The GiveWell Blog

Effective altruism

We’re proud to be part of the nascent “effective altruist” movement. Effective altruism has been discussed elsewhere (see Peter Singer’s TED talk and Wikipedia); this post gives our take on what it is and isn’t.

What is effective altruism?
To us, “effective altruism” means trying to do as much good as possible with each dollar and each hour that we have. It’s a way of thinking about morality that insists on maximization of good accomplished, and not just satisficing of rules and guidelines. To us, this implies

  • Focusing on how one’s actions are likely to affect the world, rather than on how they affect oneself and one’s feelings. Effective altruism is consistent with believing that giving benefits the giver, but it’s not consistent with making this the driving goal of giving. Effective altruists often take pride in their willingness to give (either time or money) based on arguments that others might find too intellectual or abstract, and their refusal to give suboptimally even when a pitch is emotionally compelling. The primary/driving goal is to help others, not to feel good about oneself. (This doesn’t mean that effective altruists aren’t passionate about what they do or don’t obtain emotional rewards; it means feeling this passion and obtaining these rewards comes from focusing on impact.)
  • Being open to working on any cause, rather than committing to a cause up-front based on pre-existing personal interests. Strategic cause selection is based on the notion that one can do much more good working on some issues than on others.
  • Thinking of all beliefs as being open to change, and therefore potentially worthy of debate and analysis. Effective altruists research and debate a broad range of topics, from estimates of how to improve lives as much as possible per dollar to how such estimates should be used to what counts as a life to whether it’s better to give now or later (and many more). The consequences of a shift in views – including on relatively abstract topics – can be large, so one should not take any questions “off the table” by declaring that only one answer is acceptable.
  • Being open to unconventional approaches to doing good. For example, effective altruists often choose to devote themselves to for-profit activities, perhaps because they’re earning to give and perhaps because they believe these activities are themselves promising ways to improve the world. Doing lucrative for-profit work isn’t usually – or stereotypically – identified with being “humanitarian” or “altruistic,” but for effective altruists it can be a serious option.
  • Using one’s investigative resources efficiently. We’ll never have all the information we need to make knowably optimal decisions. Effective altruism means focusing one’s debates, analysis and research on questions that will bear tangible fruit in terms of informing our decisions about how to accomplish good. 

    Different groups in the movement have different views of what this means. Some believe it is important to focus on philosophical questions (such as “should we value enabling a birth similarly to averting a death?”) that seem in some sense “fundamental” and highly consequential in determining what outcomes we should aim for. We take a different view: we believe that questions like these, important though they may be, often aren’t amenable to tangible progress through further investigation. We prefer to focus our resources on questions that combine “importance” with “tractability to further investigation.”

Effective altruism is unusual and controversial
To many readers of this blog, the above qualities might sound like self-evidently good ones. But we don’t believe that’s how most people see them.

As GiveWell and Good Ventures have explored what causes to get involved in, the single most common advice we’ve gotten has been to “choose what you’re passionate about.” When we’ve described our desire to do “strategic cause selection” – choosing causes based on how we can accomplish the most good – we’ve seen a good deal of pushback and skepticism. It’s common for people to emphasize the importance of “starting from the heart,” and to fear that our commitment to a cause won’t be genuine (and won’t be robust) if it comes from a strategic, analysis-based choice.

This concern is reminiscent of David Brooks’s reaction to the idea of “earning to give”:

If you choose a profession that doesn’t arouse your everyday passion for the sake of serving instead some abstract faraway good, you might end up as a person who values the far over the near. You might become one of those people who loves humanity in general but not the particular humans immediately around … Instead of seeing yourself as one person deeply embedded in a particular community, you may end up coolly looking across humanity as a detached god.

…when most people pick a vocation, they don’t only want one that will be externally useful. They want one that they will enjoy, and that will make them a better person. They want to find that place, as the novelist Frederick Buechner put it, “where your deep gladness and the world’s deep hunger meet.”

I believe that these concerns misunderstand effective altruism. Effective altruism isn’t an alternative to having personal interests and passions; it is a personal interest and passion. Our next post will elaborate.


  • I like this summary; it feels fair and accurate to me.

    The one thing that troubles me about effective altruism is that it takes a donor-focused rather than needs-focused approach. Effective altruists aim to maximize the good they can accomplish. This can lead them to donate only to those charities whose effectiveness can be evaluated and verified, even if those charities are not addressing the most pressing needs of the world’s poorest people.

    For example, the World Food Programme claims that malnutrition kills more people each year than AIDS, malaria, and tuberculosis combined. And yet because GiveWell cannot yet recommend any nutrition-based charities (because data are lacking on effectiveness), effective altruists are not likely to contribute to addressing malnutrition. Plenty of other people are, of course contributing to food- and water-related charities, which also means that a contribution to one of those charities would most likely have a much smaller incremental impact than a contribution to AMF or another GiveWell-recommended charity. Therefore contributing to such a charity wouldn’t count as “doing the most good” even though from a needs-based perspective one could argue that providing adequate nutrition is a higher priority than preventing malaria.

    I think this is being addressed in current debates over what constitutes “the most good,” in terms of scope and time frame, but I still find it troubling that effective altruism seems to focus more on maximizing the donor’s impact than on meeting the most critical needs of the populations who most need our help. Maybe this is just a communication problem, in which case effective altruism needs to explain its aims more clearly. But I’m not convinced that maximizing personal impact always equates to “doing the most good.”

  • Brenton Mayer on August 13, 2013 at 10:55 pm said:

    @Brad, I think that when Effective Altruists talk about ‘effectiveness’, or of ‘maximising personal impact’, they precisely are talking about ‘doing the most good’, I don’t believe there is a meaningful distinction.

    It might be true that it would be better to eradicate malnutrition than to eradicate malaria, and therefore from a global perspective that would be better. But that doesn’t mean that from a ‘needs based perspective’ that using your money to prevent malaria should be a higher priority. If you’re dying from malaria then the needs of the globe are irrelevant to you, your most pressing need is the malaria. As you pointed out, if we consider how many of those individuals can be helped to what extent with the same resources, at the moment we should address malaria, just because the number of people whose most pressing concern we are able to alleviate is higher with a given amount of resources.

    So as I see it, we’re given the option of helping less people with their most pressing need by giving to alleviate a bigger problem or helping more people address their most pressing need by addressing a smaller problem. The answer to that is pretty simple yes?

  • @Breton: “The answer to that is pretty simple yes?”

    I appreciate the logic of your argument, but I don’t think the answer is simple at all. If a pressing need is so large and complex to address that it takes many decades of sustained effort by millions of donors and hundreds of charities, progress can be achieved through the combined impact of many individuals contributing, each of whose individual effectiveness is far smaller than what they could have achieved by tackling a smaller problem. In fact I think most large, systemic problems have this characteristic. It’s similar in some respects to voting, where your individual vote has no measurable impact but when combined with the votes of millions of other people can bring about change in leadership or (in the case of a referendum) a new policy. But it’s more complicated than voting because there are many ways for charities working on big systemic issues to do the wrong thing or simply to be ineffective, so the risk that you’re throwing your donations away is much higher.

    I think the distinction I’m trying to draw is that “doing the most good” as defined currently by many effective altruists means saving the greatest number of lives (or avoiding the greatest amount of suffering) per unit of currency donated. What’s unclear to me, though, is whether the most good could ultimately be done by giving some money to high-impact charities whose effectiveness can be verified now, while also contributing a (smaller) amount to charities working on big systemic issues that could erase many of today’s gains, but whose effectiveness can’t yet be verified.

    For example, climate change is expected to increase the risk of malaria in many places where malaria is already prevalent, but climate change is also projected to cause the spread of malaria into new regions that have never experienced malaria. If you’re trying to eradicate malaria, maybe you can do the most good by supporting a charity like AMF that helps avoid malaria today while also supporting a charity that monitors climate conditions and mosquito populations so adaptive actions can be taken to protect populations in currently malaria-free regions. Your current personal effectiveness will be lower (you’ll save fewer people today per dollar or pound donated), but you might end up doing more good in the end. The “might” in that statement is what many effective altruists would object to, because it represents uncertainty and risk, and yet if we concentrate only on funding interventions with a low risk of failure, there’s an opportunity cost. Bigger, more complex problems could effectively erase our gains down the road. A life we save from malaria today could be ended tomorrow from malnutrition, drought, lack of access to safe drinking water, etc.

    I understand that many big charities that work on systemic issues, such as Oxfam, receive many times more donations than smaller, more focused charities like AMF. People are paying attention to the big issues, which leaves room for effective altruists to focus on smaller ones where the impact per donation is higher. But if the big problems need all the help they can get, even if the incremental impact is small, I wonder if the “most good” can actually be achieved by some combination of donations to short-term and longer-term efforts. I don’t have the answer.

  • Jeff Kaufman on August 14, 2013 at 3:59 pm said:


    You seem to be saying that the goal is doing as much good as possible, but that’s not under “effective altruism”. That’s not consistent with the definition Holden gives in this post, or what I’ve seen elsewhere. If people who claim to be trying to be effective altruists are doing something that you think doesn’t do as much good as possible, I think it makes more sense to respond “you could be more effective in your altruism” than “effective altruism focuses on the wrong things”.

  • Holden on August 15, 2013 at 2:54 pm said:

    I agree with Jeff.

    I also agree that to date we’ve been very focused on measurable, short-term impact. Via GiveWell Labs, we are working on changing this. That said, I think that with the information we have, donating to our current top charities is at least as strong in terms of long-term expected progress on systemic issues as supporting groups that more explicitly work on systemic issues. More on my thinking at The Root Causes of Poverty and Flow-Through Effects.

  • AnthonyC on August 16, 2013 at 7:56 am said:


    More people also die of heart disease or cancer than malaria, but that doesn’t mean that curing those would be an efficient use of charity. More important in the long run, perhaps, but not an immediate win.

Comments are closed.