The Macroeconomic Management of Foreign Aid : Opportunities and Pitfalls

Boriana Yontcheva, Peter Isard, Leslie Lipschitz, and Alex Mourmouras

• Publisher: INTERNATIONAL MONETARY FUND
• Year: 2006
• eBook ISBN: 9781589065208

NONE

1 Foreign Aid and Macroeconomic Management

Overview and Synopsis of Key Lessons

Peter Isard, Leslie Lipschitz, Alexandros Mourmouras, and Boriana Yontcheva *

I. Introduction

Since the adoption of the Millennium Development Goals (MDGs) in 2000, the challenge of reducing poverty around the world has been more prominent on the agenda of the international community.¹ Relatively slow progress toward meeting the MDGs by the 2015 target date has added to the urgency of this effort. Two influential reports—the United Nations Millennium Project Report (the “Sachs Report”) and the Commission for Africa Report (the “Blair Report”)—envisage substantial increases in aid flows to poor countries, especially to countries in sub-Saharan Africa. The international community sees increases in aid, along with improvements in recipient policies and freer global trade, as necessary for global prosperity and poverty reduction.

While it seems clear that the MDGs cannot be met without a substantial scaling-up of foreign aid, historical experience provides reason to question whether large increases in aid will translate into large strides toward the MDGs. Aid has indeed facilitated some remarkable successes in helping countries reconstruct or develop rapidly. Major beneficiaries have included the European recipients of Marshall Plan aid following the Second World War, Taiwan Province of China in the 1950s, Botswana and Korea in the 1960s, Indonesia in the 1970s, Bolivia and Ghana in the late 1980s, and Uganda and Vietnam in the 1990s. Aid has also been instrumental in eradicating certain diseases (e.g., river blindness) and clearly has the potential to facilitate large advances toward the MDGs for health and education. But the idea that countries can be lifted out of poverty simply by giving them more foreign aid belies half a century of experience.² To maximize the benefits of aid, policymakers and aid donors need to be aware of the important complementarities between aid, policies, and institutions, and to be cognizant of potential macroeconomic hazards to avoid.

The papers in this volume were prepared for a high-level seminar on “Foreign Aid and Macroeconomic Management,” which was organized by the IMF Institute and the IMF’s African Department, with cofinancing from the United Kingdom’s Department for International Development (DFID) and Germany’s Internationale Weiterbildung und Entwicklung gGmbH (InWEnt). The seminar was hosted in Maputo by the Government of Mozambique in March 2005. In opening the seminar, Prime Minister Luisa Dias Diogo emphasized that issues of macroeconomic management were complex and not conducive to simple solutions, that these issues were likely to become more challenging with a substantial scaling-up of aid, and that high-level seminars and other efforts aimed at strengthening the capacity for macroeconomic management were very important. Abdoulaye Bio-Tchané (IMF African Department) expressed similar views in his opening remarks, stressing the importance of strengthening not only the specific institutions of macroeconomic management but also the basic institutions and governance processes that support the rule of law.

The seminar brought together high-level African officials, experts from the academic community, and policymakers from the aid-donor community. While many issues relating to the efficient and effective management of aid were discussed in passing—for example, the scale of financial flows needed to meet the MDGs, and the short-run trade-offs between growth-enhancing infrastructure spending and social spending on health and education—the main focus of the seminar was on how to identify and forestall any adverse macroeconomic effects from a large-scale increase in aid flows. Do aid-receiving countries have the capacity to absorb a substantial scaling-up of aid flows? Does donor specification of the timing of spending and the precise sectoral allocation exacerbate the absorption problem? Could a substantial increase in aid shift relative prices in a way that would be detrimental to export competitiveness and longer-term growth prospects? What microeconomic constraints underlie the potential macroeconomic hazards from scaling up aid? Will the scaling-up of aid exacerbate its volatility and unpredictability? What about the impact of higher aid flows on the debt sustainability of low-income recipients? And what is the effect of aid on institutions and the political economy in recipient countries?

This chapter provides a summary and synthesis of the perspectives that emerged from the papers and discussions in Maputo. Section II places the discussion of the macroeconomic hazards of aid in a broader context. It examines the nexus between aid, growth, and poverty reduction, which was the topic of the first session of the seminar. Section III of this chapter considers the issue of aid-induced Dutch disease—that is, the possible link between aid inflows and a deterioration of the recipient country’s competitiveness in global markets—and the associated challenge of mitigating absorptive capacity constraints. It also addresses the trade-offs between different categories of aid-financed spending and the case for frontloading development assistance. This is followed by discussions of the volatility and unpredictability of aid in Section IV, and of debt sustainability and the choice between loans and grants in Section V. Section VI turns to the connection between aid and institutional quality along with the themes that rang through the three keynote speeches. Section VII summarizes the perspectives provided by the panelists in the final session of the seminar, and Section VIII pulls together the main policy messages that emerged from the discussions. The Appendix contains comments on some technical issues raised by the papers presented in Maputo.

II. Aid, Growth, and Poverty Reduction

Aid and Growth

The literature on the effectiveness of aid in promoting growth and poverty reduction is large and inconclusive: thus far, empirical studies have not found strong evidence of a robust and positive aid-growth relationship. The econometric work presented in Maputo provided new perspectives but did not lead to consensus on the effectiveness of aid in promoting growth. Some of the new evidence indicated that the relationship between certain types of aid and growth was more favorable than earlier literature had suggested; other empirical work painted a more pessimistic picture. The discussion was lively, and the debate remained unresolved.

The paper presented by Steven Radelet (Center for Global Development)³ challenged the disappointing results of most aid effectiveness studies by questioning the appropriateness of basing analysis on an aggregate of all types of aid, which is the standard approach used in the literature. The authors (Radelet, Clemens, and Bhavnani) stress that many types of aid either are given for worthy causes that have little to do with economic growth (e.g., humanitarian relief after floods, earthquakes, or other natural disasters; and aid to help democracy) or contribute to growth with long lags that short-run data cannot capture (e.g., spending on education and health). By contrast, aid to fund basic infrastructure or to support agriculture and industry ought to elicit higher growth after fairly short time lags. The absence of a strong link between aid and growth in the findings of many studies may reflect the failure to distinguish between these different types of aid.

Radelet and his coauthors therefore disaggregate aid into three types: humanitarian assistance; early-impact aid to finance infrastructure and direct investments in agriculture and other sectors; and late-impact aid to finance investments in human and social capital. Worldwide, early-impact aid averaged about 2½ percent of recipient country GDP between 1973 and 2001, and about 5 percent of GDP in Africa. Drawing on previous work,⁴ Radelet’s paper reported strong and robust evidence that early-impact aid exerts a powerful effect on growth. For the typical recipient, a 1 percentage point of GDP increase in this type of aid produces an additional 0.31 percentage point of annual growth over a four-year period. This is two to three times higher than previously estimated. By contrast, Radelet and coauthors find no clear influence of long-impact aid on growth and a negative relationship between humanitarian relief and growth, as the latter category of aid is often triggered by adverse growth shocks. They also find that the effect of aid on growth is more pronounced in countries with stronger institutions and longer life expectancy.

Radelet argued that high aid-to-GDP ratios in Africa ought not to alarm us: they reflect low incomes rather than generous aid.⁵ Crucially, however, the effect of aid on growth is subject to diminishing returns, so that raising aid beyond some point does not add appreciably to growth. He estimated that, for the typical country, the aid saturation point is 8–9 percent of GDP for early-impact aid; this corresponds approximately to 16–18 percent of GDP for total aid. For some countries, these results may be good news: they suggest that there is substantial room to scale up aid from present levels without bumping into severe capacity constraints. However, the influence of each additional dollar of aid will be lower than that of the previous dollar.

In discussing Radelet’s paper, Arvind Subramanian (IMF Research Department) emphasized the need for caution in assessing the growth effects of large aid packages and challenged Radelet’s findings based on evidence from Rajan and Subramanian (2005). The latter paper examines whether labor-intensive industries (those in which a poor, developing country should have a comparative advantage) grow relatively more slowly than other industries in countries that receive more aid. It employs an empirical methodology that exploits variation both within countries and across countries, and finds that aid inflows systematically undermine the competitiveness of labor-intensive export sectors, as reflected in a decline in the share of labor-intensive and tradable goods industries in the manufacturing sector. This result is not surprising at the theoretical level; it is consistent with most conventional models and with recent independent work by Arellano and others (2005) based on a calibrated real business cycle general equilibrium model. But the Rajan and Subramanian paper is remarkable for the robustness of the detrimental allocative effects of aid and the finding that these may occur without appreciable changes in the real exchange rate. Any link between aid and reduced export competitiveness is extremely important, Subramanian argued, because export-led growth has been so central to most successful development strategies.

Several participants raised a separate challenge to Radelet’s conclusions based on the notion that aid is fungible. If aid is fungible—that is, if recipient governments can reallocate domestic resources in response to aid inflows—attempts to channel it to specific projects or sectors will be frustrated, the composition of public spending will remain that favored by the government, and the intended growth effects of aid might not be realized. Fungibility would not be an issue if the government’s objective were to maximize social welfare and if donors and recipients had completely coincident views on how to do so. But in many developing countries the composition of public spending is biased against the poor and against growth. One explanation (Olson, 1965) recognizes the political pressure incumbents face—even in democratic regimes—from organized interest groups. These groups have power to influence the allocation of public resources in their favor, which is welfare-reducing as far as the general public is concerned (Dixit, Grossman, and Helpman, 1997).

Radelet argued that his results amounted to evidence that aid was not fully fungible, since the estimated effects of different categories of aid on growth are very different, implying that aid intended for different purposes has dramatically different relationships with growth. Goodall Gondwe (Ministry of Finance, Malawi) and others voiced support for Radelet’s finding, noting that finance ministers cannot completely circumvent the allocative conditions attached to aid.

Radelet’s paper also provoked questions about methodology. T.N. Srinivasan (Yale University) wondered whether the use of cross-section regression analysis was the right approach for analyzing the effects of aid on growth. Most developing countries had done very well until the oil shock of the early 1970s. One could argue that the global environment changed around that time, and that the change has had an impact on the development process. In Srinivasan’s view, changes in donors’ objectives and recipients’ capacities and interests in receiving aid can affect the relationship between aid and growth but are not captured in cross-section regressions.

Among others who commented on methodology, Tertius Zongo (Ambassador to the United States, Burkina Faso) noted that a careful appraisal of the growth effects of aid required evaluation of the effectiveness of public spending, and Nancy Birdsall (Center for Global Development) emphasized the importance of looking at the impacts on productivity in assessing whether the growth effects were sustainable. Aart Kraay (Development Research Group, World Bank) added the perspective that cross-country regressions assessed the effects of aid, both directly and through its interaction with policies, after controlling for a host of factors, including institutions; but they did not shed light on the effects of aid on institutions and policies. The channels through which aid matters for institutions and policies are not well understood, but are an important agenda for research.

Aid and Poverty Reduction

Along with analyzing the effectiveness of aid in promoting growth, economists have sought a better understanding of the link between aid and poverty alleviation. Aid can make a direct contribution to poverty reduction. It can also have an indirect effect through its impact on investment, employment, market opportunities, and institutions. Kraay’s paper (Chapter 3) addresses the effects of aid on poverty using household survey data from a large sample of developing countries.⁶ A key result, which is corroborated by other studies, is that growth is the overwhelmingly predominant force behind poverty reduction: 97 percent of the cross-country variation in changes in poverty can be attributed to cross-country differences in growth, and virtually none of the variation to changes in relative incomes. More controversial, perhaps, is the finding that the impact on poverty of pro-poor growth strategies that explicitly seek to speedily improve the quality of life for the poor is not much different from that of other growth strategies. The latter finding provoked David Bevan (Centre for the Study of African Economies, Oxford) to point out that different countries have very different income distributions. Bevan regarded the view that almost all poverty reduction comes from economic growth—and that the only way to reduce poverty is through the growth process—as a form of extreme pessimism about the feasibility of altering a country’s income distribution. He acknowledged that income distributions may be heavily influenced by culture and institutions, but he was somewhat more sanguine about the possibility of change.

A second key result is that the contribution of aid to poverty reduction—both directly and through its effect on growth—is quantitatively small. In a version of Burnside and Dollar’s (2000) cross-country regression, Kraay finds that only 4 percent of the variance in growth is explained by aid, either directly or through its interaction with the quality of policies. By contrast, policies by themselves account for 10 percent of the variation in growth and institutions account for another 7 percent, while more than 60 percent cann...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Contents
• Foreword
• Preface
• 1 Foreign Aid and Macroeconomic Management Overview and Synopsis of Key Lessons:
• Session I Aid, Growth, and Poverty Reduction What Should Aid Be Trying to Achieve? How Good Is the Record?
• Session II Aid Absorption Recognizing and Avoiding Macroeconomic Hazards
• Session III Dutch Disease Where Do We Stand?
• Session IV Aid, Volatility, and Stabilization Policy. Does Aid Smooth Absorption or Exacerbate Shocks? Reliability and Countercyclicality
• Session V Aid, Debt, and Fiscal Policy
• Seminar Program
• List of Participants and Observers
• Footnotes