Using Data, Networks and Complexity to Study Trade, Aid, Economics

A recent blog post of mine got all lathered up about the work of Toby Ord‘s charity-bothering organisation Giving What We Can. (Note: from the pictures on their website it seems that there are other people involved, but calling it “Toby Ord’s” is lazy journalese for the less wieldy “charity founded by Oxford researcher Toby Ord etc. etc.” Who wants accuracy when you can have brevity, eh?)

The organisation says it “provides information about the cost-effectiveness of different charities”, something which the people at the International Aid Transparency Initiative have been talking about for a few years now. (Note that if I’m quoting, it’ll be from givingwhatwecan.org unless I mention otherwise.)

Given the famous paucity of data available from NGOs about what they spend their money on, this is a pretty exciting claim. So how does it work?

Well, the first thing to say is that the only charities that are rated are those that “have a relatively narrow focus”. This makes perfect sense of course, since rating the biggies like Oxfam and MSF would involve rating hugely varied and various projects. That said, it serves to slightly water down Tim Wigmore‘s assertion in this week’s New Statesman that Ord’s research “shows there is a clear winner: charities that focus on single issues.” (Sorry Tim, I’m not intending to turn this post into some kind of hatchet job. It’s thanks to your article that I’ve heard of Toby Ord at all! But the point is probably nevertheless worth making.)

The next thing to clarify is that Giving What We Can (GWWC) are so far focused on health-related interventions, using cost-effectiveness data from the World Heath Organisation (WHO) and figures on the global burden of disease from the Disease Control Priorities Project (DCP2). This is also an eminently sensible first step since GWWC base their charity rankings on something called disability-adjusted life years (DALY), a WHO metric which accounts for both length of life and, crucially, quality of life. (By the way, the weights given to each disability in terms of its negative impact on quality of life make fascinating reading in themselves.)

This helps to explain why GWWC’s top-rated charities are all health related, namely: the Against Malaria Foundation, the Schistosomiasis Control Initiative and Deworm the World. These JPAL-style health interventions have been shown to be hugely effective in boosting test scores and school attendance as well as in having the obvious impact on quality of life which not having a disease affords you. But it’s worth making the point that these charities are the most effective of the single-issue health intervention charities, not the most effective of all charities overall. Given the huge importance of this knowledge in itself, I don’t see that the message would be robbed of relevance by making this explicitly clear.

I think projects like this are what make the open data movement in aid so important and so exciting. The idea behind the movement is that simply by making information available, passionate and talented researchers such as Ord and his team will turn the data into knowledge which can be used by the public at large (or even policy makers) to inform their donation decisions. I think that what GWWC have done with the statistics is remarkable.

But it’s also important that the limitations of What We Can Know are made clear when we decide What We Can Give.

NO COMMENTS
Post a comment