Is there a micro-macro paradox in aid?

UNU-WIDER / Oct 2012

The question of whether aid is effective in promoting economic growth is a complex and controversial one. While there is a general consensus around the idea that aid can have positive effects at the micro and meso levels, recent studies, such as Rajan and Subramanian in 2008, argue that at the macro level it is difficult to identify ‘any systematic effect of aid and growth’. This at least suggests that after 25 years of work, the micro-macro paradox first identified by Mosley in 1987 may be revived. While foreign aid has many objectives, growth remains key. The difficulty in showing that aid has a positive effect in this area has led some to suggest a complete cessation of assistance to Africa.

In the WIDER Working Paper ‘Aid, Growth and Development: Have we come Full Circle’, Channing Arndt, Sam Jones, and Finn Tarp (AJT) take a close look at the evidence and argue that over the long-run a ‘consistent case for aid effectiveness emerges’. For AJT aid is not dead, but remains an ‘important tool for enhancing the development prospects of poor nations’.

The Evolution of the Aid-Growth Debate

AJT suggest that literature on the aid-growth relationship can usefully be divided into four generations. The first generation was based on a simple model which posited a stable and linear relationship between growth and investment in the physical capital required for production. Studies assumed that aid had a positive effect on growth as it would be used to plug gaps in savings or foreign exchange, and thus facilitate investment; consequently their focus was on the degree to which aid increased savings and investment in recipient countries. The general consensus amongst these studies was that aid does tend to increase investment and savings, but not by as much as the total amount given, suggesting, perhaps unsurprisingly, that some aid is consumed rather than invested.

This focus on the relationship between aid, investment, and growth was continued in the second generation of literature. While most of the studies in this generation found a positive link between aid and investment it was Paul Mosley’s 1987 description of the micro-macro paradox that garnered the most attention. His work raised doubts about the appropriateness of the underlying growth model, suggesting both that expecting all capital investment to translate into economic output, and expecting all aid to be used as investment, were overly bold assumptions. Other doubts about previous research on the macroeconomic effect of aid were also raised. In particular it was pointed out that in order to be accurate, studies needed to take into account the fact that countries might receive more aid precisely because of a poorly performing economy.

These doubts about the assumptions at the core of previous research, as well as the new availability of panel data which allowed researches to look into the impact of aid both across and within countries over time, motivated the new approach of the third generation. A paper published in 2000 by Burnside and Dollar was particularly influential; the paper suggested that aid could be effective given the right conditions, namely good fiscal, monetary and trade policy. The third generation is characterized by a cautious optimism about aid illustrated by Roodman’s suggestion that while some aid is likely to increase development and growth, it is not the decisive factor.

This optimism is not reflected in the fourth generation which is distinguished by a general view that the aggregate impact of aid on growth is non-existent. This non-existent impact is explained by political economy dynamics; aid inflows can weaken governance by encouraging corruption and rent-seeking activities. This negative outlook in terms of the macroeconomic effect of aid is not reflected in studies of its micro and meso effects. It is this problematic that leads to the suggestion that Mosley’s paradox has reemerged, AJT aim to examine this view by reassessing the results of one of the fourth generation’s leading papers.

Reassessing Rajan and Subramanian

More than any other ‘fourth generation’ research it is Rajan and Subramanian’s 2008 study of the aid-growth relationship which found that aid has no effect on growth, that has led to the revival of interest in the micro-macro paradox. While acknowledging that Rajan and Subramanian’s work provides a great contribution to the literature and advances a number of important methodologies, AJT point out concerns with both the specification of the problem and the statistical instrumentation strategy. They argue that once these concerns are addressed, a positive relationship between aid and growth can again be established.

Once AJT’s changes to the specification and instrumentation are made, and the tests are rerun they are able to draw a number of conclusions.

  • The average effect of aid on growth is positive in both the 1960-2000 and 1970-2000 periods.
  • An inflow of aid at the level of 10 per cent of GDP spurs a more than 1 percentage point increase in annual per capita growth rate on average, in the long run.
  • In the shorter term the effect of aid is difficult to discern.
  • Combining this long term macroeconomic evaluation with the already established micro and meso effects of aid provides us with a 'consistent case for aid effectiveness’.
  • There is no micro-macro paradox; aid can be shown to be effective at both levels.

Given these results AJT argue that cutting back aid would be counterproductive in terms of enhancing the development prospects of poor nations. The challenge now, AJT suggest, is to work on how aid can be made more effective, particularly in terms of spurring productivity increases, so that over the next three decades living standards in poor countries are advanced substantially.

This report by James Stewart summarizes UNU-WIDER working paper no. 2010/96, ‘Aid, Growth and Development: Have we come Full Circle’ by Channing Arndt, Sam Jones, and Finn Tarp.

Related papers: