I’ve just started More Than Good Intentions, the new book on impact assessment in international development by Dean Karlan* & Jacob Appel, and was struck by a figure given in their introduction: US$2.3 trillion has been spent on development aid over the past 50 years. (They don’t specify how this figure was constructed, or whether it’s in nominal or constant dollars. However, Easterly cites the same figure elsewhere, so I’m going to run with it for the moment.) K&A mention this in the context of arguments about development effectiveness, with the usual gloss – the question of how that much money could have failed to spark development. Reading this now, however, I’d sooner ask the opposite question of why anyone might assume that such a trivial sum could suffice to materially transform large swathes of the world.
Think about it: US GDP in 2009 was $14 trillion in nominal terms. In a single year, the wealthiest country in the world produces up to 6 times the value of all the money spent on aid over the last 50 years. To put it in per capita terms, US per capita GDP in the same year was $45,989 in nominal dollars. Let’s assume that the $2.3 trillion in aid was spent equally over those 50 years, for approximately $46 billion in aid per year. Let’s assume as well that this aid went exclusively to the bottom billion during each of those years. That leaves us with about $46 per person per year over 50 years – about a month and a half of subsistence at an average of $1 a day.**
This is a highly stylized and inevitably inaccurate description of how aid funding is spent, but I found it useful to put these numbers into context. Why should we expect that a sum like $46 per person per year, no matter how effectively spent, might successfully pull nations out of poverty? Why should we expect this paltry ammunition to succeed against the array of historically and politically contingent reasons why countries find themselves unable to grow or to equitably distribute the benefits of growth?
Of course, this isn’t actually an argument against aid, or improving aid effectiveness. The fact that it isn’t sufficient to raise all impoverished people out of poverty doesn’t mean that the limited but real benefits that it can provide – like improving access to healthcare or education – are suddenly worthless. And certainly almost everyone in the development community sees aid as necessary yet insufficient for development. However, to echo Fukuyama’s critique of the lack of historical context in recent political science work (which you can read in my notes [PDF] from his recent talk at SAIS), I find it a bit troubling that we as development practitioners are still quite so focused on the causal link between aid and development, sometimes at the expense of broader thought about how countries develop and why. As a commentator at the Fukuyama talk said, much development work feels like it’s “trying to do history in a hurry” – and with insufficient tools at that. Alongside rigorous evaluation of the type K&A advocate, I’d love to see a stronger understanding of historical contingency & context in discourse on development.
* I worked on one of Dean’s projects with IPA, and he generously sent me a galley of the book for free.
** For the sticklers on research methods, yes, I know that nominal & real dollar amounts aren’t directly comparable; that aid hasn’t gone to a tidy billion people per year for exactly 50 years; that $1/day is actually a complex estimate of poverty [PDF]; and that living on $1/day doesn’t actually mean you get a dollar per day. It’s a thought experiment, yo.