Resource intensity, institutions, and developmentIntroductionThe resource curse andits implicationsData and empiricalprocedureResults of regressionanalysesConclusions and discussionReferencesData sources and summary statisticsResource Intensity, Institutions, and DevelopmentERWIN H. BULTETilburg University, NetherlandsRICHARD DAMANIAUniversity of Adelaide, AustraliaandROBERT T. DEACON*University of California, Santa Barbara, USASummary. — We examine the relationship between resource abundance and several indicators ofhuman welfare. Consistent with the existing literature on the relationship between resource abun-dance and economic growth we find that, given an initial income level, resource-intensive countriestend to suffer lower levels of human development. While we find only weak support for a direct linkbetween resources and welfare, there is an indirect link that operates through institutional quality.There are also significant differences in the effects that resources have on different measures of insti-tutional quality. These results imply that the ‘‘resource curse’’ is a more encompassing phenomenonthan previously considered, and that key differences exist between the effects of different resourcetypes on various aspects of governance and human welfare.Ó 2005 Elsevier Ltd. All rights reserved.Key words — resource curse, human development, resource wealth and scarcity1. INTRODUCTIONConventional economic reasoning suggeststhat increasing a country’s stock of assets pro-vides greater opportunities for economicgrowth. Somewhat paradoxically, a substantialbody of empirical evidence demonstrates thatnatural resources tend to hinder, rather thanpromote economic growth. The seminal andinfluential studies of Sachs and Warner (1997,2001) show that after controlling for a widevariety of variables, an increase of one standarddeviation in natural resource intensity leads toa reduction of about 1% per year in economicgrowth. This result has been coined the ‘‘re-source curse,’’ and it inspired a large volumeof subsequent empirical research. While the re-sults are robust with respect to the inclusion ofmany conditional variables, they have also beenchallenged (e.g., by Manzano & Rigobon, 2001,who focus on debt overhang, and Stijns, 2002,who emphasizes learning processes).Economic growth per se is a poor indicator ofwelfare. It is conceivable that even if natural re-sources are a curse for economic growth nar-rowly defined, they may lead to improvementsin other aspects of welfare—such as the preva-lence of poverty, malnutrition, and infant mor-tality. It is an open question to what extentgrowth dividends, if any, percolate to otherperhaps more vulnerable members of society.*We are grateful to three anonymous reviewers forhelpful comments and suggestions. Research supportfrom the Agricultural and Development EconomicsDivision (ESA) at the United Nations Food and Agri-cultural Organization is gratefully acknowledged. Theviews expressed in this report are those of the authorsand do not represent the position of FAO. Final revisionaccepted: January 21, 2005.World Development Vol. 33, No. 7, pp. 1029–1044, 2005Ó 2005 Elsevier Ltd. All rights reservedPrinted in Great Britain0305-750X/$ - see front matterdoi:10.1016/j.worlddev.2005.04.004www.elsevier.com/locate/worlddev1029The record of growth in recent decades showsthat many countries with low per capita growthrates have succeeded in providing food securityand meeting basic nutritional needs, while oth-ers with higher growth rates have failed (Tho-mas, Daclam, Dhareshwar, Kaufman, &Lopez, 2000). The link between resource abun-dance and underdevelopment is therefore un-known, a priori.The main objective of this paper is to ana-lyze whether the negative statistical relation-ship between natural resource abundance andeconomic growth has a parallel in measuresof economic underdevelopment and welfare.While underdevelopment and welfare clearlyare not independent of economic growth, thereare important differences between these vari-ables. Underdevelopment and welfare indica-tors are typically expressed as ‘‘levels,’’whereas economic growth is measured as achange in levels over time.1Levels capturedifferences in economic performance over longtime periods, and are directly relevant for wel-fare as measured by the consumption of goodsand services (Hall & Jones, 1999). Empiricalwork by Easterly, Kremer, Pritchett, and Sum-mers (1993) reports relatively low correlationof growth rates across decades, suggesting thatdifferences in growth rates across countriesmay be transitory and that a focus on ‘‘levels’’may be appropriate. Moreover, underdevelop-ment and welfare indicators capture distribu-tional considerations that are not capturedby aggregate growth statistics. For example,some indicators capture the population sharefor which the so-called basic human needsare met. This information complements in-come growth statistics to provide a more com-plete picture of the effect of resources on wellbeing in society (also see Acemoglu, Johnson,& Robinson, 2001b, and in particular Davis,1995, who compares indicators of develop-ment for mineral and nonmineral exporters,finding no evidence that mineral exporters per-form worse).The paper is organized as follows. In Section2, we review the theoretical and empirical liter-ature on the relationship between resourceabundance and economic growth. Building onthis literature, we develop an empirical modelto analyze the impact of resources on underde-velopment and welfare. This model is outlinedin Section 3. In Section 4, we present theempirical results and discuss the main implica-tions for development policies. Section 5 con-cludes.2. THE RESOURCE CURSE ANDITS IMPLICATIONSWhy might resource-rich countries growslower than resource-poor ones? In the 1950s,several ‘‘structuralist’’ explanations were devel-oped, focusing on declining terms of trade forprimary commodities (e.g., Prebisch, 1950),sharp fluctuations in the prices of such com-modities, or lack of linkages between resourceextraction enclaves and the rest of the economy(Hirschman, 1958). None of these explanationshave stood the test of closer empirical scrutiny(e.g., Behrman, 1987; Cuddington, 1992; Dawe,1996; Fosu, 1996; Lutz, 1994; Moran, 1983;Tan, 1983).In more recent times, economists and politi-cal scientists have advanced new theories to ex-plain the disappointing growth performance
View Full Document
Unlocking...