Giving locally vs. globally: Where can we do the most good?

“Charity begins at home” is a popular idea. At REG we’re not interested in whether ideas are popular, though, but in whether they’re based on good evidence. Where do the arguments lead in the “local vs. global” debate?

One might intuitively think there are special obligations towards “local” people. But what justifies this intuition? (It’s likely we should have it even in case it’s unjustified, given that evolutionary pressures favoured tribal thinking.) The ethical principle that all lives count equally contradicts it – and seems to be solidly grounded in the fact that equal suffering is/feels equally bad, no matter where it happens. Also, how would we ourselves like to be treated? We wouldn’t want our suffering to be discounted just because it doesn’t happen “locally”. Last but not least, if “distance discounting” is justified, how exactly does it work? What distance leads to what discount? The answer to this question is bound to be arbitrary, which further supports the non-arbitrary focus on suffering, no matter where it happens.

If we accept these arguments, we could still try and justify a “local” focus by arguing on purely practical grounds: We have a better understanding of how to improve the situation locally, distance increases uncertainty, and so the expected returns are higher if we donate towards local causes.

However, it’s not true that our understanding of more distant problems is generally worse and their solution more uncertain. Scientific charity evaluators such as GiveWell have gathered a lot of evidence about how best to reduce the suffering of people in developing countries. This evidence often comes from randomized controlled trials and is thus on a par with (or better than) the evidence about how to fix the problems of people in rich countries.

Also, money has diminishing marginal utility. Globally, about one billion people still live on less than $1.25 per day (purchasing power parity), and about 20,000 children die from the consequences every day. It is therefore very probable that we can help people to a much greater extent if we invest our charitable resources in developing countries. And this hypothesis is indeed borne out by the more specific cost-effectiveness analyses: In developing countries, it is possible to save a life for about $3,400 – while it’s simply impossible to save a life for that little money in rich countries. Health care economics shows that saving lives is 100 times more expensive locally.

So: If we decide to donate, say, an overall amount of $340,000 to charity, our choice is between saving one life or 100 lives. Imagine a firefighter standing in front of a burning house and facing the choice between saving a room containing 100 people or a room containing one person. What should he do? Sit down and do nothing; run to save the 100; run to save the one?

Whenever we spend money, we inevitably are this firefighter.