Aid effectiveness in 36 African countries

UNU-WIDER / Oct 2012

The question of whether aid is effective in promoting growth is a controversial one. Views range from those who are highly skeptical that aid has any effect on growth at all, to those who believe that aid can play a significant role in promoting economic development. In the WIDER Working Paper 'The Long-Run Impact of Foreign Aid in 36 African Countries: Insights from Multivariate Time Series Analysis', Katarina Juselius, Niels Framroze Møller, and Finn Tarp (JMT) offer an insight to this question by using a statistical model to perform an analysis of the relationship of aid and growth indicators in 36 sub-Saharan African countries. They argue that when an appropriate model is used it becomes clear that there is strong evidence that aid has a positive effect on macro variables, including growth.

Foreign aid: transmission mechanisms and variables

The main transmission mechanism through which aid is expected to have a positive effect on GDP is investment. Investment is often limited by insufficient domestic savings, low foreign currency reserves, and insufficient domestic tax revenue which constrain a country's ability to import foreign capital goods. JMT posit that if aid can fill these gaps, then it can help increase investment, which should in turn increase GDP. Even in the absence of such gaps, aid may help increase investment by, for example, improving infrastructure, and thus making private investment more profitable.

Books in a classroom, Sudan. © UN Photo/Albert Gonzalez FarranAid can also increase public and private consumption within a country. This consumption can be growth-enhancing, such as spending on health and education. In these cases it represents another mechanism through which aid can have a positive effect on GDP. However it should be noted that aid could also have the effect of increasing non-productive consumption, and as such may have a negative effect on GDP and investment.

Having identified investment and consumption as the two transition mechanisms, the authors go on to identify the key macro-variables. Clearly, as growth is the main focus of much of the literature on aid effectiveness, GDP is an important variable. The authors add to this investment, private consumption, and public consumption, all of which represent the potential transmission mechanisms identified.

The relationship between aid and macro-variables

JMT point out that there are four possible relationships between aid and the four identified macro variables (GDP, investment, private, and public consumption).

  • Case 1: Aid doesn't affect the macro-variables. The macro-variables don't affect aid levels.
  • Case 2: Aid doesn't affect the macro-variables. The macro-variables do affect aid levels.
  • Case 3: Aid does affect the macro-variables. The macro-variables don't affect aid levels.
  • Case 4: Aid does affect the macro-variables. The macro-variables do affect aid levels.

Clearly Cases 1 and 2 imply that aid does not have an effect on growth, as in both cases aid is having no effect on macro-variables. Cases 3 and 4 both may imply the opposite, that aid does have an effect on growth, as they posit aid having an effect on the macro-variables, however it is of course possible that the effect is negative.

The authors use a co-integrated VAR model to test which of these cases best describes the 36 countries they focus on. They find that in twenty-five of the countries the relationship between aid and the macro-variables is best described by Case 4, that is the relationship between aid and growth runs in both directions. This demonstrates the perilous nature of assuming that aid levels are not affected by the macro-economy of the recipient country without testing the hypothesis. The fact that Case 4 is the most common case also shows that aid does have some effect on the macro-variables. However, the size, or direction, of this effect is not yet known, if the size is statistically not significant, or the direction inverse, then these results clearly cannot be taken as evidence that aid has a positive influence on the macro-variables. This motivates the authors to undertake a more detailed analysis of the long-run effect aid has on each of the macro variables.

Results

JMT run tests that look for evidence of both the argument that aid can play a role in increasing growth and for the argument that aid has no or negative effect on growth. The key results are outlined below.

  • When looking for evidence that aid has a positive effect on growth:
  • In 27 of the countries studied aid has a positive and significant effect on either GDP, investment or both.
  • In 7 countries the effect of aid on growth in positive, but not statistically significant.
  • Aid only has a statistically significant negative effect on GDP or investment in two countries, Comoros and Ghana.
  • The hypothesis that aid increases non-productive consumption, and thus reduces GDP or investment, is found to have no support.

When looking for evidence that aid has a negative or no effect on growth argument:

  • Aid has a negative and statistically significant effect on GDP in 9 countries, and on investment in 7.

JMT point out that these results clearly show that the idea that aid is effective finds much more support in the data than the idea that it is not. The results thus represent strong evidence for the thesis that aid has in general a positive long-run effect on growth. However JMT note that aid was not shown to be effective in all countries and that further analysis on why this is the case is therefore required. They also point out that there is room for future analyses which take into account more than the four macro-variables addressed here.

This report by James Stewart summarizes UNU-WIDER working paper no. 2011/51, 'The Long-Run Impact of Foreign Aid in 36 African Countries: Insights from Multivariate Time Series Analysis', by Katarina Juselius, Niels Framroze Møller, and Finn Tarp.

Related papers: