[This is the third post in a brief series explaining the current GiveWell top charities. You can get all the information in this post on GiveWell’s website, but my blog post is both shorter and less boring. In order to reward you for reading a whole blog post about a charity you already know about, I have included at the end the Most Interesting GiveDirectly Fact.]
Whoever said that you should buy warm fuzzies and utilons separately had never heard of GiveDirectly.
GiveDirectly gives money to people in the developing world, usually about $1000. This typically about doubles its recipients’ yearly income. You may see a more-or-less random sample of recipients’ responses to this program on GD Live. Recipients’ responses are unedited and only posted if they opt in to sharing. They may seem slightly overenthusiastic but imagine how you would feel if someone doubled your income for no reason. (I recommend only reading GD Live if you’re pretty sure you plan to donate to GiveDirectly; it has a known effect of causing people to be unable to resist the temptation to donate.)
Research on the effects of GiveDirectly’s cash transfers shows the following:
- Households spent $51 more per month; about half of the money was spent on food.
- Household assets increase by $463; most commonly, the money was spent on livestock, durable goods (particularly furniture), and savings.
- 23% of recipients had an iron roof, compared to 16% of controls.
- Households spend $13 more per month on business expenses, typically non-durable expenses on non-agricultural businesses.
- Recipients spend $3 more per month on health expenses.
- Spending on alcohol and tobacco did not increase.
- Food security and sense of psychological well-being increased.
- Business revenues increased by $15; profits did not increase, but that might be a short-term effect due to e.g. investments that have not yet paid off.
Also, one guy bought a guitar and used it to write this catchy song:
My life is GiveDirectly
My house is GiveDirectly
My phone is GiveDirectly
My job is GiveDirectly
My love is GiveDirectlyMy life, Give Direct
My house, Give Direct
My farm, Give Direct
I love Give DirectMy life, Give Direct
My house, Give Direct
My job, Give Direct
I love Give Direct
While there is high uncertainty, GiveWell’s best guess is that GiveDirectly does not have significant negative effects on households that don’t receive money.
GiveDirectly is a very well-run charity. 99.7% of recipients receive all funds promised. While staff fraud has occurred in the past, GiveDirectly has responded promptly and taken more steps to prevent future fraud.
GiveDirectly is also known for its commitment to randomized controlled trials. A very high proportion of its recipients (although not 100%) are enrolled in a randomized controlled trial, such as the RCT on the macroeconomic effects of cash transfers or GiveDirectly’s basic income trial. (Oh yeah! GiveDirectly is totally doing a trial on whether basic income cost-effectively improves people’s lives!)
GiveDirectly’s room for more funding is huge, because of how easy the program is to scale. GiveDirectly alone could productively use hundreds of millions of dollars in funding– more than every other GiveWell top charity combined.
Many people support GiveDirectly not just because of its program but because of its challenge to the international aid sector. GiveDirectly asks, “if the thing you’re doing doesn’t work better than giving people cash, why the fuck aren’t you just giving them cash?” Of course, this is a very hard impact to measure, and it’s unclear if giving GiveDirectly more money would cause them to have more effect on the rest of the aid sector, compared to GiveDirectly existing at all. But anecdotally more funders are asking themselves “does this outperform cash?” and give-cash control groups have expanded in popularity.
Many other people support GiveDirectly because they care about autonomy. A lot of donors to the developing world are incredibly condescending: they buy a cow from Heifer International or a merry-go-round pump from PlayPumps or help build a school on their mission trip. Surely, however, if people in the developing world need cows or schools or merry-go-round water pumps, they can buy that themselves? Is there some reason to believe that we know better than people in the developing world just because we’re rich? By far the most popular purchase with a GiveDirectly cash transfer is an iron roof; have you ever seen a charity fundraising to buy iron roofs for people in Kenya? We don’t know what it’s like to be members of the global poor, because we’ve never been poor. The poor know better than we do what their needs are. It is respectful of their dignity as people to let them make this sort of choice.
So, GiveDirectly is great. Why don’t we all just donate there?
Well, GiveWell has this handy little chart where they calculate what you get for your dollar. It’s super-fake and you shouldn’t take it literally, but the effect is large enough that that doesn’t necessarily matter. And what it found was that the cost to achieve an outcome that is just as good as saving a five-year-old’s life, according to the median values of GiveWell staffers, is twenty-three thousand dollars.
Now, of course, that depends on your values, and I encourage you to put your own moral weights into the sheet instead of relying on the median of GiveWell staffers’ values, which is a terrible way to do ethics. But if you care about saving the lives of small children and/or don’t mind high levels of uncertainty, you will get more value for your money by donating to a different GiveWell top charity.
The problem with “do we know better than people in the developing world just because we’re rich?” is that the answer is “yeah, sometimes.” Very, very few Kenyans have access to Sci-Hub so that they can develop an informed opinion on whether deworming medicines will increase their children’s income twenty years from now. Few Ugandans can explain the connection between those vitamin A pills the health worker is handing out and their children dying of diarrhea. Of course, the same things are true for most people in the developed world, but there exist any people in the developed world who understand those things, and the rest of us can follow their donation recommendations.
People in general tend to undervalue preventative health care. People in Africa don’t buy malaria nets for the same reason you never use your gym membership. If it works, nothing happens: you don’t see the malaria or heart attack you didn’t get. Nothing disastrous will happen if you put off going to the gym or buying the malaria net till next week, and exercising is boring and if you buy the malaria net you won’t get to eat for two days, so you never get around to buying it. As an outsider, you can say “this is a known cognitive bias, I’m going to give you a free malaria net and then your children won’t die of malaria.” So charities other than GiveDirectly can be more cost-effective.
In conclusion, as promised, here is the Best GiveDirectly Fact:
GiveDirectly started enrolling recipients from Homa Bay county, Kenya in mid-2015. There, it encountered unexpectedly high rates of refusals from potential recipients; while refusal rates in Uganda and Siaya, Kenya have historically been low (around 5%), refusal rates in Homa Bay have been about 45%. GiveDirectly believes the refusals are due to widespread skepticism towards GiveDirectly’s program and rumors that GiveDirectly is associated with the devil…
November 2017 update: We requested data from GiveDirectly on refusal rates (and other metrics) for January to March 2017 for Kenya and Uganda and March to June 2017 for Rwanda. Refusal rates remained fairly high in Kenya, with 22% refusing to participate in the census and 68% of complaints submitted to GiveDirectly categorized as “program is evil/from the devil.”
Why might you donate to GiveDirectly?
- You need a lot of warmfuzzies in order to motivate yourself to donate.
- You think encouraging cash benchmarking is really important, and giving GiveDirectly more money will help that.
- You want to encourage charities to do more RCTs on their programs by rewarding the charity that does that most enthusiastically.
- You care about increasing people’s happiness and don’t care about saving the lives of small children, and prefer a certainty of a somewhat good outcome to a small chance of a very good outcome.
- You believe, in principle, that we should let people make their own decisions about their lives.
- You want an intervention that definitely has at least a small positive effect.
- You have just looked at GDLive and are no longer responsible for your actions.
Aapje said:
That song screams dependency, which is not necessarily a good thing.
In general, this seems like a highly experimental program, where GiveDirectly runs various experiments, including lump sum payments, a short term UBI (2 years) and a long term UBI (12 years). To their credit, they also have controls, where they follow villages that don’t get aid. I think that Ozy presents this too much like an established program, rather than an experiment that just got started.
It seems that the statistics on the effects of the program are mostly gotten through self-reporting. There is an obvious incentive to report that you are spending the money on things that you expect the donor to approve of, and such.
Now, I’m supportive of doing experiments, but I think that it should be made clear to people that it is very experimental still and that interventions can very easily seem more beneficial than they are.
LikeLiked by 1 person
ozymandias said:
GiveDirectly has been in operation for eleven years and cash transfers (albeit conditional cash transfers) were extensively studied before GiveDirectly existed. I’m not sure in what sense this isn’t an established program, except “they are running experiments on their programs.” But (if it’s possible) everyone should be running experiments on their programs! As we’ve seen in earlier posts in this series, such as on Helen Keller International, circumstances change and programs that were cost-effective can become cost-ineffective. It seems weird to punish GiveDirectly for doing something more charities should be doing.
LikeLike
Aapje said:
They’ve been in operation for 11 years in the sense that they started with four guys giving money directly to people in 2008. In 2012, they created a proper non-profit with bigger funds. That gradually developed and only in 2016 did they announce the big UBI experiment.
GiveDirectly say themselves that there still isn’t enough evidence to be sure that it has substantial positive long term effects (instead of subsidized apathy or such) and that we need the rigorous experiment they announced in 2016 to actually see if it works long term and to what extent.
So I am just pointing at what they themselves are telling us.
Secondly, I don’t know why you think that I am saying that people should punish them. I explicitly said that experiments are a good idea. I just think that it should be made clear to people that the long term outcomes are unclear and that they should be aware of this, so they don’t feel misled if the experiments don’t work out.
A lot of ‘rationalists’ are libertarians who actually might be super-duper excited because of, not despite, it being an experiment that may result in validation for a UBI in general & thus may also have an impact on societies outside of Africa (like the West).
Your take on this may actually be less effective at selling this charity than if you framed it as the largest UBI experiment in the world.
LikeLike
ozymandias said:
You’re linking to their UBI experiment. That doesn’t say anything about what GiveDirectly normally does (lump-sum cash transfers). There is no theoretical reason to believe that lump-sum cash transfers cause dependency, since people expect to only get them once, nor do they appear to cause dependency in the research.
Lump-sum cash transfers have been very well-studied; in fact, GiveDirectly is probably the best-studied of the GiveWell top charities.
I’m not trying to sell people on charities. I’m trying to explain the pros and cons about donating to charities so people can make informed decisions. If I were trying to sell people on charities I would have left out the whole section about how GiveDirectly is ten times less effective than the other GiveWell top charities under many sets of values.
LikeLiked by 1 person
Aapje said:
I looked some more. They seem to have three different programs, ‘standard’ cash transfers, cash transfers to refugees and the UBI experiment. When donating, you can pick which program to donate to. By default, the option is set to let them decide.
I see no data on how much they spend for each program and/or how your donation is most likely to be used if you don’t choose yourself.
I looked at one of the studies of the GiveDirectly program that shows that the lump sum payments and monthly payments result in different spending patterns, which can be interpreted in different ways, so it’s unclear to me what is better.
LikeLike
Rachael said:
Typo: first post should be third post.
LikeLike
tailcalled said:
Now I’m curious: What are the advantages of iron-roofed houses and how much do they cost?
LikeLike
Lambert said:
What kind of rooves are the iron ones replacing? If it’s some kind of thatch, I’d imagine the main advantage is not being made out of grass. Specifically it doesn’t rot, doesn’t leak, things don’t try to eat it and it’s generally a lot lower maintenance.
Corrugated iron seems to be around £20 ($30 ish) m^-2
https://www.themetalstore.co.uk/products/corrugated-roofing-sheet
LikeLiked by 1 person
F Selker said:
Metal roofs seem to reduce malaria risk and perhaps other insects that live in thatch:
https://www.nature.com/articles/s41598-019-43816-0
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5995440/
LikeLike
Jared said:
Typo: “according to the average values of GiveDirectly staffers” should say “GiveWell staffers”.
LikeLiked by 1 person
ozymandias said:
Gah, thanks.
LikeLike
Sniffnoy said:
Whoever said that you should buy warm fuzzies and utilons separately had never heard of GiveDirectly.
C’mon Ozy you know perfectly well who that was. 😛
(It was Eliezer Yudkowsky, for everyone here who haven’t read that bit of his writing.)
LikeLiked by 1 person
michaelkeenan0 said:
> I encourage you to put your own moral weights into the sheet instead of relying on the average of GiveWell staffers’ values, which is a terrible way to do ethics
I’d be interested in a blog-post-length discussion of why this is, and in what circumstances (if ever) it might not be the case. Averaging the opinions of a bunch of smart and knowledgeable people is a good method for a lot of things, like estimating jellybeans in a jar or whether psychology papers will replicate or broad domains like “Who Wants To Be A Millionaire” questions. I’d like to read about why ethics is different.
Does ethics being different rely on the ethicist having a lot of time and interest in researching the relevant ethical issues? If you don’t have time for that – maybe because you’re Jeff Bezos and you’re busy running Amazon, or because you find ethical discussions boring or aggravating, or you’d love to but researching your relationship/nutrition/trauma/career/other problems is your top priority right now – is relying on the GiveWell average the next-best thing? If not, what is the next-best thing?
The GiveWell average seems sort of similar to the enlightened preferences method I learned about from Bryan Caplan. It’s a shortcut to finding out what we might think if we were more knowledgeable. You survey people, and find people who are like you in whatever predicts your opinions (demographics, background, etc.), and then find the opinions of people who are like you, except they’re more knowledgeable/smarter. I suspect the GiveWell people are a lot like me, but more knowledgeable, maybe so knowledgeable that I couldn’t catch up in dozens of hours of researching.
LikeLiked by 2 people