Was Development Assistance a Mistake?
Total Page:16
File Type:pdf, Size:1020Kb
Was DevelopmentAssistance a Mistake? By WILLIAM EASTERLY* Development assistance is the combination development wisdom shifted away from mobi- of money, advice, and conditions provided by lizing and guiding capital accumulation.Atten- rich nations and internationalfinancial institu- tion, instead, shifted toward the success of the tions, such as the WorldBank and International East Asian tigers (South Korea, Taiwan, Hong Monetary Fund (IMF), which is designed to Kong, and Singapore),which combined export achieve economic developmentin poor nations. orientation and macroeconomic stability. This This article argues that developmentassistance became the inspiration for structural adjust- was based on three assumptions that, with the ment packages of the IMF, the WorldBank, and benefitof hindsight(although a wise few also had the "WashingtonConsensus," which called for foresight),turned out to have been mistaken. removing price distortions,opening up to trade, and correcting macroeconomic imbalances I. We KnowWhat ActionsAchieve (mainly budget deficits).The slogan of the new EconomicDevelopment wave was "adjustmentwith growth." Alas, loans to finance structural adjustment Development economists long have known met the same fate in low-income countries as the answers of how to achieve economic devel- the earlier loans to finance investment-there opment. The only problemis that those answers was little or no growth; the loans could not be have continued changing over time. repaid;and the low-income debt crisis stretched To oversimplify,the evolutionof Conventional out into the new millennium with every year Wisdom is as follows (see also DavidL. Lindauer bringing a new wave of debt forgiveness (most and Lant Pritchett2002; the WorldBank 2005; recently, the cancellation of the structural and Dani Rodrik2006). In the 1950s throughthe adjustmentloans in the MultilateralDebt Relief 1970s, development(i.e., economic growth)was Initiative of 2006). In the middle-incomecoun- a simple matterof raising the rate of investment tries of Latin America, there was, for the most to Gross Domestic Product (GDP), including part, adjustmentand debt repayment,but little public investments for roads, dams, irrigation growth comparedto expectations in the 1990s. canals, schools, and electricity, as well as pri- The hope that the "East Asian miracle" could vate investment. Private investment, however, be replicated elsewhere with the same policies was usually not trusted to do enough or to do proved illusory. The Washington Consensus the right things, and so there was a strong role then gave way to second generation reforms for the state to facilitate and direct investment, that stressed the importance of institutions guided by the developmentexperts. such as property rights, contract enforcement, Unfortunately, the debts accumulated to democratic accountability, and freedom from finance these investments turned out not to be corruption. repayable.So, there were two debt crises during Although each shift in the conventionalwis- the 1980s. Middle-income countries had bor- dom was provokedby the failure of the previous rowed from commercial banks at market rates, conventionalwisdom, the argumentwas usually while low-income countrieshad loans from offi- that previous recommendations were "neces- cial agencies at concessional rates. Both entered sary but not sufficient."As Rodrik (2006) points into a long process of rescheduling and writing out, this had the effect of placing all the blame off debt that led to a lost decade for both groups on the recipient ratherthan on the development of debtors.Understandably, inferring that unre- experts, of making the list of "sufficientcondi- payableloans were a sign of unproductiveinvest- tions" for developmentever longer,and of mak- ments, especially in Latin America and Africa, ing the conventionalwisdom nonfalsifiable. A lot of these shifts were provokedby broad * New York University, Room 510, 110 Fifth Avenue, stylized facts and compelling country examples New York,NY 10011 (e-mail: [email protected]). rather than by formal empirics. Development 328 This content downloaded from 128.122.62.112 on Thu, 1 Aug 2013 15:15:03 PM All use subject to JSTOR Terms and Conditions VOL.97 NO. 2 WASDEVELOPMENTASSISTANCE A MISTAKE? 329 knowledge could draw upon more formal record of achieving development (as opposed empirics like growth regressions. However,the to, say, totalitarian control of the economy by hope that arose in the early 1990s that the New kleptocrats).It is just that we don't know how Growth Literature at least empirically could to get from here to there;which specific actions find the answers eventually collapsed from a contributeto free marketsand good institutions; surplus of answers. Steven N. Durlauf, Paul A. how all the little pieces fit together.That is, we Johnson, and R. W. Temple (2005) pointed out don't know how to achieve development. that 145 differentright-hand-side variables were significantas determinantsof growth in various II. Our Adviceand MoneyWill Make Those studies with around 100 degrees of freedom. CorrectActions Happen When the problemsof unrestrictedspecification were reducedby testing the outcomes of the key Using the same judgment by stylized facts WashingtonConsensus variables on growth, the and country cases that has guided the evolu- results tended to confirm the casual empiricism tion of the conventional wisdom, development described above-countries as a group moved assistance has failed to achieve development. toward "better policies," yet average growth Over the past 42 years, $568 billion (in today's for that group declined for unknown reasons dollars) has flowed into Africa, yet per capita (Easterly2001). growth of the median African nation has been In the new millennium, a remarkablybroad close to zero. The top quarterof aid recipients group of academics and policymakers seem to (heavily overlapping with Africa) received 17 agree that, after all, maybewe don'tknow how to percent of their GDP in aid over those 42 years, achieve development,although they are reluctant yet they also had near-zeroper capita growth. to say so exactly. The World Bank (2005) was Successful cases of developmenthappening due either giving up or offering instantaneousnon- to a large inflow of aid and technical assistance falsifiability: "Different policies can yield the have been hard to find. South Korea is often same result, and the same policy can yield dif- cited, but it took off after aid was reduced, and ferentresults, dependingon countryinstitutional the Koreans disregardedthe advice of the aid contexts and underlyinggrowth strategies."The donors (see JamesFox 2000). Othermore recent BarcelonaDevelopment Agenda (2004), a who's examples frequentlycited (Ghana,Uganda, and who of leading economists,concluded that "there Mozambique)were cases of recovery after steep is no single set of policies that can be guaranteed collapse, and depend on rapid growth episodes to ignite sustained growth. Nations that have that usually prove to be temporary (Ricardo succeeded at this tremendouslyimportant task Hausman,Rodrik, and Pritchett 2005). Botswana have faced different sets of obstacles and have might be a better example of a long-term suc- adopted various policies regarding regulation, cess story initially financed by aid, although exportand industrialpromotion, and technologi- the most well-known case study of Botswana cal innovationand knowledgeacquisition." (Daron Acemoglu, Simon Johnson, and James Lindauer and Pritchett (2002) call it most A. Robinson2004) doesn'teven mentionforeign honestly: "It seems harderthan ever to identify aid. The currentlymost celebratedcases of rapid the keys to growth. For every example, there is a growth-India, China, and Vietnam-receive counterexample.The currentnostrum of one size little aid as a percentageof their GDP. doesn't fit all is not itself a big idea, but a way of With aid, one has an even more serious expressing the absence of any big ideas." problem than with other growth regressions of This does not mean that economists know endogeneity of the right-hand-sidevariable. It's nothing about development,or that they know very likely that low-growth countries got more nothing aboutthe manylittle pieces thatcontrib- aid because they had low growth. This calls for ute to development.Good economic analysis of more formal econometric methods to disen- problems in finance, macroeconomics,taxation tangle the aid outcome from the counterfactual, and public spending,health, agriculture,etc. has utilizing instruments such as population size held up well. Economists are reasonablyconfi- and geostrategic factors. Unfortunately, more dent that some combinationof free marketsand formal empirics on the effect of aid on growth good institutionshas an excellent historicaltrack have suffered from the same problem as other This content downloaded from 128.122.62.112 on Thu, 1 Aug 2013 15:15:03 PM All use subject to JSTOR Terms and Conditions 330 AEA PAPERSAND PROCEEDINGS MAY2007 the supportof their publics and the assent growth regressions-too many possible specifi- of their when do not need cations and not observations legislatures they enough (to begin to raise revenuesfrom the local economy, aid did not even make with, Durlauf, Johnson, as long as they keep the donors happyand and Temple's 2005 list of 145 statistically sig- willing to provide alternative sources of nificant variables appearing in growth regres- funding." sions). Aid and control variables have included