Paul Romer: evaluation

Transcript context

…You’ve been interacting a lot with epidemiologists due, in part, to your arguments for testing. What’s your opinion of that field? There’s actually an interesting parallel in epidemiology with a technical kind of issue in economics. In macro, we shifted towards model-based reasoning about macroeconomics. So representative agent — the whole rational-expectations movement was a shift towards “Let’s see what the models say,” rather than “Let’s see what the data say.” In epidemiology, there’s a very well-established model — this SIR model that is behind a lot of these predictions, but there’s an alternative — the Institute for Health Metrics and Evaluation, which has this model that’s been very influential, widely watched these days. The IHME is using a much more data-driven approach, kind of a curve-fitting approach. It’s almost like old-style Keynesian macro, where you just say, “Well, let’s just fit something to the numbers and see what comes out of that without imposing a lot of theory onto the estimation process.” And I just found it interesting to see that tension in another field, from the outside. The way it looks to me is, it’s good to have both of those wings active in a discipline. And it’s good to have them in contention with each other. If I have a criticism of macro in economics, it’s like the criticism in epidemiology — we may have biased things a little bit too much towards the models, and we’re not giving enough weight to just the facts themselves. And I think that’s because it’s actually easier to do models than to look at data, so we need to have a little bit of collective pressure to, “Yeah, yeah, that’s what your theory says, but let’s look at the numbers. ” Many people have supported mass testing plans. Of course, you’ve been in the lead here. Why do you think they’re not getting more support? Because the benefit-cost ratio, if you can pull it off, seems to be quite high.…

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