Like every Friday, from Raj Nallari and Breda Griffith's lecture notes.
Scaling Up Aid
The principle of mutual accountability lies at the heart of the Millennium Declaration signed by 189 countries in 1999, which aims to scale up aid to developing countries and to ensure that aid is used effectively in reaching the Millennium Development Goals. On the one hand, donors—multilateral and bilateral—must increase aid flows, improve quality in delivery by taking into consideration country context and by opening markets to developing countries. At the same time, developing countries must commit to development strategies and stronger systems of governance. Among the challenges arising from a surge in aid flows identified by Heller (2005) are:
- “ensuring that larger aid flows promote growth and reduce poverty;
- increasing substantially the delivery of government services and investments in infrastructure, and managing spending decisions when a large proportion of financing (aid) is outside the government’s control and uncertain in duration;
- dealing with the possibility that higher aid flows might cause an appreciation of an aid recipient’s currency, or domestic inflation, with adverse effects on the country’s international competitiveness;
- handling the increased complexity of managing monetary, fiscal and exchange rate policies when higher aid is subject to uncertainty in terms of its magnitude, timing and likely economic effects;” (Heller, 2005).
Aid and Economic Growth
Opinion on the causality between aid and growth has received huge research attention in recent decades and has, not unexpectedly, come up with hugely mixed findings. At one extreme, some research has found that aid has no effect on growth and may actually undermine growth; by contrast, some research has suggested that aid has a positive relationship on growth on average, although not in every country and with diminishing returns. In between is the view that aid has a conditional relationship on growth, helping to accelerate growth only under certain circumstances (Radelet, Clemens and Bhavnani, 2005).
A range of empirical studies from the early 1970s to the mid-1990s found that aid had no effect on growth and could actually undermine it. Reasons put forward related to aid being misappropriated and spent unproductively—limousines or presidential palaces—encouraging corruption. Furthermore, empirical evidence suggested that aid may distort and weaken private sector incentives. Aid flows may also cause the currency to appreciate, thus weakening the profitability of the traded goods sector—an effect known as Dutch Disease—and, under another adverse channel, food aid could reduce farm prices and adversely affect farmer income if it is not managed suitably.
This literature has been criticized for assuming a linear relationship between aid and growth, i.e. assuming that each extra dollar of aid has the same effect on growth as the previous dollar and thus ignoring the possibility of diminishing returns and the endogenity of growth whereby faster growth might cause extra aid. Recent work by Rajan and Subramanian (2005) suggests that a “Dutch disease” effect—aid flows giving rise to a real appreciation of the aid recipient’s currency and undercutting its competitiveness and weakening growth—lies behind their findings of no robust evidence, either positive or negative, of aid on economic growth. The finding holds for time periods, types of aid, types of donors and characteristics of recipient countries.
At the same time, several studies point to a positive relationship between aid and growth. The studies do not conclude that more aid leads to growth but suggest instead that “higher aid flows have been associated with higher growth” (Radelet, Clemens and Bhavnani (2005)). Investment in physical or human capital financed by aid is one channel through which aid could support growth. During the 1990s, this literature began to look at whether diminishing returns to aid flows—i.e., that the impact of additional aid would decline as aid amounts grew—could set in at a certain level of aid relative to GDP. Working along these lines, Clemens (2004) finds that diminishing returns to aid sets in after total aid reaches about 17 percent of GDP (or the subset of aid aimed directly at growth reaches 8 percent of GDP). Heller (2005) notes that aid to a number of African countries is already above 10 percent of GDP thus highlighting the importance of recipients managing aid effectively.
The conditional view of aid suggests that aid has a positive impact on economic growth conditional on certain variables, for example the effectiveness of donor practices and recipient characteristics. Based on research at the World Bank in the mid-1990s, several researchers noted that WB projects had higher returns in countries with stronger civil liberties (Isham, Kaufmann and Pritchett (1995)) and that aid stimulated growth in countries with good policies but not otherwise (Burnside, Craig and Dollar (2000)). Further recipient country characteristics such as vulnerability to trade shocks, climate, institutional quality, political conflict and geography have all been studied as conditional variables affecting the aid-growth relationship. Whilst Radelet, Clemens and Bhavnani (2005) raise doubts as to the statistical robustness of such studies, they note that the view of aid as working best in a country context of good policies and institutions has become the prevailing wisdom for donors, multilateral development banks and the foundation of the U.S. Millennium Challenge Account. Reflecting the importance of host country conditions, in recent years recipient countries have had more of a say in decisions about setting priorities and designing programs working closely with donors and civil society. Greater country ownership and broader participation are considered positive developments for the efficacy of aid. The effectiveness of donors has received very little systematic research although many opinions prevail here also. Multilateral aid is considered to be more effective than bilateral aid and untied aid to generate higher returns than tied aid.
Recent work on the growth and aid debate also suggests that the type of aid, in terms of its substance and timing, matters. Pioneer work by Clemens, Radelet and Bhavnani (2004) focuses on the type of aid that is aimed primarily at economic growth. Their review of the literature on economic growth and aid showed that almost all studies of growth and aid examine the relationship of total aid to growth. Yet they note that not all aid is aimed at growth, for example, humanitarian and food aid are aimed primarily at consumption not growth per se. The same argument also holds for aid allocated for purposes of education materials, democracy, and judicial reform. Furthermore, the authors note that past studies had made use of cross section data at four-yearly intervals—too short to capture the impact of aid on growth. (On the other hand the authors note that problems of causality arise if the time period is too long.)
Based on their hypothesis that “not all aid is alike”, Clemens et al. (2004) examines the effect of (i) humanitarian aid; (ii) early impact aid, and (iii) late impact aid on economic growth. Their results show that (i) and (iii) have very little or zero effects on growth, exhibit below. On the other hand, early impact aid – aid to build infrastructure, aid to support productive sectors, aid for budget support – that accounts for about half of total aid has a positive relationship with growth over a four-year period.
Among the findings on the effect of early impact aid on economic growth quoted in Radelet, Clemens and Bhavnani (2005) are:
- the positive effect of early impact aid on economic growth is not reversed over time;
- in general, growth-oriented aid aid has a positive effect on economic growth;
- each US$1 in early impact aid yields US$1.64 in increased income in the recipient country in net present value terms or a project level return of 13 percent per country;
- for Sub-Saharan Africa, higher than average early impact aid raised per capita growth rates 1 percentage point above what would have been achieved by average aid flows;
- the relationship between early impact aid and economic growth is stronger for countries with good policies and institutions;
- the aid-growth relationship is stronger in countries with higher life expectancy;
- typically the maximum growth rate a country can achieve occurs when early impact aid represents 8-9 percent of GDP;
- given that early impact aid is roughly half that of total aid, the maximum growth rate occurs when total aid reaches around 16-18 percent of GDP in the typical country;
- aid is partially fungible – aid flows intended for different purposes have significantly different relationships with economic growth” .
In sum, Radelet, Clemens and Bhavnani (2005) are optimistic about the effect of growth-oriented aid on economic growth for all the reasons summarized above.
Aid – Substance and Timing – and Economic Growth
Source: Finance and Development, September 2005
- Fridays Academy