Thursday, December 19, 2013

Deciding where to give

I kept delaying writing this post because I'm not entirely sure where I'm donating over the coming year. I tend to spread donations out throughout the year, so I don't feel a particular rush in December. But, since the end of the tax year is a lot of people's giving season, I want to share my thoughts.

My previous favorite, the Against Malaria Foundation, is not looking so good lately due to problems distributing their mosquito nets. I'm not super concerned about this – I think it's likely that they'll get their act together in the next six months or so and resume distributing nets. But I don't plan to donate to them until that happens. (This is the kind of thing I'm really glad GiveWell is here to tell me about – as a donor, I would never have picked up on the problem.)

My parents usually make a donation as a Christmas present to me, and they asked where they should make it.  I asked them to give to Give Directly.  I think the just-give-cash method makes a lot of sense, given the evidence that recipients choose quite well what they need, and I hope this idea will catch on.  The organization itself seems quite transparent and sensibly run.  I also found it encouraging that so many GiveWell staff are making their personal donations there.

I can see advantages to donating later, especially given the fact that the Against Malaria Foundation (which looked like the strongest contender for a good while) may catch up again soon.  So I'll probably hold most of my donations for about six months and see how things look then.

If I were holding off longer, like years, I would probably put my donations in a donor-advised fund like GiveWell's or Giving What We Can's.

I also plan to donate some to organizations like GiveWell, Giving What We Can, or 80,000 Hours (I haven't decided on the breakdown yet).  I think this is an important time both for exposing more people to the idea of effective altruism and for developing better knowledge about where to give.  I need to look more specifically into what those organizations would do with more funding.

Lastly, I'll make some feel-good donations to causes that aren't the most effective (a fundraiser for a friend's business, my Quaker meeting, etc.) These donations come out of my personal spending budget, not my charity budget.

One obvious question is: why divide up money rather than giving all to one place? Giving What We Can makes some good points in their recent post about why to donate all to one charity.  But I'm also persuaded that we should act how we want other people to act, and I wouldn't want the whole effective altruist community to donate to only one place.  So I'm okay with dividing things up a bit.


  1. Hi Julia, I'm a new reader of your blog and I'm pretty impressed by your philosophy of life. I wanted to ask you however why you suggest GiveDirectly. I quote some passages from GiveWell website that made me feel doubtiful:

    What were the overall effects of GiveDirectly?

    Food security. At baseline, food security was low among participants.146 Program participants reported a 0.25 standard deviation increase in a food security index over controls.147 Some food security related variables improved more for recipients of large transfers, but the study did not have high enough power to determine whether there was an effect on the overall index.148
    Health and education. The study did not detect an effect on indices of health and educational outcomes.149
    Revenue and profits. Receipt of transfers lead to a $17 per month increase in total revenues but no detectable increase in profits.150 We emphasize that these are very short-run effects and we do not know whether participants’ business investments might lead to profits in the longer run. There is no evidence that large transfers had greater effects on revenue or profits."

    To me it feels like we guarantee food security to one family for one year and then... ? What is the change on the long run we have promoted? Wouldn't business-oriented microfinance work better?

  2. Suturn,

    It would be nice to see across-the-board improvements in aspects like health, finance, education, nutrition, gender equality, etc. from a single intervention, but that might be too much to hope for. It looks like Give Directly recipients had improvements in some of those areas but not others.

    It does look like long-term finances were improved by the cash transfers: "There is also some evidence that recipients are able to invest cash transfers at high rates of return (e.g. ~20% per year), leading to long-term increases in consumption." They also cite the fact that many people invested in metal rather than thatch roofs, with an estimated 7-14% financial return (not counting the non-financial benefits of having a roof that doesn't leak on you and all your belongings.)

    The part about "increase in total revenues but no detectable increase in profits" seems to refer to business enterprises started by recipients. (I'm not totally sure on this, and just wrote GiveWell to check.) So maybe I spend $100 on a bicycle and earn $100 by running errands with my bicycle. I have higher revenue but no profit. If this is the case, "business-oriented microfinance" doesn't look so good - unless such programs have some aspect that help small businesses become more profitable. I don't know specifics on such programs.

  3. Hi Suturn and Julia:

    I work for GiveWell and wanted to address these thoughtful comments and questions. (Thanks for emailing, Julia!)

    The question of the long-term impacts of cash transfers is basically an open one, we think. A number of studies of cash transfer programs that are different from GiveDirectly's program in a variety of ways have found very high long-term returns on investment from people using funds to invest in businesses or human capital. The RCT of GiveDirectly that we summarize only follows up over one year, so we don't know yet what the longer-term impact of contributions to GiveDirectly will be. The high estimated returns on metal roofs and the increase in business revenue make it plausible to me that the long term returns to cash transfers could be quite high, but the lack of measured increases in profits makes it hard to know what to expect from the follow-up studies.

    Julia, your interpretation of the difference between revenue and profits is exactly correct, but it's worth noting that I believe the authors tried to separate out normal business spending from durable asset purchases like a bike. So a more accurate way to think about it might be that you bought $100 worth of fuel and rental time for a motorcycle and made $100 in revenue driving people around, leaving no measurable profit. It's also worth noting that measuring profits in businesses that don't keep good written records is notoriously difficult, so it's a little unclear how much stock to put in the lack of results on that front.

    One more thing I'd note is that I don't think it's right to focus too narrowly on the food security results. The RCT also find significant increases in psychological well-being and in non-food consumption spending, I find it easy to value both of those things. This isn't to say that the value is lasting, just that it might be more multidimensional than "we guarantee food security to one family for one year."

    Thanks again for the thoughtful comments and questions!


  4. Hi Alexander and Julia, thank you for your answers. Many microfinance projects focus on both giving the finances and the know-how to set up profitable businesses thus promoting long-term improvements.

    For example a woman receives a loan and participates to a class on how to grow a certain crop. She can then start to make profit, gradually pay back the loan. If she wants she can apply for a new loan for a new investment, for example to improve the house condition because now she has the means to pay it back.

    Isn't such a model more likely to bring more long term improvements?

  5. Suturn,
    That model does sound good, as do many models. Before donating to a particular charity, I would want to know that the intervention itself had been demonstrated to be good, and that the particular charity was carrying out the intervention in an effective way. I don't have the time or expertise to do this research, so I tend to trust evaluators like Innovations for Poverty Action, Poverty Action Labs, and GiveWell.

    If you know of organizations that you think are doing excellent work, maybe they should submit data to GiveWell to be considered as a recommended charity.

  6. I understand and appreciate your attitude, on the other hand I think we should keep on making questions on our own rather than just surrend to the experts.
    To summarize, the thorough research promoted by GiveDirectly so far has shown an improvement in food security, psychological well-being and no measurable profit. So should I choose them just because they use proper research instruments or shall I look for a model that seems to work better, even though they didn't use research to demonstrate it? In my opinion, if we're concerned with effective altruism, the question remains an open one.

    1. I certainly don't consider the question closed. But when the evidence for one model is "we looked into it carefully and it wasn't everything we were hoping for, but it had some good effects" and the evidence for another model is "it sounds really good", that's not really enough for me. Especially when there are actual studies on the second model indicate "It's also not as good as we were hoping for, and in fact doesn't seem to work any better than the first model."

      There's a tendency for rigorously studied interventions to look bad because they didn't do everything that's claimed of other interventions. But given that other interventions' claims often don't prove to be accurate, I'm going to wait for evidence in most cases. The major exception I can think of is new projects or interventions that haven't had a chance to collect data yet, in which case I would still want some information (is their staff good? etc.).

  7. Suturn-

    Microfinance of the kind you describe has been studied in several randomized controlled trials, and I think the results are considerably more negative than the story you're telling would suggest (and more negative than the results from the RCTs of cash transfers). This literature is constantly evolving, but this blog post by David Roodman is a good place to start:


    1. Hi Alexander, by browsing several articles on the website you suggested me I didn't find any conclusive evidence of what you say. Rather, it seems the results are pretty different for different projects and Microcredit is generally helpful for those who have an enterpreneurial attitude.

      Moreover, I don't understand what's GiveDirectly final goal. They'll give money for one year to a family and then?

    2. Hi Suturn,

      I'm not sure what's leading you to that conclusion. My read of the post I linked to above is that the most rigorous studies of micro loans to date have looked for, but generally not found, evidence of reductions in poverty.

      I think GiveDirectly might say their goal is that recipients are empowered to use the funds spent on their behalf in the way that they determine is best, though you'd have to ask them yourself if you're interested in their take. You might find this page a helpful place to start:


  8. It's worth noting that many GiveDirectly recipients have used the money to buy a metal roof and/or a cow. Both of those things represent long-term investments. Maintaining grass roofs is expensive (in most cases the grass for roofs has to be purchased) and grass roofs correlate with higher rates of malaria and other illnesses (mosquitoes can easily find their way through grass roofs). Cows produce income from milk production, the sale of calves, and the eventual sale of the cow itself. So in many cases recipients of funds from GiveDirectly benefit from those funds well beyond the period of just one year.