Jeffrey Sachs, the author of The End of Poverty, deserves a lot of credit. He was one of the first people to realize that fighting malaria was important for malaria-affected countries’ economic development. He helped found the Global Fund to Fight AIDS, TB, and Malaria, which does a lot of good work. He has certainly done more for the global poor than I have.
Which is why The End of Poverty is so interesting. Jeffrey Sachs is perhaps one of the most prominent advocates of massively expanding how much money the developed world spends on foreign aid. The End of Poverty is his booklength defense of this position, which means it’s a useful challenge for those of us who tend to be aid skeptics. And because The End of Poverty came out in 2006, we can see how well its predictions hold up sixteen years later—an angle you can’t look at for more recent books.
Unfortunately for me, a person who would quite like to be less aid-skeptical than I actually am, The End of Poverty is a terrible, terrible book.
Part of the way through The End of Poverty, you read the paragraph:
One reason for optimism, in general, is that as economic development takes place, the call for democratization and transparency grows. That phenomenon, although not universal, is very widespread, and it will definitely work in China. I say definitely because we have seen how powerfully that impulse succeeded in Taiwan, South Korea, and in other nearby cultures and countries. China will experience powerful forces for democratization from within as the rate of literacy and the level of private wealth grow, and as various interest groups in the society have more standing and greater eagerness to participate politically, in part to defend their property rights.
Needless to say, it didn’t work out like that.
I looked for Jeffrey Sachs writing a mea culpa and analyzing how he got this so wrong, but all I found is that he has decided to take up Uyghur genocide denialism in his spare time. So it goes.
To some extent, the End of Poverty is Sachs’s memoir. Sachs has done a lot of interesting things, and those chapters were interesting. Unfortunately, he still works in global development and needs to maintain relationships. Therefore, everyone who shows up in the book is an intelligent, hardworking, competent person sincerely motivated by the desire to help the global poor. He should have stuck to facts in this one and published his memoir when he retired and was ready to settle scores.
But even if it were just facts, I would have found The End of Poverty unsatisfying. It has a lot of bad reasoning and flat statements of controversial claims without any sort of backing evidence. For example, we are informed that “culture does not explain the persistence of poverty in Bolivia, Ethiopia, Kyrgyzstan, or Tibet.” That could be right! But it would take an entire book to prove. He just tosses the statement off off-handedly and moves on.
As another example, Sachs refers to “governance scores” as evidence that African countries aren’t especially poorly governed—but he doesn’t explain how poor governance is defined or operationalized or how these scores are calculated. We’re just supposed to take his word for it, I guess.
He misrepresents evidence to support his own views. For example:
The PIPA survey also found that 54 percent [of Americans] rejected the idea that foreign aid "should be strictly a private matter taken care of by individuals giving donations through private organizations." Americans understand what must be done and why it is a public duty.
It seems to me that if 54% of Americans think something, then Americans are quite divided, actually? If Sachs represents data that is literally in the previous sentence that badly, how can we expect him to accurately represent more complex evidence?
Sachs persistently confuses correlation with causality. Fertility rate is correlated with GDP per capita, so Sachs concludes that having too many babies causes lower GDP. Poverty is correlated with poor governance, and so poverty causes poor governance. (On priors, I’m skeptical that the effect of wealth increasing bureaucrats’ pay outweighs the effect of corruption making countries poorer.) Democracy is correlated with wealth, so wealth causes democracy. This is an embarrassing mistake for an economist to make.
Speaking of embarrassing mistakes for an economist to make:
The rest [of the aid], $12 [per capita], went to Africa. Is it really a surprise that we do not see many traces of that aid on the ground? If we want to see the impact of aid, we had better offer enough to produce results.
There are two problems with these sentences. First, $12 is really quite a lot of money compared to the average African’s income, so we’d expect to see some effect. Second, money… normally has diminishing marginal returns…? So we would expect the first dollar to have a larger effect, probably one we could see in the data? I feel embarrassed making this argument to a tenured professor of economics. It feels like pointing out to a biologist that young animals are normally similar to their parents.
Perhaps because Sachs feels no particular need to justify his claims with adequate evidence, The End of Poverty is extraordinarily over-optimistic. Sachs thinks there should be a science of “clinical economics,” similar to clinical medicine. He writes down an entire differential diagnosis process he thinks economists should use to figure out what is wrong with a country’s economy. At no point does it occur to him that, if clinical economics is a field, it’s about where clinical psychiatry was in the 1930s. Sometimes the patient has a problem that the clinical economist can help with. But mostly clinical economists have a choice between giving the patient malaria or put them in a coma and see if it helps, and offering them sympathy and waiting for the problem to get better on its own.
Because of his faith in clinical economics, Sachs thinks that developed-world donors can do… pretty much everything. Corruption in Kenya? Clearly, developed-world donors should call in McKinsey and get McKinsey to design Kenya a system that won’t be corrupt. It never occurs to Sachs that Kenyan governmental malfeasance might be a problem best solved by Kenyans, or that corruption in the Kenyan government is a problem almost uniquely hard to solve by giving money to the Kenyan government.
Perhaps relatedly, Sachs doesn’t take liberalism or the protection of economic and political rights seriously. The quote about China that I opened this blog post with is only one example. But there are others:
There are early stirrings of democracy in China right now, for example, nonparty elections at the village level, where individuals stand for positions rather than represent organized political parties. Hu Jintao, the general secretary of the Communist Party, recently declared, "Democracy is the common pursuit of mankind, and all countries must earnestly protect the democratic rights of the people." There is a huge gap between words and action, but the gap is very likely to narrow meaningfully in the coming years.
There are two things I notice about this. First, Sachs is extraordinarily optimistic about whether economic growth will naturally cause dictators to give up their power. Second, Sachs is almost entirely focused on democracy. He ignores free speech, freedom of religion, property rights, participation in a free market, and other basic rights which both are valuable in themselves and play an important role in development. Sachs’s section about the need for political reform in China is focused almost entirely on federalism, which I think even in 2005 was Really Not The Biggest Problem Here.
Particularly indicative, I think, is Sachs’s claim that democracies do have famines, contrary to the work of Amartya Sen. He argues that African geography causes even democracies to have famines, which shows that good governance isn’t sufficient. However, the disagreement entirely comes down to your definition of “democracy.” Sachs appears to be including some countries as democracies that most people would not consider to be democracies. In general, researchers have validated Sen’s finding. Sachs doesn’t want to admit the importance of democracy, because that would mean that governance and institutions matter, and you can’t just give governments money and expect the institutions to fix themselves. If liberal democracy is important, his entire approach would be misguided.
When reduced to its essentials, Sachs’s plan to solve global poverty is that each country can identify how it isn’t meeting the Millennium Development Goals, then create a plan to fix it, then receive money for this plan, and then the problem will be solved. While in principle developing-world governments can make their own plans, the plans should reflect the wisdom of clinical economists, and of course the Millennium Development Goals themselves were heavily influenced by developed countries. Meanwhile, aid money should be made contingent on good governance; corrupt and authoritarian regimes shouldn’t get money. (I’m not sure how this doesn’t result in exclusively giving aid to India and Botswana, but whatever.)
I am not sure why Sachs thinks that the developed world dictating what governments the developing world can or can’t have, on the penalty of significantly withholding aid, will work any better than all the election interference and coups the US government has been doing for decades. Nor is it obvious to me why Sachs thinks that it will work out well for developed-world governments to dictate the goals of developing-world governments and what they should do with a huge percentage of their economy. I have a lot of ethical qualms about providing aid equivalent to a country’s GDP per capita based on what people in the developed world think people in the developing world’s needs are. (Sachs considers but rejects cash transfers on the grounds that they just increase consumption and we need to solve the root causes of poverty.)
I am most sympathetic to Sachs’s viewpoint with regards to malaria, which is his flagship issue. Malaria is obviously bad; there is no reasonable argument that it’s in people’s best interest to have malaria. It is straightforward for the developed world to give people in the developing world insecticide-treated bednets and seasonal malaria chemoprevention. It is only somewhat more difficult to make sure people take their pills and put up their bednets. If the pills are taken and the bednets are put up, people are definitely less likely to have malaria.
But outside of global health, the track record of foreign aid often looks much worse. We don’t necessarily know what people need: is education more important than good roads for making a country richer? Even if we know education is important, it’s hard to know whether (say) better schoolbooks, teacher training, or free lunches will lead to hte best education. Bad governments can easily steal foreign aid, misallocate it in a way that leaves the poor worse off, or even use it as a way to punish political enemies. Some serious problems, like poor governance, are difficult for developed-world donors to do anything about.
Sachs fails to address what I think is the most serious problem with his approach: while the causes of development are complex and hard to understand, we know that no developed country got that way via a richer country giving them a huge amount of foreign aid. It’s possible that this strategy we’ve never tried works better than the ones we’ve tried! But that conclusion requires data that Sachs is unwilling to present.
> Needless to say, it didn’t work out like that.
Tanner Greer has an interesting piece on this here: https://scholars-stage.org/give-no-heed-to-the-walking-dead/
Basically, he says, this was a good idea and, on its own, it did have that effect! It's just that, y'know, the Communist Party of China is not stupid and acted to counter that. And the US completely failed to respond to that. But it's possible that with more follow-through, this could possibly have gone much better?