Daron Acemoglu: prediction

Transcript context

…I’m sure you know the Mankiw, Weil, and Romer paper. They have a Solow model, but they put human capital in, and human capital has a very high degree of explanatory power. Presumably, you disagree with them, like in the context of their work — a Solow model, a growth model. What exactly do you specify differently than Mankiw, Weil, and Romer? Well, I think Mankiw, Romer, and Weil is a very important paper for its time, but it also has a lot of problems. The most major problem of Mankiw, Romer, and Weil is that it’s a regression framework of cross-country differences or changes that really has no way of dealing with omitted variable bias. And when we have seen that in the context of Mankiw, Romer, and Weil, if we go a little bit into the details, is that if you compute the implied rate of returns to human capital, they are about four or five times as large as what you would get from Mincer regression. So the Mincer regression says if you get one more year of schooling, you’re going to increase your earnings by about 6 percent in the US, perhaps as high as 8 or 9 percent in some developing countries. Let’s say 10 percent. Let’s be generous. That means that if you are going to increase your rate of schooling by 3 or 4 percent, and if there are no major human capital externalities — and we can come back to that — then that’s going to increase your GDP per capita by 25, 30 percent. That’s a very small number relative to the gaps that we actually observe. So if you look at what — The cross-human externalities are incredible, right?…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.