Inequality, Artificial Intelligence, Technological Disruption, and Assortative Mating

Paul Krugman is an economist and writer, who currently serves as professor of economics and international affairs at Princeton University, Centenary Professor at the London School of Economics, and as an op-ed columnist for The New York Times. Prior to his appointment at Princeton, Krugman served on the faculty of MIT; his last post was Ford International Professor of Economics. He has also taught at Yale and Stanford Universities, and prior to that he was the senior international economist for the President's Council of Economic Advisers, under Ronald Reagan. He is a Fellow of the Econometric Society, a Research Associate of the National Bureau of Economic Research, and a member of the Group of Thirty. He has served as a consultant to the Federal Reserve Bank of New York, the World Bank, the International Monetary Fund, the United Nations, as well as to a number of countries including Portugal and the Philippines.


In December 2008, Mr. Krugman received the Nobel Memorial Prize in Economic Sciences for 2008, honoring his work in international trade patterns.   


Fall 2019

JPPE: With US income and wealth inequality at a historical high, economists like Daron Acemoglu and David Autor have discussed the issue of job polarization and the idea that artificial intelligence and other modern labor-saving innovations might contribute to the widening of that skills gap and the further privileging of high skill work. Are you concerned that modern technology will make inequality worse?


Krugman: I’m concerned but I’m not convinced. The belief that we’re living in an era of radical technological change has a problem, which is, if we were in such a period, we should see rapidly rising productivity. What we’re actually seeing is rather sluggish productivity. Rising productivity is just not being shown in the data. And then once you adopt that attitude you can ask yourself how—thinking about the kind of tangible technological innovations of our time—are we really seeing radical progress? The rise of the original smartphone or the iPhone was a really big deal. How excited are people about this year’s latest smartphone? You really can convince yourself that we’re starting to plateau. And that may not last, but it’s not clear that this is a time of very radical technological change. 


Aside from the fact that rapid technological change isn’t so obvious, the argument that technology is driving income polarization runs up against several problems. I think Autor does great stuff, and that “U shape” he finds is really interesting. But there is a problem if wage developments don’t seem to be following the kinds of labor that he says are being devalued—i.e. if middle-skill work isn’t experiencing worse wage gains than lower-skill work, which is the part that's growing. So if we’re seeing an economy that is polarizing with a greater number of low skill jobs, why are home health aids not getting better paid? Those are service sector jobs, so that makes you question whether there is some statistical artifact about the whole thing. It’s not for sure, but I’m unconvinced. 


And then there’s the general point that if we have technology that’s biased against labor, it needs to be biased towards something, which would be capital. This means returns to investments would be high, but the corporate sector is behaving as if returns on investment are low. They are not investing heavily despite extremely low interest rates. So I just think the whole thing is a story you can tell, and it might be true in the future but there really is no slam dunk evidence that it’s what is happening now. 


JPPE: Research by Robert Allen on the “Engels’ Pause” shows that because technological disruption tends to improve productivity, it also temporarily increases inequality as wages stagnate and returns to capital rise. Then eventually some leveling force brings it down. Do you think that that’s a fair way of looking at how theoretical technological disruption causes inequality?


Krugman: It can happen. To the extent that we have a theoretical analysis of what technology does, that analysis says that it depends on the technology and it depends on the bias of the technology. Technology that replaces a worker with lots of extra capital should have a negative impact on wages and increase inequality. 


That’s not a particularly new insight. David Ricardo had it in 1821, and the reason he had it is because there’s a pretty good case that that’s what happened during the early phases of the Industrial Revolution in Britain. There’s an endless debate about what happened to real wages between 1800 and 1840, but the fact that we’re even having that debate tells you that there isn’t sufficiently convincing evidence of rising real wages to override the counterarguments. 


So stagnating wages due to technology is possible. It’s not clear that it has happened again since the Industrial Revolution. There is an argument that there was a kind of technological bias towards highly educated workers, which was driving the rise in income inequality in the 1980s and 1990s. That’s more debatable, but it’s also a story that doesn’t help much in developments since 2000. 


So technology can have an effect, and it’s very easy to write down a model in which technological change is, for some period—and maybe even an extended period—, bad for substantial groups of workers. But it depends on the story you tell. 


JPPE: There was economic research that found assortative mating was responsible for twenty percent of the rise in inequality since the 1980s. Is there anything college students can do about this, or are they just the vehicles of widening inequality? 


Krugman: It's not just assortative mating; it’s assortative lots-of-stuff. At the highest levels, everyone was roommates at Harvard. But I think a lot of those assortative mating things are mostly relying on inequality as measured by survey data, which doesn’t capture the really huge incomes at the top. Those incomes are measured by other things, and that’s a large part of the inequality. But, look, if we can restore adequate funding for high-quality public education so that we can have more great students at a wider variety of places, then maybe the mating won’t be so assortative. I’m not big on the notion that any intervention in people’s lives is evil socialism, but telling people who fall in love with is beyond even what I would consider. 


JPPE: Fair enough. Walter Scheidel came out with The Great Leveler where he wrote a history of inequality. His thesis was that periods of high inequality only ever get remedied by mass military mobilization, plague, civil war, or government collapse. In a time of historically high inequality, are you worried about that? Or do you think that effective policy and effective politics can actually play a role in reducing inequality?


Krugman: The middle-class society that I grew up in—now gone— was the creation of policy. It was not the result of the invisible hand of the market, but a dramatic increase in unionization, the squeezing of wages, wage differentials, the establishment of norms, and changes in taxation, all of which were associated with World War II. So massive total war was the background for the Great Compression. 


Do we know that this is the only way that reducing inequality can happen? No. It’s the only way we’ve seen it happen in the past, but we don’t have a whole lot of samples, and you have to hope that we can do it differently. And I would say that there were significant equalizing reforms during the Progressive Era, and it’s true that some of the stuff took place after World War I, but some of it took place before. So I don't think history should give you total pessimism about our ability to enact change. And I'd like to see more equality and not total war.