The Effectiveness of Aid to Development. Focus on the Aid-Growth literature.
par François Defourny
Facultés N-D de la Paix de Namur - Université Catholique de Louvain - Master in International and Development Economics 2005
In order to make cross-country studies feasible, it was necessary to choose some kind of «common denominator» able to reflect the general efficiency of aid despite de diversity of its objectives. The aid-growth relationship is incontestably the most common relationship used in the literature to measure the macroeconomic effectiveness of foreign aid. Concretely, in the regressions, positive and significant coefficients on the aid variable is interpreted as evidence that aid was effective in enhancing growth. As Quibria (2004, p17-18) expresses: «Although poverty reduction has been accepted as the overriding objective of international development assistance, this concern has not found adequate reflection in the current research: paradoxically, much of the current research effort has been couched in terms of economic growth rather than poverty reduction.» In fact, this «second best» indicator of efficiency is essentially motivated by the poor availability of data in most developing countries4. It can also be explained by the unfortunate tendency among economist to consider growth and poverty reduction to be synonymous. The following statements of Easterly (2003, p34) reflect this misunderstanding: «the aid bureaucracies [these days] define their final objective as «poverty reduction, [which is] today's more politically correct name for `growth'.»
Nevertheless, this decision to consider the efficiency of aid only in terms of its ability to enhance national growth entails important limitations. Restricting the evaluation of aid effectiveness to its impact on national growth means that we make somehow the underlying assumption that economic growth benefits to the targeted people, namely the inferior social classes. In other words, we make the well known hypothesis that growth «trickles down» to the poor. In the late nineties, different studies found a positive relationship between growth and poverty reduction5. Some kind of consensus emerged then about the positive contribution of growth to poverty decline. But, this hypothesis is not equally valid everywhere6. As Rodrik
4 The reliability of such a statistic may be deeply questioned, as we know the importance of informal activities in developing economies.
5 We refer to Dollar and Kraay (2000) and Gallup, Radelet and Warner (1999) for the concept of relative poverty, whereas Ravallion (2000) and Collier and Dollar (2001, 2002) analyze the impact of growth on absolute poverty. They all come to the conclusion that growth is «good for the poor».
6 This is especially the case in some countries where economic growth is mainly fed by the exploitation of natural resources. (e.g. the well known «oil curse»)
(2000, p1) writes: «the poverty-reduction payoff from growth depends in part on the specific circumstances and policies in each country.» Furthermore, if the majority of authors agree on the idea that growth increases small revenues, the extent of this gain is still subject to fierce debate. In many cases inequalities appear to widen with economic growth7. So, if growth has been accompanied by an improvement in Gini coefficients in countries like Taiwan, Bangladesh and Egypt, it may also worsen social inequalities in others such as Chile, China, and Poland. This argument leads us to the discussion about the different concepts of absolute, relative and even subjective poverty. It cannot unconditionally be said that one kind of poverty should be targeted rather than the others in order to improve people's welfare. For statistical ease, «absolute poverty» is usually preferred by economists. It does not mean that it is the best indicator of (dis)satisfaction. On the other hand, is increasing inequality necessarily detrimental if it at the same time everyone has a higher income? At least, this question requires careful thought.
In addition, looking at this aid-growth relation to estimate its ability to reach its objectives may lead to other limitations. As we will see later there exist many kinds of assistance and some categories of aid may not target economic growth at all. We find it particularly frustrating to limit the assessment of support to fields like human rights, cultural emancipation, gender equity or nature protection to there impact on growth. In comparison, it is a bit like restricting the benefits of scientific research to its marketable value.
Finally, even if the analysis is limited to economic variables, one can wonder if it is appropriate to limit the concept of development to economic growth. It is a frequent mistake in the economic literature to confuse economic growth, development and welfare8. What is the contribution of economic growth to development when human indicators are pointing in the opposite direction? This important issue wanders from the purpose of our work. It is nevertheless important to remain critical when we narrow the scope of an analysis for technical reasons.
Despite all these arguments that reduce seriously the explanatory power of the aid-growth relation, the overwhelming majority of attempts to measure global aid effectiveness have been couched in terms of impact on growth. As Amprou and Chauvet (2004, p45) say, the real
7 See for instance De Janvry and Sadoulet (2000) for the case of Latin America
8 We refer to Cassiers and Delain (2006)
issue of foreign aid is poverty reduction. But very few econometric studies departed from the aid-growth relationship to analyse the real effectiveness of aid. This has given birth to a fierce and apparently inconclusive debate that deserves some critical analysis. For this reason, we will stick to this aid-growth relationship as well. In the following, we should keep in mind the previous limitations to interpret cautiously what we mean by «aid effectiveness».
9Impact, le film from Onalukusu Luambo on Vimeo.