Bill Kerr: In recent decades, America’s economic growth has been imbalanced. While some coastal cities have surged and prospered, broad stretches of the country are falling farther and farther behind, and traditional measures may miss important parts of the story.

Welcome to the Managing the Future of Work podcast from Harvard Business School. I’m your host, Bill Kerr. I’m joined today by Harvard professor Larry Summers. Larry’s exceptional career has included serving as Treasury Secretary under President Clinton, being president of Harvard University, and most recently directing the National Economic Council under President Obama. Since returning to Harvard, Larry has been researching and writing on how technological change is reshaping the economy and the future of work. He has also worked on the inequalities faced by rural areas and potential policies for relief. Welcome, Larry, and thank you for joining us.

Larry Summers: Glad to be with you.

Kerr: Larry, a lot of the discussion is about the future of work, and you can hear predictions ranging from “blissful utopia” to a jobless misery to something that might look more or less the way it does right now, just with different types of work. When you look out to the future, what do you see?

Summers: First of all, I think no one knows for sure. There are two kinds of people: those who know they don’t know and those who don’t know they don’t know. So all predictions should be made with a lot of humility. Second thing I’d say is that I think that there are some longstanding trends in our economy, and one of the most important of those longstanding trends has been toward less employment of mature men. In the late 1960s, nearly 19 out of 20 mature men—men between the ages of 25 and 54, people economists refer to as “prime age” because they’re mostly too old to be in school and mostly too young to retire, so we kind of assume that they would be working—it used to be that one out of 20 of those people were not working. And today, even though we have 3.7 percent unemployment, it’s more like one in seven of those people are not working.

Kerr: So the unemployment statistic is misleading.

Summers: And that suggests that we’ve just had a lot of people for whom the economy is not finding a job. In the vast majority of cases, they’re not showing up as unemployed. They’re showing up as so-called “not in the labor force”—not looking for a job, but probably that reflects their sense that they can’t get a job of a kind that they want. And that’s a longstanding trend that has been with us now for multiple decades. Why is it happening? Some of it undoubtedly has to do with the restructuring of our economy associated with globalization, but I think a larger part of it has to do with the implications of technological progress that has meant that there is less routine strength-requiring work to do in our economy, and that’s a trend that will only continue. And the likely consequence of the continuation of that trend is going to be more people who are not going to be able to find jobs that pay the kind of wages that they think are appropriate for them to receive. And that’s going to be a substantial challenge for us going forward. Will the people end up working for very low wages? Will the people end up not working and being otherwise supported? Will we find ways of training them and adapting them and adapting the institutions of our society so that they’re able to find higher-paying work in support of our being a more equal society? That’s going to be one of the great economic questions that will play out over the next several decades.

Kerr: This trend has been long term in the United States. Do we see it in other countries?

Summers: We do. In general, there has been a decline in the share of prime-age men who are working. It’s somewhat more pronounced in the United States than it is in other countries, but Americans should ponder the fact that, while we talk about our terrific, dynamic, flexible labor markets, and we talk about the tremendous dynamism of our society, the truth is, in terms of the fraction of people who are not working, the United States is very, very high among mature men, higher than the vast majority of European countries—certainly higher than France, which Americans are accustomed to thinking of as rather sclerotic.

Kerr: And your work has recently talked about strong geographical differences in this non-employment. Tell us a little bit about what you’re studying and the root causes of it.

Summers: It remains to be seen exactly what the root causes are, but if you look, there is a substantial swath of the United States, what you might call the heartland of the country, that runs from the Rust Belt through Appalachia. Think about a swath cutting across the country, from Michigan to West Virginia, that has a particularly concentrated non-employment of men, that has higher fractions of men who are not working, that has some evidence, a fair amount of evidence—larger problems with opiates and the like—and that has larger problems than we’ve had historically of declining life expectancy, social disconnection. That has obviously had very substantial political salience. Those states and that part of the country was decisive in the last presidential election, and the outcome was, of course, surprising to many observers.

Kerr: Yes, but your work is particularly emphasizing that it’s not just about income replacement or being poor; it’s associated with the job itself.

Summers: The evidence is really very strong that people who are working are much happier than people who are not working, and that part of what we find satisfaction in, as human beings, is in a sense that we are making a contribution. And whether one looks at the experiences of the children of the very wealthy who are removed from the obligation to work or one looks at those who have been affected by prolonged economic depressions and have largely abandoned work—with some exceptions, it’s not easy to see communities with high degrees of satisfaction.

Kerr: When I first came into graduate school in economics, the concept of convergence was widely discussed. But you’re saying we can’t rely upon that in this setting.

Summers: Right. One of the things that is also striking is that the general tendency over the long term across US regions has been toward very substantial convergence, has been toward very substantial adjustments of income, toward equalization. And that might not be happening anymore.

Kerr: OK.

Summers: And in the last couple of decades, it has seemed like the areas of the country that have been left behind have not really been having very much, on average, in the way of experiences that have enabled them to catch up. That’s why I have been an advocate of some reconsideration of place-based policies.

Kerr: So tell us about the prescriptions you then have. It’s both related to place and also to the job, or the employment.

Summers: I think what we need to do is encourage work, and the best way to encourage work is with wage subsidies, is with programs that support hiring, and particularly with programs that support hiring in places where the supply of labor is likely to be large relative to the demand for labor. That’s why I’ve emphasized wage subsidies that are geographically variant, so that we can seek to encourage employment in those parts of the country where employment is particularly lagging. And that, of course, has the important virtue that, if you stimulate it in areas where there’s already excess supply, you’re less likely to bid up wages and less likely to be setting off an inflation phenomena.

Kerr: As you think about putting this into place, does it matter whether you’re trying to subsidize at the employer level? Do you have a lot of range of how you can bring this about?

Summers: I think there are a variety of ways of doing it. My own instinct is that we’ve been very successful in the United States with employee subsidies via the earned income tax credit. But if we want the emphasis to be on job creation, I would tend to favor employer subsidies and, in particular, would support approaches based on the idea of an incremental subsidy, so that you’re supported or subsidized for the hiring that you do in addition to your previous level of employment or in addition to 90 percent of your previous level of employment. That way, what’s being subsidized is the new marginal employment, and the employment that would’ve taken place anyway is not being subsidized. And in that way, I think the government can get more bang for its buck. That kind of approach of supporting incremental employment is only really possible through subsidies that are on the employer side.

Kerr: And if you thought about this program overall, is the goal to create a long-term sustainable heartland? Is it a transition type of a program?

Summers: Who knows what’s going to be necessary in the very long run, and one always hopes that any initiative of this kind that one supports will prime the pump in a way that it won’t be necessary for further support. But I think, given the magnitude of the problems in the country’s heartland, if we were to embark as a country on a program of support, one should certainly think of it as an effort that would last at least a generation. It’s not something we’re going to do in three years and undo in three years after that.

Kerr: OK. We hear in the press and in business and even in potential presidential hopefuls a lot of talk about UBI—or universal basic income—and also jobs guarantees. What leads you to favor—or not favor—those types of approaches as alternatives?

Summers: I think UBI is very difficult when you look at the arithmetic. If you want to give everybody in the United States $10,000, you’d have to double every tax we have, and that just seems both politically impossible and—if you think about the 45 percent marginal tax rates that some people face—it’s close to economically impossible as well. I also think that UBI probably sends the wrong message. It sends a message that it’s fine not to work, and I’m not sure that’s a message we want to send to our country’s young people.

Kerr: And your surveys would suggest it wouldn’t restore the happiness …

Summers: … and I don’t think the evidence is that people who are home all the time receiving a UBI would be terribly satisfied. I also think there’d be a lot of taxpayer money for a UBI that would go to places that taxpayers wouldn’t find very attractive for it to go. Students taking a summer off for college—would they be eligible for the UBI? College students who decided they wanted to travel for a year after college—would they be eligible for the UBI? I think we have to be rather careful about providing support disconnected from what it is that people do. I think the job-guarantee concept is a better one. Whether it’s all the way to being a good one, I’m not sure. I think one has to ask, “How are the jobs going to be guaranteed? What are the people going to do? Who is going to supervise the people? What is the consequence if they perform public functions for the government for the people who are performing those functions now?” I’d like to see a fully articulated jobs-guarantee proposal, and I could imagine supporting it, but I haven’t seen that proposal fully fleshed out to date. For now, I think my preference would be to commit macroeconomic policy more strongly to full employment and to use these kinds of weighed subsidy programs that I have been talking to. But I think, particularly, the ideas around a job guarantee are worthy of a lot of future study. What you have to figure out is how to make the job attractive so that it’s a politically and economically meaningful thing for people, without having the effect of drawing a large number of people who have jobs out of their existing job and into this new public alternative. And there may be ways of doing that, but I would want to see the proposal fully articulated.

Kerr: Let’s go back to your technology and its impact on both the jobs and growth. You began by describing sort of the skilled nature of recent technological progress and the non-routine tasks and similar. Do you think that that relationship is changing or continuing in that respect?

Summers: I think there’s every reason to think so. You look at what’s coming out of Silicon Valley. You look at what kinds of things artificial intelligence (AI) is making possible. I would think it was more likely that the impact of technology on jobs was going to increase than that it was going to decrease. I think that, if you look at the emerging market sector, it is a much larger share of the global economy than it was 25 years ago. So a given growth rate for emerging markets translates into a much larger impact on the global economy than it used to, and that’s another reason why I think there’s likely to be more pressure. So my instinct is that these pressures may increase rather than decrease over time.

Kerr: In economics, we have the famous solo paradox. We could see the computers everywhere, but the productivity numbers …. And you’ve talked about how you see today’s growth rate as being ordinary growth but done by extraordinary measures in terms of monetary policy and fiscal policy. Do you anticipate a productivity increase in the future from AI and advanced technologies?

Summers: I would expect, I mean, I think it’s interesting to think about autonomous vehicles. As of this moment, there are probably tens, if not hundreds, of thousands of people who are employed in some way that involves thinking about, planning for, investing in, considering, developing autonomous vehicles. As yet, not a single chauffeur or truck driver has been replaced by an autonomous vehicle. And so, in effect, the implication of autonomous vehicles so far has been to reduce productivity as we measure it. But, ultimately, it will surely translate into a substantial increase in productivity. And I think there’s a fair amount of that going on. Take another example. When the first e-books were introduced, there was an ecosystem of people involved with those e-books, and they were all employed. Initially, there was no increase in the number of books sold, so the fact that you had the same number of total books sold and you had more people working in the book industry meant that there had been a reduction in productivity. Ultimately, the creative destruction took place, and many stores closed. There were many fewer booksellers, so you started to see an increase in productivity. But that’s a process that takes time. And while it’s happening, you may—ironically, from technological change—be seeing reductions in productivity, rather than increases in productivity.

Kerr: So it could be decades long in an adjustment process and hard to measure.

Summers: I think that’s right, and I think it’s surprising, but important, how long that was true and how much that was true of the previous increases in productivity growth.

Kerr: You’ve talked about the inflationary pressures that can come from technological progress, globalization, and jobs, and how that could be changing monetary policy. Can you say just a little bit about that?

Summers: I think you have a generalized deflationary pressure that comes from this innovation, this greater capacity for competition. In addition to that, I think we have a structural change in the economy that leads to a higher propensity to save, because people think of themselves as having a longer life expectancy. They think of there being more uncertainty about their being forced to retire. There’s more capital flowing in from emerging markets. All of these factors lead to an increase in the level of saving, and at the same time, you have a substantial reduction in the volume of capital investment. Think of the fact that we don’t need malls to the same extent because of e-commerce, or the fact that law firms used to need 1,200 square feet of space per lawyer, and now they need 600 square feet of space per lawyer. Think about all of that, and it means that you’re going to have a lower propensity to invest. My iPhone has more computing power than a Cray supercomputer did the day Bill Clinton was elected president, and it costs $600. So if you have this tremendous volume of saving, and you have a limited number of investment vehicles, inevitably what you’re going to see is downward pressure on real interest rates, and that downward pressure on real interest rates is going to mean that they’re closer to zero, and that may make it more difficult, particularly at moments when the economy is headed into recession, for monetary policies to assure that there’s an adequate level of demand.

Kerr: The other question that I think is on many people’s minds right now is global integration, ranging from migration to trade to the financial sector and so forth. As you look out, do you see what we have right now and the turbulence around it, is it passing in nature? Is it going to be sustained for decades to come?

Summers: I think that, to an extent that’s underestimated by many people, globalization has been less driven by the WTO or NAFTA or other agreements. And instead, it’s been primarily driven by two things: the rise of emerging markets and their just greater ability and knowledge and desire to participate in the global economy; and it’s been driven by technology, which has made it possible to do things like read X-rays in India or to operate global supply chains in a way that’s technologically feasible because of information technology. And my suspicion would be that those trends will continue. If the politics became sufficiently adverse and harsh, that would stop. But even if we don’t make progress in political liberalization of the global economy, I think the naturally liberalizing effects of greater technological connection are likely to drive toward more integration.

Kerr: As we come to the end of the podcast, can I ask you for parting advice for someone in Washington, DC, who’s thinking about the future of work?

Summers: We need to teach people to do things that machines aren’t going to do. That means work in groups with others. That means use emotional intelligence. That means play outside your zone, be able to respond to the unexpected. And that means being able to adapt to new circumstances. And all of those aspects, I think, are going to have to be more important in education in the future than they’ve been in the past. I also think it’s going to be essential that we think about how to find ways to make sure that work that is incredibly important in our society but may not have a viable business model associated with it—mentoring a nine year old in how to play baseball, taking care of a three year old whose mother has to work, taking care of a 93 year old who’s outlived her family—these kinds of work are extraordinarily important but may not be supported by conventional business models. And ensuring that they take place is not just important for the people who do the work; it’s also important for the people who are its beneficiaries and for the broader kind of society that we are.

Kerr: Thank you, Larry, for sharing your thoughts with us on how the economy and the roles of work are changing and the big issues that we face.

Summers: Thank you.

Kerr: And thanks to all of you for listening in.