Intro. [Recording date: March 5, 2018.]
Russ Roberts: Our topic for today is an essay he wrote last year on "The War on Work," an analysis of public policy interventions in the labor market and the state of employment of prime-age folk, 25-54 generally.... Now, you start off with the following--I'm going to quote it and then read a second quote.
In 1967, 95 percent of "prime-age" men between the ages of 25 and 54 worked. During the Great Recession, though, the share of jobless prime-age males rose above 20 percent.
and I interject and say, meaning that the number working fell from 95 to below 80%. Getting back to the quote:
Even today, long after the recession officially ended, more than 15 percent of such men aren't working. And in some locations, like Kentucky, the numbers are even higher: fewer than 70 percent of men lacking any college education go to work every day in that state.
It's kind of incredible. And then you say:
The rise of joblessness--especially among men--is the great American domestic crisis of the twenty-first century.
And, I'd like you to defend that claim, to start with; and say anything else you'd like to add in introduction about the facts and how you see this, since you wrote that essay, if anything's changed.
Edward Glaeser: So, no, I don't think that anything has drastically changed. We've had a little bit of recovery since the Recession, certainly. We focus on men rather than women not because female joblessness isn't important, but it is somewhat more complicated because, as we know over the past 60 years, the share of women in the labor force has risen dramatically; and consequently it's--there are more often cases in which we think a woman not being in the labor force is really something that's voluntarily and in many cases quite benign. But, we have a fair amount of data that suggest that the lives of the jobless, particularly the lives of the long-run jobless, really are extraordinarily tough, and in many cases much worse than the lives of those who are working and earning a little bit less income. So, if we look at life satisfaction, happiness, we look at divorce, we look at opioid use, we look at disabilities of a variety of different forms, they are found disproportionately--wildly disproportionately--among the ranks of the jobless. Now in some cases, of course--in cases of disability, often it's the disability that's causing the joblessness, not the other way around. But, it's hard not to think that joblessness is a great source of misery for many Americans. You know, when you just think about what we want from life, thinking that we solve this by just giving people a little bit of extra money just feels completely wrong to me. For so many of us, you need a purpose in life. You need a sense that you're time on this planet is being used for some general effect. The jobless lack that. And, in some sense that's why it's a particularly tragic outcome, that so many Americans have ended up in the state of joblessness.
Russ Roberts: Yeah. There's a story--I don't remember the source of it; I don't think I've told it, although I tell this story a lot, so I wonder if I've told it on EconTalk before; if I have, I apologize to listeners who may have tired of it. But, it's the story of a many who is imprisoned and his sentence is, among his tasks is to, while in prison, is to turn a large, giant crank at the end of his cell. And he spends many hours a day turning this crank. And when his sentence has ended, one of his--besides freedom, he's excited to find out what he's been doing in turning that crank. Has he been generating power for the prison? What's been the point of it? And he asks; and they say, 'Oh, it's not connected to anything.' And in many ways that makes the punishment so much worse. And why that is, is an important human phenomenon. We'll get back to this issue of meaning in life. It's a weird thing, that we should get meaning from our work solely; but I think it's increasingly the case. For two reasons. I think meaning elsewhere is harder for many to find; and secondly, many of the jobs that we have now have more opportunity for self-expression and a chance to change the world.
Russ Roberts: One thing I want to clarify, that I'd like you to clarify is: When we're talking about joblessness, we're not exactly talking about unemployed people. A lot of people think, 'Well, what's the problem? Unemployment is really low.' Explain the difference between joblessness and unemployment, technically.
Edward Glaeser: Technically. So, unemployment means that you lack a job but are actively looking for work. You say that you want a job. That's, at this point in time, less than a third of those people what would qualify as jobless. The remainder are people who have left the labor force entirely. Meaning that they say they are no longer interested in employment. Now, that distinction is real in an empirical sense. Right? In the sense that those who have left the labor force look different along lots of different dimensions than those people who are [?] unemployed. But sort of from economists' conceptual point, right--everyone is willing to take a job; it's just a question as to what's the price of it. At least that's how we[?] tend to think. So, maybe those in the labor force have a higher reservation wage, so we would say in our econ world--maybe they'd require a larger wage to get them out of joblessness. But all of this group is fundamentally lacking employment; and presumably for the right wage, they would enter the job market and start working.
Russ Roberts: And one other quote I want to read to start with, because I think this is a key part of the story:
Joblessness is disproportionately a condition of the poorly educated. While 72 percent of college graduates over age 25 have jobs, only 41 percent of high school dropouts are working.
I'm going to read that again, "only 41 percent of high school dropouts are working."
The employment-rate gap between the most and least educated groups has widened from about 6 percent in 1977 to almost 15 percent today. The regional variation is also enormous. Kentucky's 23 percent male jobless rate leads the nation; in Iowa, the rate is under 10 percent.
Now there's a bunch of different statistics in there. I want to just unpack that for a minute. First, when you say "41 percent of high school dropouts are working," is that 41% over the age of 25?
Edward Glaeser: That's--I've gotten better on those numbers. So, those numbers are right; but they are somewhat misleading because they include everyone over the age of 25. So that's going to include all the retirees as well. So, usually, I prefer to focus on the 25-55 year old age group where the gap is somewhat smaller. But those numbers are particularly dramatic, and it's precisely because you've got retirement in there as well.
Russ Roberts: And, just to make it clear for the data-oriented folks, and those who might not be data-oriented: The fact that there's been an increase in the size of the 65-and-over population, and I assume that they are disproportionately not college-educated, at least for a while, is also going to distort that gap a little bit.
Edward Glaeser: Absolutely. Absolutely, yeah.
Russ Roberts: And so, when you say the gap is, the employment rate gap between the most and least educated--when you say "6 percent," do you mean 6 percentage points, or 6%?
Edward Glaeser: Oh, I think I almost assuredly mean 6 percentage points.
Russ Roberts: I do, too. So, why is it only 15% today? Given that it's--in the number that you started that quote from, it looks like it's 31 points--72 versus 41.
Edward Glaeser: I see--I think what that paragraph unfortunately does is it combines a fact that's true for the overall over-25 with the fact--with my preferred way of presenting the facts, which is between 25 and 55.
Russ Roberts: Okay. So, which is the latter? The "6 percent"--"15 percent," is that in the 25-54 group?
Edward Glaeser: That's right. The smaller one is going to be the 25-55.
Russ Roberts: But it's still--that's a big gap. So, it's regional. It differs by region. It differs by education. What's going on? What do we think--and it's important--the reason I'm glad you focus on 25-54, a lot of the numbers are going to be distorted by the fact that Baby Boomers are going to be coming through the, are starting to retire, and that's going to change--the size of that cohort is going to change the proportion that's working over 25, just through the demographic change, not through anything going on the economy necessarily. So, until further notice, let's focus on the 25-54, or 25-55, which is called, typically, prime age.
Edward Glaeser: Yeah. Don't you find that offensive, though? I think we want to be on record as saying that prime age stops at 55 is offensive.
Russ Roberts: It reminds me of meat. And the USDA [U.S. Department of Agriculture]. But, that is one problem. Your second point is excellent--really thought about that. I'm 63. I don't know how old you are, Ed.
Edward Glaeser: I'm 50.
Russ Roberts: You're still in your prime. But, only a few years left. I'm past my prime. I'm not past my expiration date, though. That only happens, I think, when you outlive your expected life-expectancy, which I haven't got to yet. But, so, humor aside--which I appreciate--we're going to focus, until further notice on prime-age folk. Typically--which means 25-54, 25-55. People who we'd normally expect to work, or at least people who in the past have overwhelmingly been working and now much less so. Why? What do we understand as the possible causes?
Edward Glaeser: So, I think the two primary potential culprits are labor demand and labor supply. So, the labor demand view is that once upon a time America had an abundant amount of demand for men who were strong but perhaps not particularly well-educated in terms of formal schooling. They worked in factories; they worked in coal mines; they made this country run; they made this country great. And, over the course of the past 50 years, the demand for less well-educated men has plummeted. And, consequently their relative wages have certainly going down. Over some decades, even their absolutely wage has gone down--meaning wages corrected for inflation have gone down. And that's one part of the story. And, some of the geography lines up with that. When you use things like industrial shares to, say, 'Were you in a region that had industries that have particularly lost employment? Are you more likely to be jobless?' The answer certainly is yes. 'Were you in a location where people were particularly, where your industries were particularly exposed to global competition in a variety of different ways? Are you likely to be jobless?' And the answer is yes. So, that's one part of the equation. But, there's a second part of this equation, which is the labor supply part. And that has to do with, you know, 'What kind of a wage do you require to work?' And I think there are at least two different stories, one which I probably think is slightly more important, for why the labor supply of these workers has also crunched[?] in, has also diminished. And the main story for that is we've made joblessness less awful. We have benefit programs like disability insurance that actually make it possible to be jobless and to survive. In many cases, when you look at the financials of jobless in America, they are living in a household where the income is not totally unreasonable--in part because they've got a working spouse, or in many cases because they are living on their parents' sofas in their old bedroom. Right? So, they have a source of income; it's just not their own. So, the costs of being jobless, which could have, a hundred years ago would have meant staring at the possibility of starvation, those costs have gone down. And, at the same time, the costs of working, because of public benefits being tied to how much you earn, those costs of working have gone up. And so that's skewed, that's pulled back the labor supply. Now, there's one other story which I think we can associate maybe with Charles Murray, which is maybe a decrease in the soft skills needed to hold a job in certain parts of America, and I think that's much harder to quantify. The discussion around Hillbilly Elegy, for example, was very much around this, around sort of not having the emotional makeup to go to work and behave well for 35 hours a week. I don't know fully how to quantify that. But the other two, the labor demand being pulled back at the same time we see labor supply being pulled back, both because we have a more generous welfare system and more generous relatives, and we have a certain number of benefits, which, scaled down, where you earn more.
Russ Roberts: Just a reaction, just a comment on the demand side: People often say that, in 1950 a high school graduate was middle class. And, they could get a job in manufacturing or construction, and have a decent living. And it's clearly the case that having a high school education only, today, or worse, being a high school dropout, is not as economically viable as it was in 1950. And, when I think about that, though: My first thought is: In 1900, if you were a blacksmith, you probably had, made a good living. But very soon after 1900, being a blacksmith was not a good occupation. And we don't consider that a problem. We don't say, 'It's a shame: It used to be you could make a good living as a blacksmith. Now you can't.' We'd say, 'Well, that's because we innovated. People innovated. They've developed cars. Horses were less useful. Blacksmiths were less in demand. And people learned very quickly that instead of becoming a blacksmith, you had to become something else.' And it wasn't really a crisis. For somebody who had been a blacksmith for a long time and who struggled to acquire a new skill, there was genuine hardship there. But, we generally would say that that's okay. So the question is: What has changed? And I think the simple answer is that, if high school doesn't pay any more as much as it used to, and if that happens quickly, it's hard for people, and the economy, the infrastructure of the economy and the networking that connects people with jobs, doesn't have time to respond maybe as well. Do you see any difference between how creative destruction affected the labor market, say in 1900 or 1940 or 1960 and the way it does today? Why does it seem so much harder for people to adjust to the changes in, say, labor demand in response to globalization or other changes that reduce the demand, say, for less-educated people?
Edward Glaeser: So, that's a great question. I think there are a number of different explanations. So, one of which is that, if you think about the changes in 1900, 1910, there were a number of entrepreneurs, capitalists, innovators who were well-incentivized to figure out how to employ various forms of labor that had become redundant, whether or not that was because of mechanization in the agricultural sector or the fact that the blacksmiths were going out of business. But, Henry Ford, Sloan over at GM [General Motors]--they wanted less-skilled labor. They were willing to pay $5 a day for it. And they wanted to get those workers doing stuff for them. One of the things that's changed is that, whereas Henry Ford innovated in a way that provided tens, hundreds of thousands of jobs for less-skilled Americans--they just needed to learn his skills and follow his rules--Bill Gates innovates in a way that employs disproportionately high-skilled software engineers. And that's true for much of the great entrepreneurs, great innovators of the last 30 years. Not all of them. Uber certainly is a less-skill-labor skewed innovation. Even Amazon does a bit of that, too.
Russ Roberts: Yeah, quite a bit, I think.
Edward Glaeser: Absolutely. But, the changing nature of innovation has meant that there's more of a complementarity between skilled workers and other skilled workers rather than between skilled workers and unskilled workers. And in some sense, I often say that sort of every non-employed American is a failure of entrepreneurial imagination. And, I think their imagination is failing just because of the returns for innovating for the skilled are so much higher [?]. That's one aspect. A second aspect is the regional differences. And here we really do see a difference, right? Until 1992, between 1950 and 1992, the inter-county migration rate--meaning the share of Americans who moved across county borders in every year--was never less than 6%. Meaning more than 6 in 100 Americans changed counties every year. Over the past 10 years, since 2007, it has never been above 4%. There was a steady decline from 1992 to 2007, and it's sort of stayed low since then. This is a point that has been emphasized by a number of researchers. I particularly want to highlight the work of Peter [?] and Danny Schoen[?] thinking about these factors. And they've also shown that the nature of migration has changed. Which is that, prior to 1960 people moved to higher income areas. So, the farmers moved to Detroit. They moved to Chicago. The Okies in the Great Depression moved to California. So, there's this migration to high-income areas. We've seen much less of that over the last 30 years. Particularly for less-skilled Americans. There's very little that's directed toward high-income areas. And, one possible explanation for this is that this has to do with the restrictions that we've put on housing markets in these areas. That, yes, you could find work in Silicon Valley if you are a less-skilled person working in, you know, a variety of service industries; but you are going to have to pay for housing in that area. And, the overall deal doesn't look particularly good when you are kind of happy sitting there at home in Kentucky. So, this migration has really shut down dramatically; and that's a second change.
Russ Roberts: Yeah, I love that story, as an economist. It's a good story. I worry about how much evidence we have for it. It's true that it's gotten more expensive to live in cities. Right? Because of--in certain cities, anyway. Not all cities. That's one problem with that story. It's certainly true in New York City; it's certainly true in San Francisco. I'd suspect it's true in Boston. Seattle, I'm not so sure about. Less true in Houston. Less true in lots of other medium-size American cities. And it's strange that those cities can't attract, don't attract, etc., the people who would normally be attracted to that opportunity relative to what they have at home, where they are now. The other issue, of course, is that wages can adjust in those cities. If there were a large supply of folks working--if there was demand for those workers, the wages would compensate for the higher housing costs. And yet, they don't. Or at least partially compensate, somewhat compensate. But certainly relative to what the opportunities are for a person living in rural Ohio or Kentucky, where things are not so good. You'd think they'd want to try to move to Cincinnati or Cleveland. Which are, again, not having big housing booms.
Edward Glaeser: Right. Well, Cleveland in fact is a depressed city which has its own joblessness problems. The geography of joblessness is very much centered in what we are calling the Eastern Heartland of the United States, which is a swatch which starts down in Louisiana and Mississippi and runs through Appalachia up through Ohio and Pennsylvania. But, your point that there are medium-size cities that are less expensive is certainly true. Texas, for example, remains relatively affordable. I think in terms of the housing price point: I feel entirely confident that there is something wrong in the fact that for the first time in American history we are not making it easy to move into the most economically successful parts of America. That, you know, if we think about the farmers who moved[?] the West in 1820 to the far more rich soil of the Ohio River Valley or of Iowa, nothing was holding them back in terms of getting together the neighbors and raising a barn, and settling down. If you think about the people coming to the tenements of New York or Chicago in 1900, nothing was standing in their way. Now, you come to Silicon Valley and you face one of the most restrictive housing markets in the country. And I think that's a policy problem. That being said, part of me also agrees with you. I don't think that's all of the problem, at all. And, it certainly doesn't explain why, in much of the post-War era we had this remarkable income convergence that came in part from the poorer parts of America catching up. So, states like--you know, if you take Mississippi. Mississippi is the poorest state in the Union today; it was the poorest state in the Union in 1950. In 1950, there were 18 other states that had double the income of Mississippi. Today, there's not a single state in the Union with double, a per-capita income that's double that in Mississippi. Because it has caught up. And it caught up partially because of out-migration, but also because of in-migration of businesses, of capital. And there's just much less interest in doing that, in terms of moving to low-cost areas, in terms of capital migration. And, often, when capital migrates, it just doesn't employ a lot of bodies any more. When you look at Appalachian coal production over the last 30, 40 years, it's not that coal production was substantially off, at least up until maybe 5 or 10 years ago; but, it had become a far less labor-intensive occupation. And that's what we've seen in a lot of manufacturing and mining industries as well: a tendency to use machines, not to use machines and people, but to use machines instead of the people.
Russ Roberts: So, one response to that you sometimes hear--I'm going to reject it and then I'm going to try to make a case for it, and then you can respond. One story you hear is, 'Well, people don't like to move.' They're used to their family, or they're used to the culture around them. Or, maybe they're used to navigating the welfare system where they live, is another version of this--not an attractive story, but it could be true. And they don't want to move. The problem with that story is that that was true in 1950, too, as well; and people moved, like, a lot more often. So, you have to--just a side-note: Methodologically, when you want to explain something that's changed, you have to have something that also has changed. It can't be enough just to point to the phenomenon. You have to explain why it's gotten larger or smaller. So, I think that's true. [?] I think it's false--I don't think it's true that it's that much harder socially to move, culturally to move. What I think is true, potentially, and this is what I wonder about, is it's all well and good to say that wages are higher in Pittsburgh than they are in rural Pennsylvania, or in Silicon Valley than they are in, say, small-town California. The problem is whether the people who live in those rural small towns have the skills that would allow them to earn anything close to what the average is in the cities. And this comes back to an issue we've talked about a few times here on the program which I find extremely interesting, and I want to move to that in our conversation and get your reaction to it. Which is: Do people in cities make more because the cities make them more productive--that there's this synergy? Or, do they make more because the people who happen to move to cities are different from the people who don't? And so, if it's the latter, then it's not surprising that people don't move. They don't have the skillset to thrive. It's not the cost of housing. It's that the demand for their skills is not so much higher in the city than it is where they are, and so there really isn't that much return.
Edward Glaeser: So, there are three types of evidence on the urban wage premium. I think we should also be conceptually clear about two ways of thinking about the urban wage premium. So, one way of thinking about the urban wage premium is just the straight, nominal fact that people in cities earn more regular dollars. The second of which is the real urban wage premium, meaning the dollars corrected for the local cost of living. The first--the nominal wage premium--is itself interesting, because if firms are paying more for workers, we as economists naturally infer it's because the workers have to have a higher marginal product--have to be more productive--in cities. And so, it's interesting. But, of course, it doesn't tell you that you should go to cities; it doesn't tell you that you are going to earn a higher real wage. You need to be focused on the second number, which is the real wage premium, if you want to understand migration; and migration or lack of migration to cities. The second thing that we need to focus on is the three different ways we have of thinking about data on measuring the urban wage premium. The first and most naive way is simply to confirm[?] compare urban workers with rural workers and control for observable attributes on this. That makes, causes very little difference, whether or not you control for these individual attributes. It leaves it still large, but as you say, the urban wage premium can very much be driven by, you know, workers in cities being innately smarter. Certainly, New Yorkers would like to have you believe that. The third--the second piece of evidence we have is looking at migrants. So, this is comparing people's wages before and after they move to the city. I think I sort of began this literature 25 years ago with a paper called "Cities and Skills." But there's a much better paper on this by De la Roca and Diego Puga using administrative Spanish data, which shows that people who come to cities experience faster wage growth when they are there. So, it's not that you come to the city and you immediately experience the full wage burst. What happens is that once you show up in the city, year by year, month by month, you experience faster wage growth. And then you typically take that away with you if you move to another city. Now, that certainly points to the view that cities are forges of human capital, places that we get smart by being around other smart people. But, you could still make the case that I'm only going to come to the city when I know I'm in a particularly fertile time in my life in which I see lots of wage growth. And so even this is biased. And so, if you aren't going to believe the results that control for individual fixed effects basically are controlling before and after for these workers, you've got to go to the much more limited number of studies where we have, let's say, immigrants to Sweden facing programs where they randomize where the refugees get settled within Sweden. And there you also see significant effects of place. But, we're relatively limited in terms of the experimental evidence that we have for moving people around; but we do have a little bit of it. But that's the only--if you really wanted to make sure that none of it is reflecting omitted individual characteristics, you'd have to have experiments where you really put Person A in Place A and Person B in Place B. And, it was completely random.
Russ Roberts: Yeah; I want to just throw a bone to the cultural story, which I think is worth taking seriously--it just adds to the complication in trying to measure these things accurately. We as economists tend to assume as a starting place, at least, something close to perfect information. We assume that a rural Ohio, out-of-work, rust belt denizen knows that there are more opportunities in the city and should move there. Or could move there. And, of course, they may not know. Or, they may know, but they may be anxious about it. They may think, 'It won't work for me.' And to cut between the two hypotheses: If you know someone in a city who has moved and done well, you are probably much more likely to move there yourself. So, if you don't have a--what would we call it? A cultural network of folks "like you,"--and that could mean like you in cultural ways, meaning rural; or it could mean like you in skill ways, who don't have the abilities that maybe the people who mostly live in that city already have, and you don't have them, and so you're anxious either culturally about fitting in, or economically about whether it will pay off to move there. You won't have the likelihood of moving that you would have had if you have those connections. But it seems to me we have more ability find folks like us in those cities than we used to. And you'd think that would work in the opposite direction.
Edward Glaeser: You might. Certainly the evidence for chain migration, which is one version of what you are saying where, one migrant to an area then talks to his people at home and convinces them to come. The evidence for that is overwhelming. In fact, J.D. Vance in Hillbilly Elegy even talks about how his urban community was disproportionately populated with people from the same region of Appalachia. You know, when we unpack, for example, the great decades of migration in the United States, there are a couple of things that are going on there. One of which, of course, is the mass exodus of African Americans north. Which often occurred through chain migration--often occurred through one person knowing the other, and trying to find some opportunity outside of the Jim Crow South. And also, even the move to the Sun Belt. I mean, think about how many people sort of retired to Hilton Head or retired to Florida but want to be with people who are their friends. Again, these social ties are important. You are big on how technology has made that easier to find someone. And I guess there's some truth to that. But, I suspect that even looking at the Facebook friends of people from distressed communities, they are not going to have too many people who have upped and moved to Silicon Valley and can tell them how great it is.
Russ Roberts: Another way you'd think you could measure this would be you'd look for people who tried and gave up. Came back. Do we have any evidence on those kind of folks--who moved to the big city, found it overwhelming or unproductive, and went back to their home region?
Edward Glaeser: So, there is a fair amount of evidence on the Birds of Passage in the great migration north. So, reverse migration to the South, afterwards, there's a fair amount on that. So, people who came north between 1850 and 1870 and then returned back home after then. There's also a lively literature on the European immigrants who came to the United States. Many of them went back to Europe multiple times--sometimes for short stretches, sometimes for long stretches. So, that often occurs. And of course, the world of developing world of urbanization is full of people who leave their families behind and come to work for 6 months, for 9 months, in an urban area during the off season, from planting.
Russ Roberts: Let's turn to the opioid crisis. A lot of people want to put the main causal mechanism for why we have an addiction phenomenon or at least a big use phenomenon with associated tragic consequences--tens of thousands of people dying from overdoses--is because of, it's sort of [?] economic despair. We've had Angus Deaton on the program; we've talked about this. And others: Sam Quinones's book, Dreamland, we talked with Sam. There is some evidence for this. But, we don't seem to know how much. Do you have thoughts on that?
Edward Glaeser: So, it is certainly true that if you look at the geography of opioid use and opioid [?], once again, the Eastern Heartland stands out: once again, this ridge of America running just east of the Atlantic coast but not as far west as Mississippi, stands out as being Ground Zero for this. The Eastern Heartland is also the area where the fact that Anne Case and Angus identified of the reversal in gains on male mortality has been most prevalent--the Eastern Heartland are the places in which men are dying more and opioids obviously play a role in this. I think it's very hard to ascribe a single cause to this. Across counties I've done a little bit of work with David Cutler where we find the share of the population on disability in 1990 is a very good predictor of subsequent opioid use and mortality. There are of course multiple explanations for this. It can be the engagement with--it can be pure, physical pain; it can be engagement with the medical system; it can be economic blight. The one thing I do think is pretty clear, though, is that I would--economic desperation is different, or--the focus on economic desperation focuses sort of too much about income inequality and things that don't seem to me to be as powerful as joblessness for explaining this behavior. Which is: if you are working a 40-hour job and you are earning $23,000 a year, or something that's fairly low, you may feel pretty desperate. But you pretty much know that if you start using opioids, you are going to be getting yourself into lots more trouble. Right? You are going to risk losing that job. You are going to risk having things go from bad to worse. So, the people who are on the economic margins would seem to me to be far less people at risk than the people who are genuinely jobless, who are just looking to fill up an ocean of time and hopelessness. So, again, I think it's sort of important to take whatever share we think is going to economic desperation--I think it almost surely is associated with not working than it is with working for lower wages.
Russ Roberts: But, as you pointed out--and we talked with Eric Hurst about this--again, it's hard to know exactly what's going on; at least, I have trouble wrapping my head around it. But I think we know a little about what's going on. A lot of folks who are not working aren't struggling to fill up the time. They are living at home. They sleep late; they play video games, supposedly. I really don't know--I'd really like to know about the point you mentioned earlier about the resources available. Disability--which has, and we talked about this with David Autor on EconTalk--disability has become much easier to get. And that could be because it's a response to the fact that it's harder for people to find jobs: we don't know which way causality runs there. But, it's not terribly generous. It's not like it replaces most of what you earned. And, yes, there are other people who have their parents to fall back on; there are people who, as you say, might rely on their spouse's income. But, the number of two-earner households has gone down since 1980, contrary to what I think most people think. Most people think, 'Oh, there are more women working.' True. More wives working. True. But there are a lot fewer marriages. So, it's I think harder for people to rely on a spouse. I'm wondering how the folks who don't have a job--what are they getting by on?
Edward Glaeser: So, in terms of our data, and I can refer you to, I'm presenting a new paper called "Saving the Heartland," which is joint with Ben Austin and Larry Summers, at the Brookings Panel this Thursday. And that paper should be on the Brookings website after Thursday.
Russ Roberts: And this episode will air after that. So, listeners can go back and find it. We'll put a link up to it.
Edward Glaeser: Okay. So, in our data, we have, in terms of the share of prime-aged men, we have in our sample from what used to be the Current Population Survey [CPS]; now it's the Annual Social and Economic Supplement [ASEC] from Census Data--we have 18.6% of men who are jobless in our sample. Of this, only 2.8 percentage points of the total population are people who are living alone. The other 15.8% are living with somebody else. Okay. Which means that even thought their individual income--so, if you are living with someone else, your individual income is $8,219; and of that, more than 50% is from Disability. Now, this doesn't mean that every one of those people--and there's a lot; that's 10.1% of the male population, those are long-term jobless living with others. So, you know, the bulk of their thing is from Disability. But many of them are not getting Disability. And, their total family income is over $40,000. But, because they've got someone else with them who is--now, they're not married. That's certainly true. But, you know, a significant fraction are living with their parents. A significant fraction of them have some other person that they've figured out how to get by with, and they're living off that. Or at least that's what the ASEC tells us. Now, the people who seem to live really dire lives are, let's say if you take the long-term jobless who are living alone. That's 1.8% of the male population in our sample. And, of them, you know, their family incomes are $12,500, and $9,800 of those dollars are coming from government support. And $7,300 are coming from Disability.
Russ Roberts: I want you to go over those again. Because I got a little lost in the proportions. First, in your sample, 18% of what are what?
Edward Glaeser: 18.6% of prime-age men are jobless in this sample.
Russ Roberts: Okay. So--so that's--
Edward Glaeser: It's a little high. But because of the sparse nature of the ASEC, of the Annual Social and Economic Supplement, we're pooling the 2010-2015 surveys. So, you are getting some remaining of the Great Recession. That's [?]
Russ Roberts: And you are also getting some folks who might be--are they jobless all the way through the whole thing?
Edward Glaeser: No. It's an annual survey. It's not [?].
Russ Roberts: So, when you say, 18%, you mean over the course of the panel?
Edward Glaeser: That's right--
Russ Roberts: That's in one year of the panel?
Edward Glaeser: On average, if you take the whole panel pool together, 18.6% of the respondents are jobless.
Russ Roberts: Okay. Are jobless, when, though?
Edward Glaeser: At the point in time at which they've been surveyed.
Russ Roberts: Okay. So, in some years it could be higher. In some years, a little lower. But it's about--
Edward Glaeser: Correct. Correct.
Russ Roberts: So, that's a big number.
Edward Glaeser: A big number.
Russ Roberts: Now: Within the jobless population, tell me about the resources they have and how many of them live with--I just got confused in your narrative about what's the denominator.
Edward Glaeser: All right. So, we're going to do a 2-by-2 split of the jobless. If you're living alone, living with others; jobless less than 12 months, jobless more than 12 months. Okay? So, we've got 18.6% of men to parcel out. And so the percents I'm going to give you are the entire male population, not of the jobless.
Russ Roberts: Got it.
Edward Glaeser: Okay. So, living alone, short-term jobless, that's 1% of men.
Russ Roberts: Wow. So, it's just shocking. In a way. Because you think of it as this--jobless, they're moping around; they've got nothing to do. But you're saying--this is the prime age, again, right? So, 1% of the prime-age male population is jobless and living alone.
Edward Glaeser: Jobless less than 12 months.
Russ Roberts: Jobless less than 12 months. Sorry.
Edward Glaeser: Over 12 months, living alone, that's another 1.8% of men.
Russ Roberts: Okay.
Edward Glaeser: So, the total living-alone share is 2.8%.
Russ Roberts: Okay. Still small.
Edward Glaeser: Okay. Now:
|UNEMPLOYMENT AS A PERCENT OF PRIME-AGE MEN|
| ||Length of Unemployment|
| ||Under 12 months||Over 12 months|
|Living with others||5.7%||10.1%|
Less than 12 months, living with others, 5.7%. Which leaves you, more than 12 months unemployed, more than [?] more than 12 months non-employed--so this is, a lot of these people will have left the labor force entirely--living with others, that's 10.1% of men during these years.
Russ Roberts: Which is a little more than half of the jobless folks in your sample.
Edward Glaeser: It's by far the largest category. More than half. Yes.
Russ Roberts: So, let's look at that group. So, this is jobless, prime-age men living with others. Right?
Edward Glaeser: Yep.
Russ Roberts: What are they living on? Excuse me. What is their household income? How much of it is from government transfers and how much of it is from a partner or spouse?
Edward Glaeser: Their family income averaged over this period is $42,000. And, of that, $8000 is coming from themselves. So, $30,000 is coming from someone else in the family.
Russ Roberts: And the $8000 isn't earnings, because they don't have any. Excuse me--they have no measured earnings. They could have under-the-table earnings.
Edward Glaeser: Correct. Although they are of course supposed to tell ASEC about all of their earnings, legal and illegal. But we do think that people might under-report illegal earnings.
Russ Roberts: Yah. Or, they might report them and just not report them to the IRS [Internal Revenue Service], thinking that it's two different things. I could imagine both. It is important to mention that Disabled does not formally mean can't work. It means something more complicated than that. So, people who are formally Disabled, they can't work legally and maintain their Disability payments. Right? Is it 100%, implicit functional tax?
Edward Glaeser: It's a small--well, it's much more than 100% after--I mean, you lose your disability payments once you get over the threshold. So, there's a very mild threshold. Once you get over it, so the tax is enormous. The implicit tax on Disability is enormous.
Russ Roberts: Okay. So, what you're telling me is that the prime-age male population is living in circumstances where their income, over half of the folks in that group, the amount they bring to the, the bacon they bring home, is relatively small, dwarfed by the resources that come, not just from the other member of the household.
Edward Glaeser: That's right. And in fact it's much--remember, living with others for less than 12 months is another 5.7% of the population. So, only 2.8% of men are living alone and jobless; and that's the one share of the population that are not relying on the kindness of family members.
Russ Roberts: So, we don't know whether they are living with their parents or they are living with a spouse, do we?
Edward Glaeser: We do. We have--I don't have this--
Russ Roberts: I know the matrix is getting kind of large.
Edward Glaeser: The matrix is getting large. Surprisingly large number are living with parents. And a surprisingly small number are living with a formal spouse.
Russ Roberts: So, a chunk of this--a nontrivial chunk of this is a change in cultural expectations of what people in their 20s should be doing. So, if they're living at home, I assume they are disproportionately young. And their spouses are okay--for the ones who are married who aren't living with their parents, their spouses are okay with them bringing in their Disability check and not working, having the other spouse, having the other spouse bringing in most of the income. These are men, again. So, I said 'spouse.' These are men.
Edward Glaeser: These are all men. Yeah. I don't know--are they 'okay with it' seems to be a debatable--
Russ Roberts: It's a bit strong. Sorry.
Edward Glaeser: Remember: the divorce rate has also spiked significantly for this population. So clearly not all spouses are pleased with this turn of events. Especially when we look at the time-use of the jobless, where, you know, if what you saw is that jobless men behaved like jobless women and were doing a lot in terms of home care, that would be one thing. But, the overwhelming time use is in watching television, in terms of the big shift. Not in terms of taking care of the kids.
Russ Roberts: I think it's a bigger cultural phenomenon than just sharing household responsibilities equally. I think it has to do with what we think of as masculinity and what we think of as the appropriate role for men and women. And, I'm not going to comment on whether that's good or bad. But they've changed. And they've changed a lot in the last 30, 40 years, for sure. In the last 20 years, probably.
Russ Roberts: Last part about these data, and then I want to move on to thoughts about what might be done. These numbers are very different for college graduates versus non-college graduates.
Edward Glaeser: Yes.
Russ Roberts: We talked about that at the beginning. And I assume that's true in your data as well.
Edward Glaeser: Absolutely.
Russ Roberts: For some folks, this implies: We just need to send more people to college. We recently had Bryan Caplan on the program talking about education as a signal rather than a producer of human capital. That's obviously very relevant for this question of whether we should be encouraging more people to go to college. But, in the data, many people who confuse correlation and causation would be inclined to conclude that: Well, obviously, the way to "solve this problem," if you think it's a problem, is to get more people to go to college. What are your thoughts on that?
Edward Glaeser: So, I think I have some sympathy with Bryan; but also some disagreement. I do think that education is part of the solution. But education needs to do a much better job of providing skills that are actually valued in the labor force. So, if the view is we're just going to send everyone off to college and let them major in concentrations that have absolutely no value in the workforce, I don't see how that helps anything. If we're going to ask ourselves about how to rethink American education, particularly in distressed areas, to provide skills that are actually valuable--and that may involve better community colleges in different ways; that may involve alternatives to community colleges; that may involve, you know, competitively sourced vocational training programs, after school, on the weekends. I think we just need to recognize that skills are going to be at least some significant part of the answer to this. And, we need to recognize that we don't have any easy answer in terms of skills. And, I join Bryan in the view that a simple-minded approach which is, 'Oh, boy, if I could only get the college-going rate up 5 percentage points, that would make everything better,'--that just seems completely wrong to me. But, I do think putting a fair amount of effort into trying to make sure that education is more productive and more widespread is certainly likely to be part of the solution.
Russ Roberts: Earlier you mentioned that some of the technological innovation we've been seeing don't particularly help people with less education or lower skills. And I can't remember that you mentioned Amazon first and I emphasized that, or I brought it up--but, I've read stories recently that Amazon is, for obviously economic reasons of all kinds is putting warehouses in places with low real estate values. You'd think that would be an attraction to find workforces that are relatively less employed. I wonder how helpful that might be, as a way to reduce the geographic problems we were talking about earlier. And, it is a way that at least some entrepreneurs are finding to use workers who are relatively, in less relative demand.
Edward Glaeser: It seemed hard to imagine that it could hurt. The question is whether it's going to deliver anything meaningful beyond a few jobs in the warehouses. Right? Will this actually spin off related businesses in some form? And I think we don't know about that. I mean, incidentally--this relates to the bid for Amazon or different companies--I think it's always important when cities and communities bid on businesses, they really need to ask themselves not how many jobs are being created in the short run, but whether or not this is going to create new business formation in the long run.
Russ Roberts: Yeah. I agree with that.
Russ Roberts: Now, a lot of what we're talking about is about today. But it also applies to tomorrow. And tomorrow is a world a lot of people are very scared about, where technology and artificial intelligence--increasing use of AI [Artificial Intelligence] and computers--is going to make it even harder for people with low skills and less STEM-like [Science, Technology, Engineering and Mathematics] majors to thrive. Are you worried about that?
Edward Glaeser: A little bit. I mean, like you, I also see plenty of upside in this sort of technological innovation. I think it's going to require the smart to be slightly more social. And that's in some sense the monster in the back of this room: is that, the sector that has, that in 1980s came and replaced manufacturing was of course services. Especially for less skilled Americans. Now, services have had a less good 10-15 year track record than they had in the 1980s. But, services require social skills. And, indeed, some of David Deming's work has shown that if you look at wage growth over the last 15 years, even in the STEM-intensive, the skill-intensive occupations, the occupations that have done well are those that actually involve interpersonal interactions as well as being smart. Not just being smart. And indeed, it's easier to imagine outsourcing to a computer something that just involves being smart. So, I think that's likely to be true going forward, as well: that, those things that involve the interplay of making use of our very human skills of communicating with each other, making each other feel comfortable, bringing joy to our fellow humans--those are likely to remain robust. Whereas, those that are, you know, doing something that a computer can easily mimic are not. Incidentally, this is, I think, important in terms of the issue of, particularly, prime-age male joblessness; and it reminds us that schooling can't be all the answer. One of the problems that 50-year-old men often have is they're not great at being nice to other people. And I say this particularly as a 50-year-old man who knows of what he speaks.
Russ Roberts: The curmudgeon effect.
Edward Glaeser: The curmudgeon effect. Which, you know, you can get away with, if you are a skilled worker at a factory for 30 years, or in a mine--you've been a curmudgeon, that's your--or in a tenured academic job. But if all of a sudden you are now looking for a job in the service economy, being a curmudgeon isn't great. And so, sort of, you know, thinking about how to make things work for the 50-year-olds is also a challenge.
Russ Roberts: Yeah. I don't know about this, the social thing. I go back and forth on whether it's going to be the way to the future or not. It does seem like being a human being is going--the human side of human being is going to matter a lot more. But, I just don't know yet how much people are going to be willing to pay for that. You might hope that because are going to make us so much more productive, we'll have a demand for humans, the human touch that maybe we don't care so much about now. But, when it's so much--everything else is so much cheaper, maybe it will be more attractive. I don't know.
Edward Glaeser: Well, I certainly can't claim to have an infallible crystal ball, either. I think it's [?].
Russ Roberts: We opened with--you raised the question early on about meaning in life. A lot of people are pushing for Universal Basic Income as a way to soften the blow of joblessness from technology, if it is coming, if it is as bad as it might appear to be, for some. Sounds like you're not a big fan of that solution.
Edward Glaeser: A world in which 30%, 40% of America all lives on the dole? Is that a good world?
Russ Roberts: Well, they wouldn't be living on the dole. They'd be living on the 70%. It's important to--'the dole' always makes it sound like it's coming from somewhere else. It's coming from the 60-70% who are working.
Edward Glaeser: That's right. For sure. For sure. But I think it's not just an issue of--yes. I mean, I'm less focused on the fact that we would have to--assuming that you and I were still among the working--that we would have to pay for it than I am that it's just--it's a horror for the 30%. It's not just that the 70% have to pay for it. I mean, you're telling them their lives are not going to be ones of contribution; their lives aren't going to be producing a product that anyone values? That they are just going to be taking a check from somewhere else? How in the world does that not lead to incredible social problems within the United States? How does that 30% not become, does not, bored, become incredibly resentful and hostile towards the 70% that write the checks? This is--I think this is a horror story that has been dreamed up. And it's one that, you know, just not the slightest knowledge of what human beings value in their lives. The idea that happiness just involves getting a check seems crazy to me.
Russ Roberts: Yeah. I point out in an essay that if you teach your students that utility maximization is--that the essence of economics and the essence of life is maximizing utility subject to an income constraint, you might tend to overemphasize the material relative to the nonmaterial. We do tend, with our focus on utility maximization, income, GDP [Gross Domestic Product]--and I say this as an economist who often has written on the fact that utility can include non-monetary things, of course. But, we do tend to focus on the material. And, our policy advice overwhelmingly focuses on the material. So, I think you and I are in a minority in worrying about, as economists, worrying about whether a world where 30% or 40% of Americans are not working is, might be a not-so-healthy place.
Edward Glaeser: I think that's right. But I think it's not very hard once you get people thinking about this and once you get people looking at the data of what the lives of the jobless are, to move them in this area. When you tell your students--when you talk to your students about this--right? When you talk to your students about utility, and purpose in life, they don't immediately push back and say, 'Oh, no, the point of my life is to maximize cash.' I've never once had an undergraduate who said that was the right answer. I think you're right: Our profession has done some bad in terms of making it seem as if cash is the dominant thing in the world. It's an important part of the world; but it's certainly not the dominant--certainly not a wise goal for unique maximization.
Russ Roberts: Yeah. I worry that we're different, you and me. You and I are different, in the following way. I don't play the lottery. I'm not going to ask you the personal question whether you play the lottery, Ed. But I don't play the lottery. But if I did play the lottery and I won, I think I'd still keep my job. I like my job. I like what I do; and I like--I like talking to Ed Glaeser, on a Monday morning. It's a treat. So, I think I'd still do EconTalk if I were filthy rich. But maybe I'm different. A lot of people, at least think that they are playing the lottery so they can retire, and live a life of leisure. Certainly, Aldous Huxley in Brave New World, and others, other writers of fiction, have tried to deal with the fact that people maybe aren't so eager to find meaning in their life. They are maybe happy to be on TV, or soma--the drug of Brave New World--or whatever it is that makes us forget, or makes us high, or whatever it is. What do you think?
Edward Glaeser: Uh, I don't know. It seems like these people who, as we said, are living with some other person in their house; they've got, you know, it's not that they are starving; their family income is $42,000, if they're in the long-term jobless or over 60[?] in the short-term jobless; and yet, their self-reported life satisfaction is pretty low. Their suicide rates are pretty high. It's hard to look at the data and be convinced that the Huxley vision was particularly sensible. Now, maybe--and I've heard this line from Silicon Valley--maybe we could restructure society in some kind of way that gives people ways to contribute to the world even when they're not working and enables them to do so in ways that are more pleasant than their old jobs. I just have no faith that America will be able to pull that off. I have no faith in our--I mean, as an economist, I certainly, I believe that culture matters; but I certainly have no idea how to change culture. Whereas, if you are going to work every day, it's not just that you are doing something that someone's likely to be giving you positive feedback for, but you are also surrounded by your mates. You are surrounded by people around you who give you some sense of social connection, and that's also part of why having a job really matters.
Russ Roberts: Yeah, you'd do that in a different way, though, if you don't have a job. You'd go--you'd go root for your football team. Your bowling buddies. Lots of different ways outside of work to find those things. And I sympathize with your concern that we could redesign society. I think that would be a mistake. And, of course, introducing a Universal Basic Income is a very dramatic way to try to redesign society. But, if we went that way, there would be a lot of things that would be set into motion. I would think religion would find some innovative ways to reach out to folks. And, of course, YouTube does that every day. That's what the opiate crisis is partly about--we don't know how much. But, certainly, lots of things come into play. I think it's really hard to know how much people care about "contributing." You and I may have it as--we may see it as a noble cause, that we're contributing: we're doing something that produces value. Other people may see that as a kind of a neurosis: Why do you care?
Edward Glaeser: Well, remember that--so, there's a bit of truth to that. But remember, it's not necessarily that you're going to work in a fast food restaurant and you're feeling, 'Boy, I'm contributing to the world.' It's more that you've got people around you who are telling you you're okay; if you're not getting fired and you are doing reasonably well, you are getting some positive feedback from your boss; maybe you're even getting some positive feedback from your customers, if you're kind of good at being in a service job. Let me just give you a fact, though, about, you know, the socializing of the jobless. So, there is a category in the Time Use Surveys called socializing--and I'm just going to take those people living in, you know, in the Eastern Heartland, since I've split them up into different regions--that the jobless have an extra 360 minutes a day that they've made available by not going to work. Okay? How much of that is allocated toward socializing? 20 minutes. So, it's a very small share of their extra time, they are spending with other human beings. As opposed to, of course, the 160-odd minutes they are spending extra watching television.
Russ Roberts: Yeah. I think actually--that cuts both ways, obviously. I just wonder--again, we're at a particular cultural moment where people are somewhat content to spend time with their screens relative to human beings. I do think there will be something of a pendulum swing back toward being around human beings. And you see that in many ways. Just to make clear what I meant about utility maximization in the material world: In theory, most economic models that we teach our students, and that we use, assume that your satisfaction--your utility--is the same whether you earn your income or whether it's a check from the government. And I agree with you 100%: I think that's probably not true. And I don't think we understand that very well, how it's not true. And I think sociologists--I don't know if they understand it pretty well, but that's what they should--in theory, that's what they should be doing; I don't know if we should be doing it. But, we clearly don't, I think, understand that difference; and I think we ought to think about it.
Russ Roberts: I want to close with licensing. We've had a couple of guests recently talking about the harm from it, and perhaps the potential benefits: recently, going back to sociology, a recent guest, Beth Redbird talked about that--talked about the possible benefits, that it makes people aware of what opportunities are out there, who might otherwise be unaware of occupational options and ways to get into the field through education. I'm a skeptic about that work; but it's interesting. Maybe there's something to it. I'm much more on the other side--that I think licensing is a terrible barrier. Where do you--what are your thoughts on that and what we might do?
Edward Glaeser: So, I mean--I don't know how much you are going to buy these regressions, but certainly in our Brookings paper we have a couple of regional regressions where one of our controls is the share of the workforce that's licensed at the local level. We've done this, I think, both at the states'[?] and the public-used[?] microsample[?] area. And certainly we find more licensing, a more licensed workforce, is associated with more joblessness. Again, I have no magic bullet in terms of causal inference on this. I don't have some randomized treatment that does this. There are lots of stories that you can tell that's just this and causal [?]. But it's certainly true across place, that places where licensing is more prevalent, joblessness is higher.
Russ Roberts: Do you have a magnitude you could share with us? Approximate?
Edward Glaeser: I---oh--I wouldn't want to be--you can take a look at it. I'm sorry: I don't want to be on the record of making a claim that could be off by an order of magnitude.
Russ Roberts: I appreciate the honesty.
Edward Glaeser: There is--it's basically significant in all of the past 4 decades. So, it's--so, certainly, I think, you know, the map that I usually show is the occupational licensing of opticians, where of course--not the optometrists who actually look at your eyes and tell you what prescription you need. They are the guys who actually tell you whether or not you should wear round glasses or square glasses given your face shape. And the idea that this, or many of our occupations need to be licensed, seems absolutely crazy to me. More generally, you know, I think it's outrageous in this country that we regulate the entrepreneurship of poor people so much more strictly than we regulate the entrepreneurship of rich people. If you want to start your Internet phenomenon in Cambridge or Palo Alto, you can get yourself to 50 million users with barely any regulatory oversight whatsoever. If you want to start a small grocery store that sells milk products, you have to walk through 17 permits, in many large cities. This is crazy. And, you know, it certainly seems--and the things that I would be pushing on, along with education, to be fighting for more employment, are reductions of various forms of regulatory barriers, including occupational licensing, but also barriers to starting small businesses in different parts of the country. I'm particularly a fan of one-stop permitting, as being a way of sort of centralizing things and making sure you have a licensing official that can actually talk the language and connect with the relevant populations. I, you know, believe very strongly in rethinking the way that we've designed our social welfare programs that tax earnings in different ways--Disability, of course, particularly strong; but Section 8 Housing Vouchers, SNAP [Supplemental Nutrition Assistance Program], Food Stamp assistance--all come with implicit taxes of earnings that can easily get quite high. And, of course, on the other side, while I'd like to maybe cut back some of those, I'd like to ramp up employment subsidies--maybe given to the worker, maybe given to the firm. But do it in a really obvious way that raises the effective wage, sort of an earned income tax credit on steroids.
Russ Roberts: Other than just not subsidizing education generally--because you, understand, you agree with Bryan Caplan that not every educational experience is productive--how would you improve education?
Edward Glaeser: I think on this, if we are focused specifically on the jobless side, I would be focused on vocational training. And, you know, starting on the fact that we don't know exactly what works, setting up programs where you actually are going to be able to evaluate whether or not the program worked upon the point of graduation. Doing things that are competitively sourced after school, [?], are not necessarily using the traditional educational apparatus but having alternative providers--you know, teaching plumbers, carpenters, various service jobs--training them and sort of seeing what happens--that's probably where I'm most excited. And certainly healthy engagement with the private sector, as well.