I think most discussions of charity are way too fuzzy. Your standard fundraising proposal has several adorable pictures, a couple disturbing ones, and one or two numbers that seem to have been shouted in a moment of impassioned inspiration, with nary a source to be found. Then, though, there is the occasional economist/mathematician/cyborg who comes along and decides enough is enough – no more guessing and generalizing, we’re going to get everything down to one number come hell or high water. This annoys me even more than excessive fuzziness. At least when someone’s crying incoherently, I can tell they’re human.
There’s a common insistence on stating the cost of, say, malaria in terms of impact on GDP. You can see a great example here … with a footnote to an academic paper that I’m guessing put a sick amount of effort into coming up with these numbers. Frankly, I don’t want to think about this effort, let alone read about it, because this number is meaningless.
Allow me to explain. According to these guys, malaria costs Africa about $12 billion per year. … What does that tell you? Does that help you understand anything about the impact of malaria? Well, maybe you just need some context, so here, I’ll put it in context for you. It’s about half the cost per year we would incur by dismantling Citigroup, a little more than dismantling AIG (here’s my very questionable source for those claims), 1/1000 the annual output of the United States, 1/50 the annual output of Brazil, and half the annual cost of racism (OK, I made that last one up). The death of Alex Rodriguez would cost the world about $25 million next year, or the same as the death of 315.14 college professors. We’re totally clear now?
To whom does “$12 billion” mean more than “2 million people”?
Maybe the problem is that this number just isn’t complicated and counterintuitive enough. That must be what the World Health Organization was thinking when they put together Disability Adjusted Life-Years. Read that formula. Then tell me what “5 Disability Adjusted Life-Years” means to you.
Of course, the fact that something is complicated doesn’t mean it’s a bad idea. I’m a fan of the VORP metric in baseball, for example, because having read the description of how it’s calculated, I feel reasonably confident that if one player’s VORP is way higher than another’s, he’s probably a more valuable hitter. Putting all causes in the same terms tries to do for charity what VORP does for baseball: let us compare everything in the same terms.
The problem is, it just can’t be done. Deciding whether fixing education or malaria is more important involves a ton of philosophical judgment – enough that you can’t reasonably entrust that decision to an “expert” with a calculator. Once people start trying to make major philosophical decisions with formulas, they come out with numbers that no longer tell you what you want to know.
Our goal at GiveWell is to measure what we can measure, compare what we can compare, and acknowledge when we reach the point at which decisions become philosophical judgment calls. That way the donor doesn’t have to trust our reasoning … and that way we can make sure we’re still talking about people.