Cardiff Garcia of the Economic Innovation Group interviews Paul Krugman at The New Bazaar website (October 9, 2024). The interview has a number of points of interest. Here are two themes that caught my eye.
The study of “economic geography” focuses on why economics activity may tend to cluster, or to spread out, or to happen in certain locations. The general theory is that there are “agglomeration economies,” which refers to the idea that clusters of certain kinds of firms, workers, and suppliers benefit from being close together, perhaps with clusters of consumers too. But on the other side, clusters of economic activity can also lead to congestion, crime, even disease. Thus, there is a push and pull of the clustering of economic activity over time. In the last few decades, should improved information and communications technology lead economic activity to spread out, or to become more centralized? Krugman notes:
I mean the idea that agglomeration economies of one form or another exist is obviously not new … There’s forces pulling things together and forces pulling them apart. And the balance can tip one way or another. In fact, over the past 200 years, it has tipped first one way, then the other, then back again.
And the modern forms of agglomeration are different from the ones that prevailed in the 19th century. I like to say that the models that I was writing down 30 years ago had this kind of steampunk feel. They all kind of were very much focused on manufacturing and on industrial clusters, and we all lavished attention on these great stories — like the detachable collar and cuff industry of Troy, New York, and that sort of thing. And which mostly have gone away in the United States, although not totally — they do exist to some extent even in manufacturing, but these days if you really want to find old style industrial clusters, you go to China.
So if you actually look at — there’s a variety of measures — but basically, in the United States, there was a lot of regional convergence, convergence in incomes, convergence in basically regions becoming more similar, from the 1920s up until about 1980. Then in 1980, they started pulling apart again, and you started to see metropolitan areas with highly educated workforces pulling in even more educated people, pulling in even more of the information economy — and stranding regions that didn’t have those preconditions. … [F]for prep for a future conference, I’ve been looking at just kind of within New York State, and looking at greater Buffalo versus greater New York City. Buffalo was not all that poor or backward compared to New York City in 1980, and now it’s vastly poorer. …
So we’re back in a world where we have these extremely localized clusters. And modern communication technologies, Zoom, work from home, they make it possible to do some things without being in the place, but they’re not a full substitute. And in a rich society other things tend to matter. Particularly for high skill, high pay workers, you need amenities. High tech workers are not going to move to someplace in the middle of the country, even if they have excellent internet access, because where are the good restaurants? Where are the live concerts? So in some ways the fact that we’re rich enough that people can make decisions on that basis matter. So I think we’re in some ways back to the kind of unequalizing development that we had in the late 19th century
While it’s agreed (among economists) that productivity growth is essential to growth of the standard of living, the exact recipe for sustained productivity growth remains elusive. Indeed, standard calculations of productivity are carried out by trying to measure what part of overall output growth can be accounted for by changes in hours worked, skill level of workers, and capital investment–and then “productivity” is measured as the leftover or residual amount of output growth that could not be explained by the other factors. Because the economist Robert Solow did some of the classic work in this area, the productivity residual is sometime called the “Solow residual.” Here’s Krugman on the mystery that is productivity growth:
But if you look at a chart of U.S. potential GDP growth over the past 50 years — we had some pretty big political swings in there, big changes in tax policy. If you didn’t know that there have been changes of administration, changes in tax policy, you would never guess. It’s pretty much just a flat line. The reason is that economic growth is largely driven by the Solow [productivity] Residual. And who knows how to make that change very much.
If someone says “I have a policy that could raise potential GDP growth by a quarter of a percentage point,” I’d say, “okay, I could possibly believe that, show me the details.” If somebody says they could raise it by one percentage point, I think that’s crazy. Nobody knows how to do that.
My favorite Bob Solow quote, although there are so many of his … he was actually talking about Britain lagging behind post World War II, but it applies to lots of things, he said, “every attempt to explain this ends up in a blaze of amateur sociology.” That, in the end, we really just don’t know very much about why countries have different rates of economic growth. We still don’t know why productivity slowed down in the early ’70s. We kind of know why it had a bump for a while around IT [information technology], mid ’90s to mid 2000s. But ultimately, talking about innovations, there’s a big mystery now. It feels like we’ve had a lot of technological change these past 16, 17 years, and yet, if we believe our numbers, which maybe we shouldn’t, but if we believe our numbers, total factor productivity growth has been really pretty lousy for that whole period.
