Against Swooping In

When effective altruism proves to be less than effective, and also rife with political problems.

Sigh. For years, in my introduction to ethics courses as a philosophy professor at Cornell, I’ve taught Peter Singer’s famous argument to the conclusion that we, the presumptively affluent, are morally obligated to give away much more of our incomes to charity than we do customarily. Here’s the more committal form of the argument, as I present it in my lectures, taking the quoted sentences verbatim from Singer’s article: 

Premise 1: “Suffering and death from lack of food, shelter, and medical care are bad.”

Premise 2: “If it is in our power to prevent something bad from happening, without thereby sacrificing anything of comparable moral importance, we ought, morally, to do it.”

Premise 3: It is in our power to prevent suffering and death from lack of food, shelter, and medical care by sending money to a worthy charity.

Premise 4: In sending money to a worthy charity, we would not thereby be sacrificing anything of comparable moral importance, unless we need the money to ensure the basic health and survival of ourselves or our relations.

Conclusion: Therefore, we ought morally to send money to a worthy charity (e.g., those recommended by GiveWell), up to the point of marginal utility (or at least the risk of serious suffering) for ourselves or our relations—that is, the point at which we’d lose more by giving than they’d gain by getting.

So far, so good, more or less. There’s certainly room to disagree with the argument, but precisely for that reason, it seemed to me to provide a good basis for student engagement and discussion. 

That being said, I often worry about assigning a piece that effectively assumes that students have, or will soon have, disposable income. I point out that that’s not only not a safe assumption, but it is essentially a biased one. After all, white families in this country notoriously have nearly eight times the median wealth of Black ones. And building a safety net for you and yours, given the egregious history of redlining and similar forms of racism in this country, could hardly be a more legitimate goal for BIPOC students (among others).

One of the texts I always assign alongside Singer is Tressie McMillan Cottom’s characteristically brilliant piece, “Why Do Poor People ‘Waste’ Money on Luxury Goods?” which makes the point that what counts as a luxury good for a white person may be a literal lifeline for a person of color. A camel-colored suit and pearl earrings helped her mother get her neighbor the benefits to which she was entitled at the social service agency. The lack of a silk shell cost a Black woman a desk job at a cosmetology school McMillan Cottom once worked for. It was denied to her by a Black VP, notably. Looking neat and presentable in minimally respectable, clean clothes is enough for someone with Singer’s privilege and prestige; for a poor Black woman, it may be grossly insufficient. As McMillan Cottom writes:

Gatekeeping is a complex job of managing boundaries that do not just define others but that also define ourselves. Status symbols—silk shells, designer shoes, luxury handbags—become keys to unlock these gates. If I need a job that will save my lower back and move my baby from medicaid to an HMO, how much should I spend signaling to people like my former VP that I will not compromise her status by opening the door to me? 

There are other class and gender issues related to Singer’s argument that have long given me pause too. On the former score, it’s been shown repeatedly that poorer people are more generous in their charitable donations. So, in a way, teaching his argument in a diverse classroom setting (including, yes, at Cornell) targets the people who are most likely to need to accumulate wealth to secure their own futures, and who are the most susceptible to having it extracted due to their moral conscientiousness. 

I don’t know what to do with that. But we do talk about it in my class—together with the fact that Singer speaks only of moral obligations. And the flipside of an obligation is, in general, an entitlement. There’s something strange about making such a sweeping argument in ethics that assumes that its addressees are solely subject to the obligations in question, and are not among those who are entitled—in this case, to better and more from a world that has neglected or betrayed them.

Singer has a similar tendency to assume in his piece that “the poor” are a group of passive recipients of aid, not agents and vital participants in—indeed, the rightful leaders of—ongoing efforts and conversations about how to improve their situation. (“Nothing about us without us,” as the slogan goes.) This often involves fighting not just material conditions but the ongoing exploitation and profiteering of dominant white elites and their favored institutions.

But a more structural and political conception of the barriers to people’s well-being—in this country as well as developing nations—is not really in Singer’s wheelhouse.

A word about gender is in order here too. Anecdotally at least, it often seems like my female students feel the guiltiest for the small luxury purchases they make that sometimes get them through the day. True, some of these may be excessive, and it may simply be true that these students ought to be giving more to charity in lieu of such frivolous purchases. But even something as seemingly trivial as a pair of new shoes can help some women to feel more comfortable in the bodies that we’ve been taught to rue and wrestle with. For fat women in particular, investing in clothes that make you feel comfortable and powerful is crucial to being able to make your way in a world that often derogates, even hates, your body and your person along with it. And fashionable clothes for fat bodies are notoriously expensive. Singer’s argument never even begins to consider these different reasons for such expenditure. He writes:

When we buy new clothes not to keep ourselves warm but to look ‘well-dressed’ we are not providing for any important need. We would not be sacrificing anything significant if we were to continue to wear our old clothes, and give the money to famine relief.

The royal “we” is as inapt here as it is irritating.

I’m also reminded of the story in Larissa MacFarquhar’s excellent book about a woman, Julia, swept up by Singer’s effective altruist movement (of which more shortly) who longed for a candy apple at a fair. It cost (an admittedly exorbitant) $4, she recalls. When her boyfriend Jeff bought her one as a treat, she was consumed with sadness later. She couldn’t even look back on the experience fondly, so great was her sense of guilt at this small, sweet extravagance. Instead, she wept bitterly. For:

Julia realized that, if Jeff was going to start giving away his earnings [via effective altruism], then, by asking him to buy her the apple, she had spent money that might have been given. With her selfish, ridiculous desire for a candy apple, she might have deprived a family of an anti-malarial bed net or deworming medicine that might have saved the life of one of its children. The more she thought about this, the more horrific and unbearable it seemed to her, and she started to cry. She cried for a long time.

As a dedicated opponent of diet culture, as well as fatphobia, this story makes me deeply uncomfortable. As did the occasional female student who came to my office to discuss a concrete plan for how they could give up more of what they had, following our class discussions of Singer’s paper. I would remind them that, yes, giving is good and important. But women are not society’s giving tree—recent SCOTUS rulings notwithstanding—and it is important to hold onto that as a bulwark against our culture of exploitative misogyny.

So. I am certainly no newcomer to politically-inflected criticisms of Singer’s famous argument. I have dutifully taught it anyway, with some pretty deep reservations and worries, out of a vague sense of obligation to assign perhaps the most popular contemporary paper in introductory ethics courses. 

But there is an aspect of so-called effective altruism that, as a philosopher, I naïvely never thought to question: the information upon which our charitable donations might be being directed by Singer and his ilk might itself be completely unreliable. And the ongoing confidence of the recommendations belies the ongoing scientific controversy over some of them.

Specifically, Singer’s favored source of information on where we should direct charitable donations, GiveWell, has been advocating for many years for programs that involve deworming children in the developing world. (This is one of their top recommendations, along with programs that distribute malaria nets and similar.) The idea was based on a 2004 paper that argued that these measures increased children’s attendance in schools, which might be assumed to increase good outcomes like educational achievement and social mobility. It was theorized that severe worm infections were reducing children’s ability to focus and overall health, most likely by causing chronic, endemic anemia.

The trouble is, this paper was mistaken. Along with major methodological flaws, such as not separating out the effects of education on sanitation from giving children the deworming pills themselves, it made a serious (inadvertent) calculating error. Basically, some of the coding was done erroneously in the equivalent of an excel spreadsheet, leading to arithmetical mistakes proliferating throughout their findings.

I learned of this on Tuesday from the new episode of Aubrey Gordon’s and Michael Hobbes’s excellent podcast Maintenance Phase. But they are tracing a controversy that apparently rocked methodology Twitter back in 2015. I am ashamed to say I hadn’t heard of it. Meanwhile, GiveWell has doubled down on advocating deworming programs, on very flimsy grounds. And Singer has continued to tout GiveWell as the organization that can make our giving more rational.

In a 2015 piece for The Boston Review, “The Logic of Effective Altruism,” Singer criticizes “‘warm glow givers’ for being “not so interested in whether what they are doing helps others.” Indeed, he holds:

Most [charitable donations] are given on the basis of emotional responses to images of people, animals, or forests that the charity is helping. Effective altruism seeks to change that by providing incentives for charities to demonstrate their effectiveness. Already the movement is directing tens of millions of dollars to charities that are effectively reducing the suffering and death caused by extreme poverty.

He claims that GiveWell has gone a long way toward telling us what we need to know in order to give effectively:

The quality and availability of research on the effectiveness of individual charities has risen dramatically over the past few years, largely due to the existence of GiveWell, a research organization set up in 2007 precisely to fill the vacuum that existed previously.

This turns out to be markedly over-confident. Later that same month, in July 2015, The Guardian published a story entitled, “New Research Debunks Merits of Global Deworming Programmes,” which broke the bad news thus:

Deworming children, once ranked by Nobel laureates as the fourth most effective intervention to solve the health problems of the whole world, offers very little benefit despite the millions of dollars spent on it, according to a re-analysis of evidence.

Much of the previous optimism had rested on that one 2004 study, which was carried out in Kenya, by Edward Miguel and Michael Kremer. Unfortunately, the above calculation errors rendered many of its key results wrong, and other researchers also “found there was missing data and some evidence of bias in the finding that school attendance improved. And there was no effect—as had been claimed—on exam results.”

Moreover, as Michael Hobbes says in the aforementioned episode of Maintenance Phase, once the calculating errors were corrected:

All of the health effects of deworming disappear. The anemia link is gone. The kids’ difference in heights? Gone. Difference in weights? Gone.

Finally, the largest available meta-analysis of deworming programs, published in 2015, concluded that:

For routine deworming of school children in endemic areas, there is quite substantial evidence that deworming programmes do not show benefit in terms of average nutritional status, haemoglobin, cognition, school performance, or death.

Millions of dollars, partly in charitable donations, may hence have been wasted. Oopsie.

But that’s not all. It would be one thing if GiveWell had admitted its mistakes (which, to be fair, it has in other cases), and Singer had used his enormous platform to course-correct. But the opposite has happened. Since then, GiveWell has doubled down on its advice—with deworming programs still constituting 4 of its 9 top recommended charities—even though it admits the following on its website:

Mass deworming is generally very inexpensive on a per-person-treated basis (in the range of $0.50). The benefits are potentially major, but also debatable: parasitic infections rarely cause mortality or other acute effects, and the evidence on their impact on quality of life is thin. In brief:

There is strong evidence that administration of the drugs reduces worm loads, but weaker evidence on the causal relationship between reducing worm loads and improved life outcomes.

Evidence for the impact of deworming on short-term general health is thin. We would guess that deworming has small impacts on weight, but the evidence for its impact on other health outcomes is weak.

This is difficult to reconcile with their advice to keep on giving. In the end, they seem to lean on the low cost of treatment offsetting the low probability of improved outcomes, together with one ten-year study that the hosts of Maintenance Phase note doesn’t inspire much confidence.

And Singer has not, as far as I can tell, acknowledged the controversy at all. In the 10th anniversary edition of his book, The Life You Can Save, which was published in 2019, he writes:

There are many more examples of how a relatively small donation can do a lot of good. If you are considering donating to a charity recommended by The Life You Can Save, you can use the organization’s Impact Calculator to show what the amount you donate will achieve. On current estimates, a $50 donation could: 1. Deliver treatments through the Schistosomiasis Control Initiative or Evidence Action’s Deworm the World program to protect an estimated 100 or more children from parasitic worm infections.[1]

The citation points, as usual, to GiveWell’s endorsement of Evidence Action’s “Deworm the World” initiative, itself set up by one of the authors of the debunked 2004 paper, Michael Kremer, among others. Evidence Action notes that “The Life You Can Save, founded by leading moral philosopher Peter Singer, features Evidence Action on its list of 21 best charities in the world.” Its boast? “We measurably reduce the burden of poverty for hundreds of millions of people and we have the data to prove it.” It continues to list deworming as one of its flagship programs on its website.

And so, we come to this troubling situation. A large philosophical community, known as the Effective Altruist Movement, led by Singer among others, has been essentially guilting and shaming people for giving money to non-GiveWell-approved charities, without seeming to ensure that its advice is still valid.

I can’t help but reflect grimly on the fact that, like philosophy in general, the effective altruism movement is worryingly white male-dominated. And there’s a lesson to be learned here about the dangers of a certain masculine technocratic arrogance and white savior complex and their intersection. (As I quipped on Twitter, the Venn diagram of such traits is troublingly close to a circle.) 

There’s also a worry I have about the images and stereotypes that inform such people’s sense of which charities are worthwhile. Gordon and Hobbes in their episode mention the desire for a silver bullet to solve the problems in the developing world—the more Ted Talk-ishly counter-intuitive the supposed solution, the better. I also wonder about the extent to which deworming in particular has caught the imagination of well-meaning but implicitly racist whites who all too easily envision people in the developing world as “infected” or “contaminated,” given the power of racist disgust in moral psychology. (In reality, as Gordon and Hobbes explain, the light to moderate worm infections that many people have in some regions are largely asymptomatic.)

Finally, I continue to find in Singer a troubling common thread that I would characterize as extractive. He wants you to give stuff up—rightly or, sometimes wrongly. He wants you to give up your money (fine for some), your meat (ditto), your attachment to particular charities (hmm), and, most troublingly and distressingly of all, your attachment to a sense of severely disabled people’s value and humanity (unforgivable, in my opinion). Sometimes our human, emotional attachment to these things is in good order and not irrational. Indeed, it may represent the deepest and best parts of ourselves.

And even when we consider Singer’s apparently most ethically promising arguments about our charitable obligations, swooping in in this way—both in the developing world, and to redirect people’s charitable donations more rationally—is not always productive or compassionate. When it comes to doing good, we must first do no violence. This includes epistemic violence of the kind that Singer arguably does when he declines to work with people in the developing world on what they want or need, as opposed to what he and his white brethren think they must need, surely. In an interview with Daniel A. Gross in The New Yorker last year, Singer was asked about his tendency to cite mostly white male thinkers from other elite institutions who “take up a lot of space in intellectual conversations” already. Singer replied:  

I want to work with people whose ideas are, you know, at a level of discussion that I’m interested in, and that I’m progressing. If you’re thinking of the work of Africans, for example, I don’t know the work of many of them that is really in the same sort of—I’m not quite sure how to put this—participating in the same discussion as the people you’ve just mentioned [e.g., white male academics like John Rawls and Michael Sandel, among others].

The lack of moral and epistemic humility here is breathtaking. But this is not, as we might have hoped, a mere aberration. Indeed these problems, as we have seen, may be endemic to Singer’s thinking.

P.S. Peter Singer’s egregious fatphobia is also a whole thing. I’ll have a lot more to say about it in my book on this topic.

---

[1] Aside from addressing anemia, Singer also claims that deworming can prevent “life-threatening conditions” including bladder cancer, kidney malfunction, and spleen damage. His only citation is the GiveWell site, which notes that schistosomiasis can cause “urinary tract infections and other bladder problems at late stages of the disease, potentially leading to bladder cancer and kidney failure; bloody diarrhea, bloody stools, abdominal pain, and liver failure; death.” But, they continue, “death due to schistosomiasis is quite rare; we have no good quantification of the contribution of schistosomiasis to non-fatal problems along these lines.” Moreover, “We have only identified one review (referred to here as “Van der Werf et al.”) that systematically attempts to attribute severe symptoms (organ damage/malfunction, death) to schistosomiasis. This review discusses symptoms such as the presence of abnormal amounts of blood in the urine, self-reported painful urination, enlarged liver/spleen, and various signs of kidney stress or malfunction. However, this review is problematic for several reasons,” including: “We find its methodology for attributing mortality/morbidity to schistosomiasis to be highly problematic,” and “Nearly all of the schistosomiasis consequences discussed in Van der Werf et al. (a) can sometimes be attributable to factors other than schistosomiasis; (b) range in severity and do not necessarily have any detectable impact on quality of life.” So the idea that effective altruists should donate money to deworming programs for the reasons listed by Singer above seems markedly under-evidenced or at least under-cited.