[Part of an occasional series in which I explain the evidence for various EA top charities. You can get all of this information on GiveWell’s website, but I’m more fun and swear more. I want this post to be accurate and will correct it for factual mistakes. If you are reading it more than about a year in the future, it is probably not currently accurate.]

There are lots of good things about Give Directly, and because I’m a nice person I’m going to start with the good stuff.

Give Directly is organizationally competent. Cash transfers make up 83% of Give Directly’s overall expenses and 85-91% of expenses paid for by donations from the public; while overhead is a bad way to pick charities in general, it’s not a bad way of assessing Give Directly, which is supposed to, well, give people money directly. They have responded well to cases of staff fraud and carefully monitor to make sure that the money gets to where it’s supposed to be.

Give Directly is one of the effective altruist organizations with the most room for more funding. Prima facie, this seems fairly obvious (there are more of the global poor than there are of the global poor who also are at risk of contracting malaria). They have a track record of scaling up their operations quickly when they get more money, including growing their cash transfers by a factor of two in a single year. While it is possible that they will not be able to grow this quickly in the future, even so, they are likely to grow fast with more money.

Give Directly works with a high level of certainty. Poor people have less money than rich people; therefore, if you give people more money, they will be less poor. The worst-case scenario is that they spend the money on eating better, experiencing more entertainment, and getting more health care for a year or two (as opposed to the worst-case scenario for most interventions, which is that it doesn’t have any benefit or even harms people). Give Directly’s research suggests that recipients of cash transfers do increase their assets through purchasing livestock, durable goods, and iron roofs, and through increasing their savings, so Give Directly does outperform the worst-case scenario, but it’s comforting to know that even if the research is wrong the intervention really definitely does improve people’s lives in the short run.

Give Directly is amazingly empirical. Their goal is to enroll literally every participant in Give Directly in a study or campaign variation; they deliberately choose experiments that not only improve Give Directly as an organization but also can influence policymakers (for instance, through studying the effects of cash transfers on inflation and job creation). Give Directly’s previous randomized controlled trial was preregistered, showing a commitment to good social science that a lot of social scientists don’t have. Therefore, giving to Give Directly is making a contribution to human knowledge about cash transfers and development as a whole.

One particular study Give Directly is running might be interesting to my readers: Give Directly is testing a basic income guarantee. Give Directly’s basic income guarantee will be the first basic income guarantee that’s long-term (more than ten years), universal, and enough to live on. It is significantly cheaper to give people in East Africa enough money to live on than it is to give people in America enough money to live on. And while some aspects won’t generalize, human psychology is similar across countries: this study could be a tremendously useful source of information on whether basic incomes stop people from working, allow people to take more risks, and whether it improves psychological well-being.

Give Directly’s premise is something that goes along with a lot of people’s ethical systems. A lot of people tend to be opposed to telling other people what to do; they know their lives better than you do, and you shouldn’t try to make decisions for them. They don’t like paternalism. If you’re sympathetic with that– and I am– Give Directly is a really big source of warm and fuzzy feelings. You’re just giving people money, and then they can decide what matters most to them, whether that means expanding a business, getting medicine for a sick child, or buying musical instruments.

So here’s the bad: cash transfers aren’t that good.

I mean, they’re good! Cash transfers definitely improve the lives of the global poor, like I said above. Cash transfers are probably one of the best charities you can donate to. But they’re not the best of the best charities. GiveWell’s latest cost-effectiveness analyses suggest that other GiveWell interventions are somewhere between twice and twenty-three times as effective as cash transfers.

As the saying goes, if your intervention doesn’t outperform giving people cash, you should just give people cash. But there are interventions that outperform giving people cash. You can have economies of scale, in which it’s easier to purchase ten million bednets than for ten million people to each purchase an individual bednet. You can have situations where, through no fault of their own, the global poor are incapable of going on Google Scholar and researching the latest public health and development economics papers to influence their purchase decisions. And some interventions, like institutional reform, are really hard for individuals to do on their own without becoming a charity.

In short, donate to Give Directly if:

  • You’re a basic income nerd.
  • You care a lot about autonomy.
  • You care a lot about randomized controlled trials.
  • You want to be really, really sure that your money is doing something good, even if it might not be the best thing.