Waylaid Dialectic

August 3, 2010

Counting Your Chickens Once They’ve Grown Up

Filed under: Aid — terence @ 9:13 am
Tags: , , ,

Like I said, caveat lector on all aid growth regressions, but this one’s kind of interesting:

Development aid and economic growth: A positive long-run relation

Camelia Minoiua and Sanjay G. Reddy

The Quarterly Review of Economics and Finance
Volume 50, Issue 1, February 2010, Pages 27-39
(Gated here; un-gated working paper here)

We analyze the growth impact of official development assistance to developing countries. Our approach is different from that of previous studies in two major ways. First, we disentangle the effects of two kinds of aid: developmental and non-developmental. Second, our specifications allow for the effect of aid on economic growth to occur over long periods. Our results indicate that developmental aid promotes long-run growth. The effect is significant, large and robust to different specifications and estimation techniques.

Interesting simply for the two ideas that it bases its analysis on:

  1. That the impact of aid on growth may be most clearly realised over long time-frames (once those healthy better educated kids have entered the workforce, for example).
  2. That different donors have different effects. In particular, those that give aid for developmental as opposed to geopolitical reasons are much more likely to give aid which works.

Point one is just an interesting point. Point two, I think, is essential to regonise explicitly whenever we talk about whether aid works or how we should give it better. Even if you’re sceptical about the regression evidence, there is so much evidence from the cold war of rent-a-dictator type aid going badly that the point is beyond debate.

And it’s worth acknowledging – before you start talking about Searchers or Planners, or the need to tailor aid programmes carefully to context (one of my main beliefs) – that aid needs to be given with good intentions. If it’s not (and too much of it hasn’t in the past) it’s highly unlikely to work.

The old saying is that ‘the road to hell is paved with good intentions’. Maybe, but almost always it was designed, engineered and hewn long before with abundant ill intent. Remember that – if you want aid to work, at the very least you’ve got to give it with the intent of helping people.


  1. I very much agree with your conclusion, however I am wondering to what extent you can distinguish donors giving to really help and donors giving for geopolitical reasons (I did not go into your working paper to be honest). I would imagine most donors, like the German government, have a perfect blend of geopolitical, security, economic and moral motives in the way they give aid.

    Comment by Claudia Schwegmann — August 4, 2010 @ 4:28 am

  2. Thanks Claudia – I agree. Even the best donors will give some aid for less than perfect motives. The difference is one of degree. The worst agencies will give almost all of their aid for these reasons. In their paper Reddy and Miniou categorise as follows (from p30 of the published paper):

    In contrast, aid from Nordic nations “seems remarkably free from self-interest and, indeed more oriented towards their stated objective of poverty alleviation, the promotion of democracy, and human rights” (Gates & Hoeffler, 2004, p. 16). Alesina and Dollar (2000) report that small donor aid has the highest elasticity to recipient income, while Gates and Hoeffler (2004) show that Nordic countries (Denmark, Finland, Norway, and Sweden) give grants and concessional loans to poorer countries, many of which are in sub- Saharan Africa (SSA). “Norway and Denmark are lauded for their singular focus on development.” (Brainard, 2006, p. eight) Some donors send little money to former colonies (the Netherlands, 17 percent) or have little scope for fostering global strategic interests due to a lack of colonial past. Alesina and Dollar (2000) conclude that “Certain donors (notably, the Nordic countries) respond more to the correct incentives, namely income levels, good institutions of the receiving countries, and openness” (p. 33; italics in original text). Similarly, Gates and Hoeffler (2004) argue that Nordic donors as an aggregate differ markedly from other donors in their allocation of aid: their recipients are more likely to be democracies and have a better human rights record. At odds with the findings of Alesina and Dollar (2000) that countries open to international trade are favored by Nordic donors, Gates and Hoeffler (2004) report that the same countries still direct significant amounts of aid to recipients with “poor” trade policies.12 Based on this evidence, we assume that Scandinavian donors and selected additional donors have aid programs that are more likely to target developmental aims (especially economic infrastructure, poverty alleviation, and social services), and that their aid is more likely to be spent in a growth-enchancing manner. We consider the following two distinct proxies for DA: total bilateral aid from Denmark, Finland, the Netherlands, Norway, and Sweden (‘G1’), and total bilateral aid from a larger group of countries comprised of: G1 plus Austria, Canada, Ireland, Luxembourg, and Switzerland (‘G2’). There is some contention in the aid allocation literature that the Netherlands and Canada are similar to Nordic countries, although there is no definite evidence on the matter (Gates & Hoeffler, 2004).

    Scandinavian donors in G1 fared well according to the 2007 ranking produced from the Commitment to Development Index (CDI). The CDI assesses the performance of rich nations along various dimensions of policy, including aid, trade, investment, migration, security, environment, and technology. One of its components, the Aid CDI ranks donor nations after adjusting their aid figures for the type of aid extended to recipient countries (Roodman, 2005, 2006, 2007). In particular, the index penalizes donor countries which offer tied rather than untied aid, loans rather than grants, and too many small aid projects which are likely to burden the recipient government with administrative responsibilities.13 Four of our G1 donor countries (Denmark, Norway, Sweden, and the Netherlands) were the highest-ranking according to the 2005 Aid CDI. This is not surprising since a small share of Nordic aid is tied (with the exception of Denmark) and it is concentrated on social infrastructure, especially in the health sector (Gates & Hoeffler, 2004). Assuming that the highest-ranking nations on the quality adjusted aid ladder are more likely to provide DA, we choose one more group of development-friendly countries which rank in the top 10 according to the 2005 Aid CDI. The third proxy for DA is pooled bilateral aid from the following ten donors: G1 plus Belgium, France, Ireland, Switzerland, and the United Kingdom (‘G3’). Notably, this group includes donor countries that have been shown to allocate aid in a geostrategic manner, so we expect a lower effectiveness of its aid. Once DA and MA are extracted from TA, the remainder is viewed as NDA (NDA= TA−MA−DA) and is also allowed to have a distinct impact on growth.

    They find aid from groups G1, G2, G3 as having statistically significant positive relationships to economic development. Aid not from these groups doesn’t appear to have a relationship of statistical significance.

    Just to reiterate – such cross country regressions, while providing an appearance of certainty are very fraught. So this result, I think, should be treated as interesting but not definitive.

    Comment by terence — August 4, 2010 @ 9:53 am

  3. […] Except that there is evidence that aid leads to higher rates of economic growth. For example, here and here). It is true that the methodological issues associated with aid growth regressions mean […]

    Pingback by All the Views Fit to Print « Waylaid Dialectic — September 24, 2010 @ 8:29 pm

RSS feed for comments on this post.

Blog at WordPress.com.

%d bloggers like this: