Foreign Aid: A Thicket of Causation, Correlation, and Everything in Between

My fellow blogger Shap did a wonderful job of introducing us both the other day, so I need not preface my post with any further remarks on that front beyond saying that I am very excited to be blogging here and that I will be the one posting with British spellings. Because, you know, I can.

This particular post is, admittedly, rather influenced by a particular final exam I had to take last semester for my Introduction to the Politics and Economics of Development course, in which I was asked to analyse and evaluate the effectiveness of foreign aid. The concept of foreign aid -- here defined as a bilateral or multilateral transfer of income to a country for the purpose of development -- is simple enough: country X gives low-income country Y a certain amount of money, which is supposed to manifest itself in higher economic growth per capita, lower mortality rates, or whatever development indicator of your choice. It's a straightforward enough notion, but a contentious argument is unceasingly waged over just how effective this approach is.

On one side, thinkers like Jeffrey Sachs argue that not enough aid has been given to the developing world. The opposing point of view insists that aid has not produced economic growth (William Easterly, one of aid's more prominent critics, makes this argument in his book The Elusive Quest for Growth: Economists' Adventures and Misadventures in the Tropics) or, worse, that aid inhibits development efforts, a line of argument advanced by the recently published Dead Aid: Why Aid Is Not Working and How There Is a Better Way for Africa. Meanwhile, economist Paul Collier walks a more moderate line, writing in his book The Bottom Billion: Why the Poorest Countries Are Failing and What Can Be Done About It that, although aid has been instrumental in maintaining economic growth in those poorest countries, the usefulness of aid is increasingly experiencing diminishing returns.

The root of this debate lies in the very stuff of foreign aid: numbers. If someone could just demonstrate an ironclad relationship between the amount of aid given to a country and its resultant economic performance (for the remainder of this post, let's posit that the economic indicator of importance is GDP growth), then all disagreement could be theoretically resolved. Alas, if it were only that easy. In trying to pin down a relationship like this, the litany of other factors that can affect GDP growth (political stability, geography, demography -- really, the list is practically endless!) must be, in essence, taken out of the calculation and held constant, therefore isolating the growth from external, non-aid influences. Alas, if it were only as easy as that! As many thinkers in the field of development have noted, foreign aid is itself affected by other variables. It is often observed, for instance, that the bulk of foreign aid is given to middle-income countries -- countries with something of a track record of minimally competent governance and economic growth -- rather than to the poorest countries. Therefore, any serious study of the relationship between aid and growth must take into account all of the variables that might aid as well.

Clearly, one doesn't have to be a professional econometrician to see that this seemingly simple exercise can get very tricky very quickly. From the simplest of models, the single-variable regression that seeks to answer, "When X changes, what happens to Y?" (a concept known as correlation), economists have derived ever more complicated models that can not merely control for any number of variables -- that is actually quite easy for an economist to do, provided that the economist has a working computer -- but also attempt to answer the only question that, in the end matters: when X changes, does this change in X cause a change in Y?

Given that all sides of this debate have amassed a plethora of evidence for themselves, the only conclusion that I can try to draw is that there just isn't a conclusion. And nobody should expect one either: the country, as an economic unit, is a creature so multifaceted that to grasp the whole of it with a single statistical model seems, well, foolish. At a certain point, perhaps foolish is what this entire endeavour is. After all, it isn't as if developing countries are machines that take aid and magically form it into improved economic growth; rather, the channels of aid-growth-development causation are less channels than they are an indiscernable thicket of strings with dead ends and knots in the middle. What more, the debate about the effectiveness of foreign aid overshadows the most pressing issue of all: the livelihood of billions of individuals trapped in poverty around the world.

Perhaps, then, the macro approach to aid is not a viable way of either dispensing aid or evaluating its usefulness. In its lieu, I would suggest a more micro approach that may be less ambitious in its goals but is ultimately more effective in facilitating development. Foreign aid should be specifically linked to a specific objective and then designed to achieve it. This method is less ambitious in its goals but may ultimately be more effective in facilitating growth and development: if the target of aid is more explicitly defined, then could it not be easier for donors to structure it in such a way that provides incentives for behaviour that leads to higher growth? In recognition of the fact that the many facets of development are intrinsically linked, aid can be organised into individual projects, each one oriented toward a particular goal. In this manner, perhaps positive spillover effects can be generated, thus improving lives the world over.

There is, of course, a great deal of unsubstantiated idealism left in this idea, but it cannot be any more inconclusive than the foreign aid status quo.

~ Min