Philosopher Larry Temkin was once a champion of the moral case for donating to aid NGOs. As I explained in my first blog post, his views have now changed. He explains why in his book, Being Good in a World of Need.

Temkin still supports donations to NGOs in some situations, but he no longer thinks we should donate to NGOs working in the poorest, most aid-dependent countries. This is because he no longer believes we can simply assume NGO work succeeds. He also believes that Effective Altruists are far too confident of their ability to choose the most successful aid projects or NGOs to support. On top of that, influenced by the arguments of economist Angus Deaton, Temkin believes that there is a real risk that aid, including aid from NGOs, makes governance worse in countries where aid flows are high, and that this ultimately impedes development progress.

I have some sympathy for what Temkin has to say (full disclosure: I was sent an earlier draft of the book to comment on). Aid sometimes fails, and Temkin is right to worry whether tools like randomised control trials (RCTs) really allow us to spot the best interventions, or choose the best NGOs. (I’ve discussed the pros and cons of RCTs before.) I’m no naive aid optimist. Helping is harder than it’s sometimes made to seem.

But while NGO work isn’t perfect or guaranteed to succeed, it can be done quite well. Temkin focuses a lot on the worst examples of NGO failures in his book, ostensibly because he is trying to demonstrate that aid can’t be assumed to help, but his book would be better if he’d focused on what NGOs typically achieve.

In my experience studying aid, including NGO aid, many NGOs do some good quite often. Some NGOs work very well most of the time. And, even if we can’t draw on evidence with the certainty that some Effective Altruists claim to possess, it’s possible to use heuristics to pick better NGOs. Do they conduct meaningful evaluations of their work? Do they publish them online? Are they signatories to broader codes of conduct? Do they have enduring partnerships with local organisations? It’s possible to thoughtfully give aid to NGOs, even ones working in poor, poorly governed, aid-dependent countries, and be fairly confident that your donation funds do some good.

Of course, if Angus Deaton is right and aid undermines governance, this could still be the wrong thing to do. We might be able to find good NGOs and donate to them, but if NGO aid undermines the crucial link between citizens and governments, the achievements of individual projects don’t count for much. This is a serious concern. Or it least it would be if Deaton was right. Is he?

I doubt it.

For a start, there’s the question of volume. Temkin provides examples of some countries where inflows of aid from government donors and multilateral organisations are very high compared to the revenues of those governments. Maybe aid distorts the political economies of these countries. Yet the subset of the world’s aid-recipient countries that are actually this aid dependent is small. Averaged over the years from 2015 to 2019, in more than half of all aid-recipient countries, aid from governments and multilateral organisations was less than 10% of the size of total recipient government revenues (see my workings). What’s more, NGO aid is much less than aid from governments and multilaterals. (In Australia, where we have good data, NGO donations are about a quarter the size of government aid.) In most developing countries donation-funded NGO work is too small to seriously skew politics.

It’s also questionable whether Deaton’s argument is relevant to NGO work even in very aid-dependent countries. This is because his argument pertains to money from donor governments to recipient governments, which supposedly spares recipient governments the task of taxing their citizens, thereby – in theory – reducing accountability. NGO aid, however, doesn’t go to governments. It goes to communities or families. It doesn’t free governments from the task of raising tax revenue. Conceivably, perhaps, thanks to NGO work, people might find themselves affluent and complacent, and stop holding their governments to account. But people are comparatively affluent in Australia, and it doesn’t stop them from demanding accountability. Indeed, people who are better off may actually be better able to hold their governments to account in the way Deaton wants (this would fit with the findings of work on clientelism in political science).

What’s more, Deaton’s argument is primarily theoretical. Yet there’s also a rich body of empirical evidence from studies of the impact of aid on governance, which barely gets referred to in Temkin’s book. The evidence comes in many forms, ranging from regressions to experiments to qualitative work. Much pertains to government aid but most is relevant to Temkin’s concerns about NGOs. The findings are nuanced, and suggest different types of aid have different effects, but overall the body of literature does not fit with the belief that aid, particularly NGO aid, is likely to undermine governance. (See, for example, work on aid’s relationship to citizens’ legitimating beliefs and their demands of government, democracy and democratic change, political institutions and trust in them, government spending, regime type in recipient countries, and the long-term effect of NGO provision on political attitudes and behaviour.)

The chance that donations to NGOs, even NGOs working in the world’s most aid-dependent nations, are likely to seriously damage governance in those countries is slim.

In Being Good in a World of Need Temkin takes an interesting journey from philosophy into the often complicated world of aid practice. Some of his conclusions are eminently sensible – particularly the important point that there are other ways we can help above and beyond donating to aid NGOs. Yet his claim that we shouldn’t donate to NGOs in aid-dependent countries is dubious. It would entail abandoning people in acute need, on the basis of very questionable beliefs.

I can’t see how this could possibly be morally desirable.

This is the second blog in a two-part review.