Bangladeshi workers arrange boxes of relief supplies: new study confirms that IMF-backed programs support full use of aid (photo: Farjana Khan Godhuly/AFP)
Low-Income Countries
The advice that the IMF gives to member countries about when and how to spend aid money has received much attention in recent years. Many people have asked about the role that aid volatility, capacity constraints, and concerns about macroeconomic stability plays in shaping the IMF's advice to countries.
Last year, the IMF clarified the principles for its advice on the use of aid: the IMF supports the full use of aid over time, taking into account the need to safeguard macroeconomic stability and limits on productive spending. But what has been the Fund's actual advice to countries receiving aid?
Aid spending versus aid absorption
Our paper, The Spending and Absorption of Aid in PRGF-Supported Programs, answers this question by analyzing the extent to and pace at which countries that have a program under the IMF's Poverty Reduction and Growth Facility (PRGF) are expected to use the aid they receive. Specifically, the study looks at how much of the anticipated aid inflow the government is expected to spend, and how much of the corresponding foreign exchange inflow the central bank is expected to sell.
To clarify the distinction, imagine that aid dollars are given to the government. First, the government sells these dollars to the central bank to obtain local currency. Second, the government decides how much of the local currency to spend on new projects or programs. Separately, the central bank decides how much of those aid-related dollars to sell to importers, thereby allowing for higher net imports.
We measure the government's decision by looking at the increase in the fiscal deficit (net of aid) that is explained by the increase in aid (so called "aid spending"). We also measure the central bank's choice by looking at the increase in the current account deficit (net of aid) that is explained by the increase in aid (so called "aid absorption").
New data examine projected use of aid over time
In 2007, the IMF's Independent Evaluation Office (IEO) found that the programmed use of aid in sub-Saharan African countries with a PRGF-supported program was sometimes limited. However, the IEO study focused on the amount of aid spent during the same year when aid is received, which provides a rather limited picture on the use of aid by countries. Indeed, many countries choose to spread out the use of aid over several years. For example, if aid is volatile, initiating long-term projects that require steady financing could generate financial difficulties in the future if all aid were fully spent in just one year.
Our research goes beyond this and other studies in several important respects:
• It estimates the programmed use of aid over time rather than just in the year when aid is received.
• It analyzes how macroeconomic stability influences the programmed spending and absorption of aid—that is, the plans for spending and absorbing aid yet to be received.
• It uses comprehensive data from IMF program documents since the inception of the PRGF in 1999 until end-2007—in all, for 378 episodes of program requests or reviews.
IMF-backed programs help smooth aid
Our paper finds that PRGF-supported programs are designed so that countries can fully use increases in aid over time. Specifically, programs are designed so that countries can spend 58 percent of expected increases in aid immediately, when the aid is received, and an additional 26 percent in the following year. Thus, in total, 84 percent of a projected increase in aid is programmed to be spent within 2 years. Similarly, countries are programmed to absorb 56 percent of increases in aid immediately, and a further 16 percent in the following year. The study also presents several alternative ways of measuring the programmed use of aid, which broadly confirm these findings.
Another important finding of our study is that the programmed use of aid does not depend on simple thresholds related to macroeconomic conditions. For example, the IEO found that only countries with low inflation and adequate reserves were programmed to spend and absorb all or almost all aid.
By contrast, based on a more complete dataset, our study indicates that while high inflation and low reserves can indeed limit the degree to which increases in aid are programmed to be used (rather than saved), program design does not follow simple rules. Rather, the programmed use of aid reflects a country-specific analysis of its opportunities and vulnerabilities.
Spent but not absorbed
How does actual spending and absorption compare with programmed spending and absorption in PRGF-supported programs? First, our study suggests that actual spending of aid is full, but there is no smoothing of aid-based spending. By contrast, actual absorption of the aid received is lower than what was planned, at only 32 percent. These results are consistent with earlier case studies which suggest that, in many cases, aid is spent but not absorbed.
Second, our paper also finds that the presence of a program does not impair the actual spending and absorption of aid: the actual use of aid is broadly the same in years with and without a program. For the government, this means that the program conditionality does not impose a binding constraint on the path of aid spending. The lower-than-programmed absorption means that the central bank often keeps more aid-based-foreign exchange in reserves than the minimum required under the program, thereby limiting the scope for higher imports.
Of course, that's not to say that programs do not still guide the way the program is financed, for example, by helping to ensure debt sustainability and avoid risks of crowding out private investment or inflation.
Comments on this article should be sent to imfsurvey@imf.org
