The GiveWell Blog

Don’t take our breadth away

For our first round of grantmaking, we chose to investigate five broad causes, two in the developing world (saving lives and fighting poverty) and three in the developed world (early childhood care; K-12 education; employment assistance).

This decision – doing five causes, instead of narrowing our scope from the outset – had serious costs. Even with all the time we’ve spent – far more than a typical donor can – we still feel that our understanding of each cause is very far from thorough. Trying to deal with all five causes at once has been logistically tough, and we even missed the deadline we set for ourselves for giving out our grants (we awarded only three grants at our December 2007 meeting, and will be awarding the remaining two within the next few weeks). In other words, we absolutely bit off more than we could chew.

Yet as we look to next year, I find myself wanting to bite off even more – to go broader, not narrower. Part of the reason is that doing five causes has allowed us to provide information to a broader set of potential donors; while we didn’t analyze these causes as deeply as I would have liked, I still feel that we have found is more than what existing donor resources offer, and far more than we could find when we were casual donors ourselves.

Doing five causes also had another benefit: while we didn’t become experts in any one cause, we learned enough about each to radically change the way we think about them – and prioritize them. Last August, I wrote that K-12 education was the cause that excited me most; yet having learned more about a donor’s options in each cause – and how much each costs, how reliable each seems, etc. – I’m now most excited about global health, and I’d rank K-12 education last out of our five causes. In a nutshell, though I most want to help Americans take advantage of all the opportunities America offers, I now believe that the developing world’s needs are so much more drastic that I’d rather help them. My new view is still open to revision, but it’s better informed than the one I came in with. And I wonder whether I’d be even more excited about something else, if we’d gotten a chance to look into charities that focus on abandoned children, or on disease research, or on the environment.

It may be more typical for a grantmaker to start its work by picking an area that it can really become an expert in. Donors are often told to start their search by volunteering, so they can see their area of interest up close. But if you care both about how you impact lives and how many lives you impact, I’d argue that your most important decision is the top-level, bird’s-eye one – whom you help, and how. I feel there’s more to be gained by learning a little about all your options than by getting it perfectly right within a predefined scope. Our approach is often frustrating in its lack of depth, but our goal isn’t to become the best at understanding any one area; it’s to give as well as possible.


  • michael vassar on February 20, 2008 at 2:47 am said:

    It seems to me that the devaluing of high level decision making in favor of near-random one-time choices followed by intense specialized effort and solipsistic loyalty to the community joined through the initially random choice is one of the increasingly severe cultural problems of contemporary America. It’s a serious problem in the context of giving, but also in practically every other context. One of its costs is that it strongly pushes people to continue to dedicate their time to what they know to be bad causes and goals simply because their human capital is specialized for working on those causes and goals.

  • Sean Stannard-Stockton on February 20, 2008 at 9:55 am said:

    I think the problem you’re facing has less to do with something innate to philanthropy or a function of your resources and is more a problem with the underdeveloped infrastructure in our industry.

    Investment managers who run stock portfolios generally hold stocks across many sectors/industries. But they are not trying to be experts in every field. Instead, they have access to an army of specialized analysts who focus on only 5-10 companies and understand all the dynamics in their sub-industry.

    Without these underlying experts, you have to do all the work yourself. If you look back at the financial markets in the 1920-1950 time period, you would find a similar underlying dynamic. At that time, investors might have held fewer stocks across fewer industries than they do today, but I think you would find support for your choice to focus across a range of “industries” rather than just picking one.

    I at least am glad that there is some breadth to your research and would rather see more causes than a deeper dive on what you’re already looking at. Hopefully you can do both.

  • Michael Vassar: Can you explain this more? What do you think explains these “random choices” in the first place? Are people acting locally without thinking globally? Why might this be? Do they perhaps know more about the “randomly” chosen organizations than about others they did not choose? Do they have a special affinity for the cause they support, because of a “random” life experience, perhaps? You’ve made quite a strong statement–that the way people choose causes to support is a “severe cultural problem”–but I’m not sure I follow your logic here. Could you explain, perhaps from a high level?

  • I think Sean is right, and that one thing Givewell is missing, and has been criticized for in the past, is that people are already working on these issues, and that one small organization like Givewell can’t solve this “problem” where all others have failed. For example, Holden’s point that he thinks global health is a cause ripe for funding has obviously already been discovered by none other than Bill Gates, employs many staff who research causes in the way Givewell proposes, but along the well-tested model Sean describes. Again, this points to the naive quality of the entire Givewell enterprise.

  • michael vassar on February 20, 2008 at 1:31 pm said:

    The analogies between investment and philanthropy are poor because personal wealth has strongly diminishing returns so risk aversion is rational while philanthropic spending has very weakly diminishing returns so risk aversion is only rational with VERY large portfolios. Consequentially, rational investors investigate many possible investments in low depth and choose a large number of them but rational philanthropists invest a diverse array of approaches but choose only one until that one is fully funded.

    The advantage of focusing on the big picture comes from the fact that the expected returns in philanthropy, unlike those in fully liquid investment sectors, varies radically between sectors. There is nothing like this in liquid investing. Tech stocks might systematically perform twice as well as industrials, or in another market one half as well, but they won’t ever have 100X the expected return. If you are committed to making liquid investments then its more rational to spend the effort to double your expected returns in a random industry than to try to find an industry that will do twice as well as another with a bird’s eye view. OTOH, if you look at non-liquid investments, the birds eye view becomes relevant. Someone deciding between investments in jewelry, Florida real estate, timber, baseball cards, great master paintings and clean energy start-ups may rationally expect the best performing of these to return 100X as much as the average and 10,000 times more than the worst. Deciding what broad class of investment to make is likely to be more important than the decisions made within the class so the birds eye view is appropriate even if you are risk neutral.

  • “It seems to me that the devaluing of high level decision making in favor of near-random one-time choices followed by intense specialized effort and solipsistic loyalty to the community joined through the initially random choice is one of the increasingly severe cultural problems of contemporary America.”

    Michael, I have to say that the above sentence wins my personal award for the most difficult-to-parse sentence I’ve read all day.


    Maybe it’s because I had to look up the meaning of the word “solipsistic”.

  • Anyways, I agree with Michael’s comment #5, that there is a much greater differential between expected returns among typical philanthropic choices than among typical investment choices.

    Lots of people have relatively poorly structured investments. They aren’t diversified enough, and therefore have too much risk relative to their expected return. (Or conversely, they aren’t being paid enough return to compensate for the risk they are taking). And some of them invest in mutual funds or perhaps hedge funds that charge high expenses for returns that can reasonably be expected to match or slightly underperform a simpler investment in the market via an inexpensive index fund.

    If I had to make a wild guess, I’d say that moving from a ‘typical’ portfolio for someone with a 6 figure portfolio to a still simple, but sound and diversified Vanguard portfolio might move their expected returns up by 1-4% annually, with slightly less risk than they had before. With further research, there are a few more things that might push your returns up another percent or two.

    But I think with the same amount of research put into charitable giving, you can probably move your impact up MUCH more.

    (Well, it’s tricky, because portfolios are usually measured over time, and compounding effects may magnify 1-4% annual differences, given decades of compounding, but still).

    For many folks, charity research doesn’t make too much difference. If you’re only giving away $100 per year, then spending a lot of time to improve the effectiveness of that donation is not going to have too much net impact. But at higher levels of giving, it’s, IMO, very important to give in a reasonably informed manner.

  • Final comment for now…

    The original post, as well as several of the comments, suggest that the bird’s eye view is very important. I would agree. I think it’s more important to make the right decision between, say, African health and US K-12 education than to make the right decision WITHIN the chosen area (within reason – of course one doesn’t want to support a scam organization in any field, and one should make reasonable efforts to find a good organization).

    In turn, this opens up the question of how should GiveWell’s research be focused. Is it better to choose 5 or so topics based more or less on gut feel, then find a few exemplars within each topic (which in turn, allows readers to compare the topics to an extent), or is it better to focus on the bird’s eye view in the first place? i.e. I’ve thought for a while that GiveWell should put more emphasis on high level charity issues – mapping the charitable landscape and the bird’s eye view. But I see merit in the approach that GiveWell has taken, as well.

  • MV, I think you are also describing short-term versus long-term investing. Another major difference is that there is a far simpler bottom line when investing to make money than when investing to improve the human condition. In the latter case it could be perfectly rational for a donor to choose a relatively less effective organization or method if that donor has judged that he or she values that cause. I know you are trying as hard as you can to divide the world up between “rational” people and everybody else, but I don’t think you are going to be able to do that with your starting point. You are over-thinking some of these issues.

  • michael vassar on February 20, 2008 at 6:39 pm said:

    Andrea: thinking globally is a very unusual and unnatural thing for humans to do. like many other uncommon beneficial activities, when they actually do it the consequences are likely to be very large.

    Think how little information children have about any given career when they commit years of their life to a career track pointing in that direction, or for the most elementary example think of the supposed conviction that most people have in whatever religion they are born into. Great effort is expended in a direction determined by a non-deliberative decision despite great disagreement between agents. For a classic experiment on this, see the Robber’s Cave experiment

    As a strong general rule, a person is acting unwisely and irrationally if they invest great prolonged effort in any task without great deliberation although others in approximately their circumstances dedicate their efforts to a different task.

  • MV, Yes, this is the human condition. Not quite sure where you’re going with this, unless it’s to solve the human condition, which I have a sneaking suspicion you think you have done. Or, some might think it’s bs.

  • michael vassar on February 21, 2008 at 11:16 am said:

    Michael: People tend to act in the following unwise manner…

    Andrea: Yes, they do, and you are a person so you probably do it too nya nya

    Michael: OK then, if you don’t believe in trying to do better than the behavioral patterns that humans default to then suit yourself. However, since the point of this blog and this organization is to try to do better in a particular way we aren’t going to care much what you say about the matter if you don’t think that trying to do better is a good idea.

  • MV: You missed the point entirely. You said, and implied, far more in your posts about how irrational people are, than that we should simply try to improve. Nobody’s against improvement, just wasting time going in unproductive directions. I believe that’s because you, like Holden, are naive in the way kids are who are used to being praised for high SAT scores, but who don’t have enough experience yet to be able to judge the value of their own conclusions. Young academically accomplished people who get used to people calling them “brilliant” are handicapped by lack of humility. If you want to dismiss this description of your thinking as naive, which is perfectly legitimate, by mocking with “nya nya” go ahead, but you haven’t learned anything from the feedback, and therein lies your problem. We’re not going to bio-engineer our way out of being human. That kind of thinking used to be called “utopian.” I would be interested in hearing what you have to say in 50 years, when you’ve (one hopes) out-grown your wise acre, naïve belief that your high SAT scores somehow mean you know better than all the wisest, smartest human beings who’ve thought about things for centuries.

  • I agree on breadth over depth. I think that on the breadth issue you can continue to spark a very useful, much needed discussion about how priorities are chosen in the philanthropic sector. You also will, I think, be closer to the big questions that ought to trouble the average donor, who is unlikely to become an issue expert or hire issue experts (a la a foundation) to direct their giving.

    To employ an inadequate analogy from the political rather than financial sector: it’s the equivalent of helping distinguish between political parties in the general election rather than focusing on the differences between candidates in the primaries. It’s not that parsing the details of the primary candidates’ positions isn’t important, it’s that the differences tend to pale in comparison to those between the parties. That’s why in most cases (perhaps the current presidential primaries excepted…), the first is usually the hobby of small number of engaged individuals, while the latter is of much more widespread concern.

  • Sean Stannard-Stockton on February 21, 2008 at 9:46 pm said:

    Michael, your point about the lack of benefit to diversification in philanthropy is something I agree with strongly and have discussed in other forums. My point was not that GiveWell should cover multiple sectors in order to reduce risk. I was just saying that it is more difficult to cover multiple sectors in philanthropy because the underlying experts are not there (or not contained within an easily accessible system).

    If GiveWell was only concerned with maximizing the impact of their own funds (as most foundations are) I would suggest narrowing their focus. But the part of the GiveWell experiment I find most interesting is the way it can help inform other donors. If they have a narrow focus, they will only be a resource to donors with that same narrow interest. If they have a broad focus, they will be able to attract multiple donors and more people can learn something from the process.

  • michael vassar on February 23, 2008 at 1:53 am said:

    Sean: I think that we agree on all the points you discussed.

  • Holned Karpronsky on February 28, 2008 at 8:01 am said:

    It’s obvious that Holden, Ellie and the whole GiveUp crew are incompetent and, worse by far, self-important imbeciles whose main objective is to “bite off more than [they] can chew”.

    The obvious solution is: JUST SUCK ON IT, GiveWell.

    If this comment gets deleted it’s just because the moderators can’t deal with the transparency of it.

Comments are closed.