Most people’s attention amid the USAID spending freeze has rightly been focused on the immediate life-and-death matters of health and humanitarian assistance. Matt Yglesias suggests that in the future, US aid should focus more narrowly on these kinds of proven life-saving interventions, and less on projects that Samantha Power describes as “less visible but pay dividends in the longer term, such as giving girls a chance to get an education.”
Matt writes that “it’s very unclear whether the United States is actually effective at this … they don’t make a ton of sense on a dollar-for-dollar basis.” Matt is often right, but I think he’s wrong on this one. There is good evidence that USAID education spending has been effective, and likely extremely good value for money—just as Samantha Power says, over the longer term.
Overall, education is a small slice of all USAID spending—around 3 percent of the $40 billion in disbursements recorded on foreignassistance.gov for 2024. Within education, most USAID programs focus on early grade reading and mathematics skills. These skills are the foundations for the rest of education—if you can’t read you can’t keep up as the curriculum progresses. The US has helped lead a global movement here focusing attention and action on the millions of children who are in school but just not learning very much. These programs are highly leveraged, as most education spending comes from domestic governments. The small contribution of international aid just makes local spending more effective.
Is the US any good at running education programs? A recent paper by my CGD colleagues Justin Sandefur, Thomaz Alvares de Azevedo, Xiaomin Ju, and Thi Le review the entire set of almost 50 programs run in the last 15 years. They estimate that on average USAID reading programs have improved outcomes by 0.3 standard deviations. What does that mean in real terms? The standard presentation in the sector has been to translate that into a number of “equivalent years of schooling.” Or “learning-adjusted” years of schooling.
To get a dollar-for-dollar comparison with health investments requires a dollar costing of those educational benefits though. The challenge here is that whilst we have copious estimates of the dollar returns to more schooling (over 700 in a recent review), we have very few estimates of the returns to more learning. The main reason for that is to estimate the returns to schooling you can just ask adults how much school they completed. You can’t ask adults how good their third grade reading skills were, you need to measure them 20 years ago when they were children. As far as I know, there is only one published study, in China, that does this. In a forthcoming paper, I do something similar for Indonesia. Both papers show similar results; that a 0.3 standard deviation improvement in a child’s learning is associated with a 3 to 4 percent increase in annual earnings for that individual as a young adult. With this dollar return to learning estimate in hand, we can value the return on USAID’s investment in learning.
GDP per capita (and so average earnings) in the countries that USAID ran education programs in was around $2,500 per year. If a person’s earnings increase by 3 percent, we can estimate the total benefit over their lifetime. Assuming someone works for 40 years, with their salary growing by 2 percent each year (in line with GDP growth), with an annual discount rate of 8 percent, the net present value of lifetime benefits amounts to over $1,000. To get that, USAID spent an average of $34 per child, implying a benefit-cost ratio on the order of 30-1. That’s great value. And that’s even without valuing any of the non-monetary benefits, the intergenerational benefits, and the social benefits. Of course, you can argue with the parameters here. Perhaps 0.3 standard deviations is too generous, or a 3 percent annual earning gain is too high. But the underlying logic is clear.
A little extra learning leads to a little extra earning, which adds up over a lifetime, and the cost is minimal. These gains, small and spread out, aren’t going to make a ton of headlines. But they are still a great use of money, and make plenty of sense on a dollar-for-dollar basis.
The US and other donors should continue to scrutinise their aid portfolio to minimise waste and inefficiency, but there is a solid track record on global education to build on.
CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise.
CGD is a nonpartisan, independent organization and does not take institutional positions.
