It’s a familiar story, with a twist. On your way to work, you pass a shallow pond. In this pond, a toddler has fallen face down and is struggling to become upright. But unlike other times you’ve passed this pond before, on this occasion you cannot save the child by wading in and muddying your expensive shoes. In a Faustian turn, to save this child you must slice off the arm of a passerby with your pocket knife and extend it to the child, allowing them to grab hold of the bleeding stump so that you can drag them to the shoreline. This is what Larry Temkin worries giving to global health charities is like.
Being Good in a World of Need is a soul-searching tour of the messy ethics and facts of global aid. In its 350 pages, Temkin amasses numerous considerations relevant to the practical ethics of giving aid, making a vigorous case for two central conclusions. First, we need to seriously consider a minority view in development economics (but with some esteemed defenders like Nobel Laureate Angus Deaton): that global aid has done more harm than good by weakening the institutions protecting the global poor and facilitating corruption. And second, Temkin argues, global aid researchers must avoid narrow metrics of cost-effectiveness analysis and employ a morally pluralistic approach, a lesson that Temkin thinks students of effective altruism and donors to GiveWell, in particular, need to take on board. These conclusions are interrelated: as the spillover effects of global aid multiply, the moral considerations we need to consider likewise proliferate.
Empirically, Temkin raises concerns about corruption internal to aid agencies, brain drain from them and how their intentional and accidental actions undermine local culture. He also reconstructs a version of Deaton’s argument: that global aid introduces a “resource curse” where governments can forego taxing their citizens, in turn weakening citizens’ bargaining power with their government and weakening overall governance. Some of the considerations are novel and important: “internal brain drain” from poor countries to aid agencies is not widely discussed and merits additional analysis.
The empirical chapters (4–11) exhibit considerable humility and are accessibly written. Temkin is careful not to overstate the implications of his arguments and shies away from analyzing empirical data for greater readability. A downside of the approach is that it makes it difficult to infer how widespread these problems are or their significance.
If the problems are not ubiquitous, one natural conclusion is to simply avoid supporting bad aid initiatives and support good initiatives instead. Temkin considers the possibility that his concerns are really an argument for effective altruism, whose (quasi-)empirical core idea is that some ways of benefiting others do vastly more good than others and so we should use evidence and reason to identify them and put them into practice. He engages several individual effective altruists at length in chapters 12 and 13. These effective altruists point out that while most aid might well be ineffective, GiveWell’s top charities are buttressed by dozens of RCTs from different countries demonstrating their effectiveness, well-validated mechanisms of action and detailed follow-up inspections to ensure they’re working properly. If this book had been written two years later, they might have further added that the Against Malaria Foundation, one of GiveWell’s top recommended charities, has saved an estimated 150,000 lives as of March 2023. But Temkin appears unconvinced. He thinks effective altruists may be paying insufficient attention to extremely subtle spillover effects.
Consider the 200,000 donors who have given to the Against Malaria Foundation. Each donor can look at GiveWell’s RCTs and confidently conclude that they’ve saved a life. And they can look at follow-up reports and be confident that their donation has not had ill effects. This is consistent, Temkin argues, with the conclusion that all 200,000 donors together have had a substantial negative effect much bigger than the positive effect that each one had individually. Perhaps they together weakened the governance in these countries. Or even if the Against Malaria Foundation hasn’t caused directly any negative effects, they have at least supported indirectly the idea of global aid by making a global aid organization a bit more successful, thereby indirectly supporting all of the bad global aid in the world.
Here it is difficult for the reader to determine whether Temkin is fretting about inscrutable butterfly effects or pointing to serious risks. Temkin, his readers and effective altruists will have to fall back on their background beliefs about how much harm and benefit is caused by global aid. For non-experts, epistemically modest background beliefs would anchor on the mean view of development economists. But Temkin (following Deaton) freely admits that most development economists believe that the effect of foreign aid as a whole has been weak but positive, on account of its effects on growth. So why shouldn’t Temkin’s reader assume the same? And absent some specific evidence about a GiveWell charity, why not infer that charities comparatively more effective on narrow metrics have had an even greater positive effect?
But Temkin is no utilitarian. So his criticism of global aid may have less to do with a belief that it is on net harmful and more to do with the pluralistic approach to giving he endorses. Temkin raises a range of ethical considerations relevant to global aid (Chapters 3, 12–15), covering aggregation, virtue ethics, complicity, equality and more. He claims that the ethics of global aid are far too complex to reduce to the utilitarian calculus of maximizing expected welfare benefits.
If the book is meant as a call to vigilance, a reminder of the ways that helping can hurt and a criticism of the many aid efforts that continue to wound, then it is a welcome contribution. History is littered with attempts to do good that have backfired from hapless engineering failures to heinous crimes of empire wrapped in the garments of British utilitarianism. Certainly economists should continue to argue about the effects of global aid in the academy.
But at times Temkin seems to be defending an entirely different alternative to GiveWell. This comes out in Temkin’s extended dialogues with effective altruists and in his concluding recommendations that readers support basic science research and international policy reform. This is a different ambition needing more work.
An alternative thesis we should consider is this: perhaps in a world that is so interconnected and so unequal, where half a billion people live on less than $2 per day and 200,000 people die each year from preventable neglected tropical diseases, the primary thing that matters really is how much utilitarian benefit people get from aid, while other moral considerations are minor by comparison. When the median American’s salary could save ten lives if gifted to GiveWell, perhaps maximizing welfare really is the correct shorthand.
Temkin deploys thought experiments like the one mentioned at the outset meant to challenge conventional utilitarian reasoning: could you justifiably hack off a passerby’s arm to save a drowning child? In ordinary cases of giving, though, we typically don’t use bystanders as a means and the harm is not so consequential. So the relevance of such evocative thought experiments to global aid is questionable.
Still, Temkin raises numerous other ethical considerations that count against, or complicate, utilitarian reasoning when helping others. He argues that helping others is more important when we are face to face with an identified victim than when we are helping an unidentified statistical victim. He urges us to consider whether the victim is innocent or responsible for their plight; whether we are respecting the victim’s customs, way of life and autonomy; whether we are treating them as an equal; and whether we have reasons of partiality to instead prioritize people in our own community.
But in emergency situations, we unavoidably have to weigh up costs and benefits. And sometimes non-welfarist considerations are minor in light of the benefit we can confer on others. The more globalization increases and the more inequality grows, the more well-off people can do to help badly-off people. Everyone agrees that at some point, if the benefits are sufficiently large, the welfare benefits of helping others dominate pluralistic concerns around intentions and personal relationships.
From the starting point that the Against Malaria Foundation has saved a population the size of Charleston, South Carolina – and that each $5,000 donation will mean that one more person gets to live long and well on our unlikely planet – objecting to the GiveWell donor has a high burden of proof. What comparable moral harm could justify letting those 150,000 people die? I’m not sure that Temkin convincingly argues that there could be grounds for thinking that we ought to.
