of Endogenous Fis-
Sachs, Jeffrey D., and Andrew M. Warner. 1997b. cal Deficits and Delayed Fiscal Reforms.” In Fiscal
“Sources of Slow Growth in African Economies.” Institutions and Fiscal Performance, ed. James M.
Journal of African Economies, 6(3): 335–76. Poterba and Jürgen von Hagen, 37–57. Chicago and
Sachs, Jeffrey D., and Andrew M. Warner. 2001. “The London: University of Chicago Press.
Curse of Natural Resources.” European Economic Vicente, Pedro C. 2010. “Does Oil Corrupt? Evidence
Review, 45(4–6): 827–38. from a Natural Experiment in West Africa.” Journal
Sala-i-Martin, Xavier. 1997. “I Just Ran Two Million of Development Economics, 92(1): 28–38.
Regressions.” American Economic Review, 87(2): Vincent, Jeffrey R., Theodore Panayotou, and John
178–83. M. Hartwick. 1997. “Resource Depletion and Sus-
Sala-i-Martin, Xavier, and Arvind Subramanian. 2003. tainability in Small Open Economies.” Journal of
“Addressing the Natural Resource Curse: An Illus- Environmental Economics and Management, 33(3):
tration from Nigeria.” National Bureau of Economic 274–86.
Research Working Paper 9804. Warrack, Allan A., and Russell R. Keddie. 2002. “Natu-
Sandbu, Martin E. 2004. “Taxable Resource Revenue ral Resource Trust Funds: A Comparison of Alberta
Distributions: A Proposal for Alleviating the Natural and Alaska Resource Funds.” Western Centre for
Resource Curse.” Columbia University Center on Economic Research Information Bulletin 72.
Globalization and Sustainable Development Work- Weinstein, Jeremy M. 2005. “Resources and the Infor-
ing Paper 21. mation Problem in Rebel Recruitment.” Journal of
Sarraf, Maria, and Moortaza Jiwanji. 2001. “Beating Conflict Resolution, 49(4): 598–624.
the Resource Curse: The Case of Botswana.” World Wick, Katharina, and Erwin H. Bulte. 2006. “Con-
Bank Environmental Economics Series 83. testing Resources—Rent Seeking, Conflict and the
Sauré, Philip. 2010. “Overreporting Oil Reserves.” Natural Resource Curse.” Public Choice, 128(3–4):
Swiss National Bank Working Paper 2010-7. 457–76.
Sefton, J. A., and M. R. Weale. 1996. “The Net National van Wijnbergen, Sweder J. G. 1984a. “The ‘Dutch
Product and Exhaustible Resources: The Effects of Disease’: A Disease after All?” Economic Journal,
Foreign Trade.” Journal of Public Economics, 61(1): 94(373): 41–55.
21–47. van Wijnbergen, Sweder J. G. 1984b. “Infla-
Sefton, J. A., and M. R. Weale. 2006. “The Concept of tion, Employment, and the Dutch Disease in
il-Exporting Countries: A Short-Run Disequilib-

O conomic Growth in Resource-Rich Countries.” In
E
rium Analysis.” Quarterly Journal of Economics, Resource Abundance and Economic Development,
99(2): 233–50. ed. R. M. Auty, 76–92. Oxford and New York: Oxford
Withagen, Cees, and Geir B. Asheim. 1998. “Charac- University Press.
terizing Sustainability: The Converse of Hartwick’s World Bank. 2006. Where Is the Wealth of Nations?
Rule.” Journal of Economic Dynamics and Control, Measuring Capital for the Twenty-First Century.
23(1): 159–65. Washington, D.C.: World Bank.
Wood, Adrian J. B. 1999. “Natural Resources, Human Wright, Gavin, and Jesse Czelusta. 2002. “Exorcizing
Resources and Export Composition: A Cross- the Resource Curse: Minerals as a Knowledge Indus-
Country Perspective.” In Development Policies in try, Past and Present.” Unpublished.
Natural Resource Economies, ed. Jörg Mayer, Brian Wright, Gavin, and Jesse Czelusta. 2003. “Min-
Chambers, and Ayisha Farooq, 39–52. Cheltenham, eral Resources and Economic Development.”
U.K. and Northampton, Mass.: Elgar. Unpublished.
Woolcock, Michael, Lant Pritchett, and Jonathan Wright, Gavin, and Jesse Czelusta. 2004. “The Myth of
Isham. 2001. “The Social Foundations of Poor the Resource Curse.” Challenge, 47(2): 6–38.

Natural Resources: A Blessing or Curse? Factors Determining Economic Outcomes

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

Natural Resources: A Blessing or Curse? Factors Determining Economic Outcomes

Uploaded by

Copyright:

Available Formats

Journal of Economic Literature 2011, 49:2, 366–420

Natural Resources: Curse or Blessing?

1. Introduction such as Botswana or Norway, are more suc-

M any recognize the opportunities natu-

* University of Oxford, VU University of Amsterdam,

Figure 1. Growth and Natural Resource Dependence

Natural Produced Intangible

Low-income countries 1,925 1,174 4,434 7,532 26% 16% 59%

Note: All dollars at nominal exchange rates. Oil states excluded.

Percentage share of total wealth

Source: World Bank 2006, p. 29.

of natural gas found in Slochteren (M. Ellman yields P ε = H[F(1 − LN) + E]/G(LN), where

Figure 2. Natural Resource Dependence Reduces Competitiveness

Annual growth in real Based on data in Sachs Mehlum, Moene,

Initial income ­–1.76 (8.56) ­–1.28 (6.65) –1.26 (6.70)

be careful about drawing inferences about whether it is openness to international trade,

Entrepreneurs Rent seekers

Figure 3. Rent Grabbing and Producer Friendly Institutions

Ores and metals Mineral Production of gold,

Worst institutions –0.548 –0.946 –1.127 –1.145

increases in technical appropriability of Persson and Guido Tabellini (2003) and

of resource dependence is mainly driven by of growth that cannot be explained by the

separatist conflict. Lootable resources such a Heckscher–Ohlin model of international

Figure 4. Genuine Saving and Exhaustible Resource Share

• Trinidad and Tobago

Figure 5. Counterfactual Hartwick Rule

and εE ≡ ​ _

labor mobility, no technical progress, no 4.4 Fractionalization, Voracious Depletion,

Figure 6. Alternative Prescriptions for Harnessing Natural Resource Windfalls

​k​ ′N ​ ​ = −g′ PQ /g′′ < 0.

Ht = (1 + θL Tt−1)Ht−1, θ > 0 and [

xN = XN/L N H = G(1, kN) ≡ g( kN). = (1 − L Tt)g(kN(t)),

can be written as a stable difference equa- max–min egalitarianism which leads to a

Figure 7. Absorption Constraints and Dutch Disease Dynamics

Funds for Nonrenewable Resources.” International 89(3): 379–99.

­ il-Exporting Countries: A Short-Run Disequilib-

You might also like

Initial income –1.76 (8.56) –1.28 (6.65) –1.26 (6.70)

and εE ≡  _

k  ′N    = −g′ PQ /g′′ < 0.

Ht = (1 + θL Tt−1)Ht−1, θ > 0 and [

il-Exporting Countries: A Short-Run Disequilib-