The Effectiveness of Aid to Development. Focus on the Aid-Growth literature.

7. The second part of the literature: reactions to Burnside and Dollar (1997)

Because of its important political implications Burnside and Dollar (1997) have provoked a huge amount of comments and criticisms. Because of its emblematic role, it has attracted much attention²⁶. But most of the remarks that follow do not exclusively concern this initial article. Therefore they can be extended to several other articles of the same nature²⁷. We think important to present that critical literature since these reactions constitute the core of the ongoing debate. Dalgaard et al. (2004, p197) write about this: «It should be clear by now that much of the current discussion centres on the question if bad policies - in addition to being detrimental to growth - imply that aid is wasted.»

7.1. The indicator for «good» economic policy

Burnside and Dollar (1997) estimate the quality of economic policies by compiling a simple economic policy index²⁸ out of three measures of inflation, trade openness and budget deficit. This method presents several shortcomings revealed in the following studies. Berthélémy and Varoudakis (1996) demonstrate the limited impact of trade openness for economies with underdeveloped financial system. Lensink and White (2000) suggest that the relation between

26 It is important to notice that the following reactions to Burnside and Dollar (1997) may be either more pessimistic or more optimistic regarding the effectiveness of foreign aid.

27 For instance, «Assessing Aid: What Works, What Doesn't and Why» of the World Bank (1998) was completely in line with Burnside and Dollar's conclusions.

28 This economic policy index was later referred as the Burnside and Dollar policy index.

inflation and growth is probably non-linear and so could not be catch by a fixed index. Amprou and Chauvet (2004), emphasize that budgetary surplus is not automatically favourable for growth. Easterly (2003) criticizes the subjective and opaque aspect of the trade openness variable²⁹ used by Burnside and Dollar (1997). When he replaces the latter by more sophisticated alternatives, despite a significant correlation between the new policy variable and growth, the «aid×policy» interaction term looses its significance.

Furthermore, restricting economic policy of a country to these three variables may be a little too synthetic. Indeed, Dalgaard and Hansen (2001), Hansen and Tarp (2000, 2001), Hudson and Mosley (2001) and Lensink and White (2001) all test, in different ways, the relationship between aid and the «Burnside and Dollar policy index». They all found insignificant results. This means that the identification of which policies really matter for growth remains problematic. In the following literature, authors usually prefer the concept of «economic and institutional performance» illustrated by the much more evolved CPIA index³⁰, measured by twenty institution and policy components³¹.

More fundamentally, as Quibria (2004) expresses, what donors usually require as «good» polices may sometimes be historical, contextual and path-dependent and so not widely accepted. Their selectivity process is culturally biased. Hence, the basis for assessing one country's performances should be the concrete and measurable results of foreign aid and not «subjective assessments of policies and institutions against the benchmark of the imaginary ideal setup» (Quibria, 2004, p.34).