The $138.5 billion question: When does foreign aid work (and when doesn’t it)?

Jonathan Glennie, Andy Sumner


When does foreign aid work and when doesn’t it? This paper considers peer-reviewed, cross-country, econometric studies, published over the last decade in order to propose areas with policy implications related to the conditions under which aid is more likely to be effective. The paper discusses the nature of evidence on aid and why assessing its impact is so difficult. It looks at when aid is most likely to work, as opposed to just whether aid works or not. The paper reviews aid’s impacts on economic growth and social development in general before focusing on conditions identified in the aid and growth literature under which aid is more likely to be effective.

The paper suggests that there are four broad areas where the evidence reviewed shows signs of convergence that have direct relevance for policy decisions on aid effectiveness. These areas are: (i) aid levels; (ii) domestic political institutions; (iii) aid composition, and (iv) aid volatility and fragmentation.

Key findings:

  • It is logical that aid is likely to have diminishing returns as it grows relative to the size of the economy or government expenditure, and even turning negative. The last decade of evidence backs this up. Evidence also suggests that aid at low levels may have little impact on growth. This is an important finding not because it is surprising, but for the neglect that there has been in policymaking circles of this critical element of aid effectiveness. In the most important aid effectiveness process, the Paris agenda and its successor meetings, the issue has barely merited a mention, and there appears to be no mechanism whereby donors and recipients can analyse appropriate aid levels and moderate them up or down according to effectiveness criteria. Instead, there appears to be a generalised push for more aid for the poorest and less aid for countries reaching middle-income status, a policy seemingly directed more by political concerns than by the aid effectiveness evidence. Unlike other contextual issues, an aid level is an area entirely susceptible to concerted action by donors and recipients.
  • The effectiveness of aid depends on what the aid is intended for. For example, aid effectiveness for growth is improved if aid focuses on ‘developmental aid’ rather than ‘non-developmental aid’, or the composition of aid is directly aimed at affecting growth (building roads, ports and electricity generators, or supporting agriculture) or focused on agricultural aid. Further, budget support/ ‘programme’ aid and project aid given for real sector investments is likely to be more effective for growth than other types of aid, but caution is required as aid in other sectors such as health and education may only affect growth after a long period of time and thus may be difficult to detect rather than be non-existent.
  • There are two areas where there is little convergence in the evidence of aid effectiveness, despite oft-cited claims to the contrary. First, on macroeconomic policies, papers published since 2004, starting with Easterly’s important rebuttal, overturns a previously core belief in official development circles i.e. that aid supports growth when the recipient country is implementing certain macroeconomic policies generally described as ‘good’, meaning orthodox policies. Second, the debate between proponents of grants and loans also has a long history in the literature. The findings of this latest generation of evidence simply confirm that this remains a disputed area in the academic literature. This is not to say that in different contexts grants may be more appropriate than loans, or vice versa, simply that there are no generalisations that can currently be made on the subject.
  • Source

    Glennie, J. & Sumner, A. (2014). The $138.5 billion question: When does foreign aid work (and when doesn’t it)? Washington DC: Center for Global Development.