292 32841 privatesector P U B L I C P O L I C Y F O R T H E NUMBER NOTE 2005 Aid Effectiveness JUNE Michael Klein and Can Aid Agencies Be Smarter Than the Invisible Hand? Tim Harford Private financial flows such as foreign direct investment seem to Michael Klein encourage economic growth and relieve poverty in part because they (mklein@worldbank.org) is chief economist at the create excellent incentives for transferring know-how and in part International Finance because they are subject to a stern market test that ensures they are Corporation (IFC) and joint IFC­World Bank allocated and monitored carefully. For aid flows, not automatically vice president for private PRESIDENCY subject to these disciplines, it is difficult to be as effective. This sector development. Note argues that aid agencies, by learning what makes private flows VICE Tim Harford (tharford@ifc.org) is so effective, can bring better aid to the poorest. an economist at the World Bank and IFC. Aid is no longer the major source of money vate and official flows to developing countries, flowing into developing countries, which now assess the indirect impact on growth, through DEVELOPMENT This Note is part of a enjoy inward foreign direct investment, remit- government spending. They look at private series exploring trends tances from migrant workers, loans from banks loans and equity flows to the private sector, gov- in the aid industry, SECTOR and the bond markets, and substantial grants ernment borrowing from the private sector, and including patterns of aid flows, competition, and from nongovernmental organizations (see remittances--and the results look encouraging. the effectiveness of Note 290 in this series). These diverse sources The private flows do not raise government con- PRIVATE different types of aid. of money are certainly not perfect substitutes sumption (which is good, since high govern- for aid. They flow to different countries and to ment consumption dampens growth), but they different people within countries, on different seem to increase total investment, which accel- cycles, bundled with different ideas and with erates growth. GROUP different aims--and, not surprisingly, have dif- This is not surprising. Ever since Adam Smith ferent results. What do they teach us about conjured the image of the "invisible hand," the BANK smarter aid? ability of private enterprise to generate wealth for all has been widely appreciated. So should Private finance and growth we abandon official aid and rely on private First consider results. A large body of research finance? Of course not, for several reasons: WORLD argues that private-to-private flows directly boost Private finance flows only to where money is economic growth. Djankov, Montalvo, and to be made (see Note 290). That inflows of THE Reynal-Querol (2004), in a new study of both pri- portfolio equity increase economic growth A I D E F F E C T I V E N E S S C A N A I D A G E N C I E S B E S M A R T E R T H A N T H E I N V I S I B L E H A N D ? or that foreign direct investment can be a Why does aid seem to be falling short of its superb way to transfer technology is no com- potential? The answer may be macroeconomic. fort to the victim of a tsunami or a civil war. For example, foreign aid may damage the com- The technology transferred by private petitiveness of the industries that developing finance is focused only on increasing the pro- countries would most expect to fuel growth. ductivity of particular supply chains. Private Rajan and Subramanian (2005) provide innova- investors have little interest in building rural tive evidence for this explanation. roads or helping to design a legal system. An alternative explanation is that the failure of Even if private finance worked superbly, we aid to promote growth is a statistical illusion: pri- 2 would always hope to do more to reduce vate for-profit flows seek successful economies, poverty. aid flows seek struggling ones, and statistical analysis has failed to fully control for this. Why does aid fall short? But matching the benchmarks carved by the Benevolent, monitored, and "smart" invisible hand of private finance is difficult for A possibility equally worth considering is that official aid. One problem is that aid flows may some private flows outperform aid because they weaken governance in developing countries, generate the right incentives for those sending for example, by triggering a political struggle and receiving the money, incentives that are key to control the cash (see Note 291 in this to producing development results. At the risk of series). some simplification, it could be speculated that A second problem is that official aid does not the development results of financial flows into seem to boost economic growth, at least not as developing countries depend on three qualities irrefutably as private finance does. Djankov, (table 1): Montalvo, and Reynal-Querol (2004) find that How benevolent is the finance? The most benev- official development assistance as a whole olent finance would flow to the poorest peo- directly reduces economic growth and also indi- ple in the poorest countries exactly when they rectly retards growth by increasing government need it and would never need to be repaid. consumption and reducing total investment as How well is the finance monitored? Perfectly a share of GDP. This finding is broadly consis- monitored financial flows would go exactly tent with recent research on aid effectiveness, where their owners want them to go. which fails to find a strong positive effect of aid Imperfectly monitored flows might be spent on economic growth. Djankov and his co- on pet projects, stolen, or wasted. authors also find that when a recipient's aid is How much knowledge flows with the finance? mostly in the form of grants, the negative effects Knowledge matters, whether provided as stand- on growth are more severe. alone advice or alongside financial flows. Much These findings come from cross-country official aid is bundled with technical advice, but regressions, always treacherous. One possible some private flows like foreign direct invest- complication is that bilateral agencies in the ment also come with advice and training. past tended to hand out mostly grants, while The ideal development assistance would be multilateral agencies typically used loans--so benevolent, monitored, and "smart" in the the apparent superiority of loans may result in sense of providing valuable know-how, but part because multilaterals give more effective finance that falls short of this ideal can still be aid, whether grants or loans. Moreover, what was hugely useful. For example, the bond markets true in the past, especially during the cold war, provide developing countries with finance that may not be true in the future. And of course eco- is indifferent to development results and con- nomic growth is not the only measure of devel- tains no advice but without which the borrowers opment results, nor the aim of all development would surely be poorer. Workers' remittances assistance. Nevertheless, the results are in- are "dumb" too, but since they are well meaning triguing and consistent with earlier research and well aimed, small wonder that development (see Note 287 in this series). professionals are beginning to be excited by their potential to relieve poverty. Foreign direct Table Which financial flows have the key qualities? investment is entirely indifferent to the devel- 1 opment results it may produce--but those Qualities Example results often materialize where the investment is Benevolent, monitored, smart Ideal development assistance packaged with cutting-edge technological know- Indifferent, monitored, smart Foreign direct investment how. The question is, how to provide aid agen- Benevolent, unmonitored, smart Careless development assistance cies and recipients with the same incentives to Benevolent, monitored, dumb Workers' remittances teach, learn, and achieve results as are automat- Indifferent, monitored, dumb Bond market finance Benevolent, unmonitored, dumb Populist emergency aid ically created by the invisible hand? Can aid flows be monitored as well as private flows? Can aid flows be as smart as foreign direct It could be argued--though it's hard to prove-- investment? that many private flows are monitored better The smartest finance, in the right circum- than official aid flows. In some cases the moni- stances, is foreign direct investment. Many stud- toring is inherent: a migrant's remittances to his ies confirm this. Consider the manufacturing of family back home can be monitored because the car seats. Sutton (2005) finds that multinational personal relationship is so close. (And some joint ventures in India can bring error rates Mexican migrants open store credit accounts for down quickly: one new factory moved from their relatives rather than sending cash, so they 2,085 errors per million to 65 in just three years. can check that the money is spent prudently.) Domestic firms in India, emulating the multi- But most financial flows are flows of other nationals and using knowledge gained as their people's money. Aid agencies spend taxpayers' suppliers, can also make dramatic progress: one money. Private banks lend depositors' money. manufacturing firm eliminated 99 percent of Multinational companies invest shareholders' errors within five years of adopting cutting-edge money. Inevitably, such money will be assigned techniques. with less care than the personal income of those There is no mystery why foreign direct invest- who manage these flows--the bureaucrats, ment sometimes provides such tremendous bankers, or managers. expertise: What stops complete chaos? Competition The investing firm has strong incentives to and good governance. improve the expertise of local workers, sup- Is an organization like the World Bank or pliers, regulators, and partners. United Nations Development Programme gov- Workers and suppliers have a lot to gain erned less well than a publicly listed private com- (high salaries, lucrative contracts), and pany? Evidently not, if the company is Enron or multinational auto manufacturers are care- Parmalat. Yet companies will always have a gov- ful to monitor closely and reward excellence. ernance edge over governmental organizations: The largest alternative source of smart investors and depositors have a wide choice on finance is development assistance coupled with where to put their money, which sharpens the technical assistance, largely advisory services pressure to offer excellent corporate gover- and training. For example, a loan to fund con- nance and increases the returns to analysts and struction of a new electricity grid and power sta- rating agencies that review that governance. tions would come with advice on the scale and Given sufficient competitive pressure, the mar- design of the new grid and on the regulatory ket should deliver good corporate governance: framework needed to make it run efficiently. Enron went bankrupt; aid agencies never do. The potential of smart, well-monitored, and Since aid agencies are subject to neither intense benevolent development assistance is huge. But competitive pressure nor scrutiny from analysts development assistance may not create the same and rating agencies, maintaining the highest incentives as foreign direct investment--for the standards of corporate governance therefore recipients or the technical experts--to make the requires all the more effort and goodwill. lessons stick. A I D E F F E C T I V E N E S S C A N A I D A G E N C I E S B E S M A R T E R T H A N T H E I N V I S I B L E H A N D ? The strongest incentives to learn arise if the for themselves--which implies embracing grow- flow of money will be cut off for those who fail to ing competition in the industry (see Note 277 in do so. This condition may apply for a supplier of this series) and encouraging searching evalua- a multinational corporation, but not typically for tions by disinterested rating agencies. anaidrecipient.Foreigndirectinvestmentisauto- Second, foreign direct investment often cre- viewpoint matically subjected to a tough market test, but aid ates projects dependent on rapid technology projects are not. If aid agencies want a tough test, transfer. Everyone involved has a strong incen- they have to create it themselves using more rig- tive to learn or to teach as appropriate, and a is an open forum to orous evaluation (such as the randomized trials stern market test weeds out those who do not. encourage dissemination of done by the pharmaceutical industry) or rigorous Aid projects are not subject to any such test public policy innovations for benchmarks of the quality of a country's gover- unless the agency decides to use rigorous evalu- private sector­led and nance (such as the World Bank's Country Policy ation. Sovereign loans for aid projects are typi- market-based solutions for and Institutional Assessment, or CPIA). cally repaid regardless of the quality of the development. The views The strongest incentives to teach arise if technical assistance; nonsovereign loans and published are those of the repayment is impossible unless the technical equity investments provide sharper incentives to authors and should not be assistance works (or if recipients pay directly for ensure that appropriate lessons are learned. attributed to the World the advice). This condition may apply for aid Unfortunately, the most desperate situa- Bank or any other affiliated agencies that make nonsovereign loans and tions, where we would want to help the poorest organizations. Nor do any of equity investments or stand-alone consulting with grants and free advice, are also those where the conclusions represent arrangements, but an aid agency making a it is most difficult to maintain the appropriate official policy of the World sovereign-guaranteed loan knows that the likeli- disciplines and make aid effective. But one pos- Bank or of its Executive hood of repayment has little to do with the suc- sibility is to give performance-based grants (see Directors or the countries cess of the project. Note 270). they represent. In all cases aid agencies can raise their Remedies game by providing better information about To order additional copies Aid agencies strive to provide aid that is carefully how aid is being spent and embracing competi- contact Suzanne Smith, monitored and bundled with high-quality tech- tive pressures. High-quality evaluation of proj- managing editor, nical assistance, but more can always be done. Aid ects and of aid agencies, combined with the Room F 4K-206, flows could be as well motivated and as tightly political will to deliver better aid, should in a The World Bank, 1818 H Street, NW, monitored as the remittances workers send back competitive environment improve the gover- Washington, DC 20433. to their families--and could carry the expertise nance of aid agencies and the quality of the aid of foreign direct investment, the kind that can they deliver. In a well-functioning market for Telephone: induce a hundredfold reduction in errors in just aid, aid agencies can perform their roles even 001 202 458 7281 a few years in India. How could that happen? better than the invisible hand. Fax: Aid agencies should learn from the distinc- 001 202 522 3480 tive advantages of different private flows. Email: Remittances seem to be well aimed and well ssmith7@worldbank.org timed. Radical proposals to give aid vouchers to References the poor are usually dismissed--but these aim at Djankov, Simeon, Jose G. Montalvo, and Marta Reynal- Produced by Grammarians, providing purchasing power and choice to those Querol. 2004. "Helping the Poor with Foreign Aid: The Inc. who need it most, when they need it, with a min- Grants vs. Loans Debate." World Bank, Washington, D.C. imum of waste and misdirection. Rajan, Raghuram G., and Arvind Subramanian. 2005. Printed on recycled paper Foreign direct investment has even more to "What Might Prevent Aid from Enhancing Growth?" teach aid agencies. First, the standards of gover- International Monetary Fund, Washington, D.C. nance required to make such investments work Sutton, John. 2005. "Competing in Capabilities: An are high and rising. Foreign investors are mer- Informal Overview." First Development Economics cilessly unforgiving of poorly governed coun- Lectures, World Bank, Washington, D.C., April 21. tries and partners. Aid agencies, if they wish to help the poorest, cannot be so choosy, but they can seek the highest standards of governance T h i s N o t e i s a v a i l a b l e o n l i n e : h t t p : / / r r u . w o r l d b a n k . o r g / P u b l i c P o l i c y J o u r n a l