Nathan Nunn: evaluation

Transcript context

…But does this get at a broader question about persistence and the importance of persistence? If you look at Ethiopia in the 1970s, it’s a poster child for extreme poverty. If we look at Ethiopia in the time of the slave trade, Eastern Africa — as you point out in your work, a very large number of slaves were taken from Ethiopia, which is supposed to predict low trust and bad long-term outcomes. And now, they’re good at nation building because of events millennia ago. Over what time horizon does persistence operate? Or is it always supposed to be moving in the same direction? Or is there also some kind of mean reversion? That’s a great point. When you think about persistence, one thing that’s important is how much of the total variation does this factor in the past explain? That is going to be relevant for the type of question that you ask. If we see, in the periods of decades or even centuries, a lot of movement and correlations that are in opposite direction, that’s fine, and that’s completely consistent with persistence, as long as persistence isn’t explaining 99 percent of everything, or 97 percent. The way I think about it — and hopefully this is answering your question — we’re all pretty confident that smoking, in general, causes lung cancer, so that’s kind of a long-term factor. But you can definitely find counterexamples of that, and that’s individuals who maybe smoked but are healthy and lived a lot longer. Or throughout your lifetime, you might have differences in levels of health. At one point you could be healthier or less healthy. Exactly similar argument with education. We think, on average, more educated people have higher levels of income, but throughout a person’s life, there could be ups and downs, and the correlation could be different amongst subsamples. I think that analogy, if you apply it to any country and any country’s experience, goes a long way to understanding why countries are moving often in opposite directions than what we would think. But if you look at a broad cross-section of countries, there is this nonzero correlation that we call persistence. Let’s say you’re in a tournament with a hedge fund manager, and you’re both having to invest in emerging economies, and you understand persistence — you and Melissa Dell probably better than anyone in the world — and the hedge fund manager hasn’t mastered your work. So even if persistence is explaining, say, only 5 percent of the variation, which would be a modest claim, you should be able to beat that person, right? Because you and Melissa understand persistence, and you will pick the winners, on average, somewhat better than the hedge fund manager. Or no?…

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