Dan Hamermesh: Time, Beauty, and What Economics Gets Wrong About Work

Ben: Why do Americans work so much more than people in other rich countries? What does it actually mean to live a good life in a society organized around work? My guest today is Dan Hamermesh, one of the most influential labor economists of the last 50 years. Dan's work has explored how we spend our time, discrimination, beauty premiums, and the economics of leisure. He has a rare ability to take everyday aspects of life that most people overlook and reveal the economic forces underneath them. Here's our talk.

Ben: Welcome, Dan, to The Economics of Work. I think of you as one of the most expansive labor economists, really pioneering the field and getting into some interesting topics: time, beauty, discrimination. It's a pleasure to have you, and welcome to the show.

Dan: Thanks for having me.

00:39 Grading Keynes's 15-hour workweek

Ben: I want to get into this question of leisure and work and how those interact. As you know, almost 100 years ago, in 1930, John Maynard Keynes published "Economic Possibilities for Our Grandchildren," where he predicted we'd be working 15-hour weeks in 100 years and otherwise living a life of leisure. Does he still have time? Is he right? Is he wrong? Where do we stand? How do you grade him?

Dan: On that one, I'd give him probably a D-minus.

Ben: Wow. Harsh.

Dan: Only because it takes guts to make predictions, though it's easier to do if you're not going to be around to have your mistakes pointed out to you, which obviously he knew he wouldn't be. No, we're nowhere near that. In fact, it's really quite remarkable. The US, by the way, is the worst culprit among rich countries. In the mid-70s, we were in the middle of the rich countries in terms of hours worked per year. Currently, we work the most hours per year. People think, "Oh, the Japanese work harder." No, they don't. They work less than we do. The Central European countries, France, Germany, work much, much less. That's a change over 50 years. We haven't cut back in this country.

Ben: You said the worst. Is that a bad thing? Why do you think it's bad?

Dan: Yes, that's absolutely a bad thing, for the following reason. The purpose of life, in my view, is to enjoy things. Very few people get pleasure out of their work time. I do. I'm totally neurotic. I'm not even being paid for anything, but I still work half time. I would call it work, but for me it's fun. For most people, that's not true at all. So the question is, why are we doing this? Why are we so different from people in other countries?

Ben: Couldn't you make a case that there's some positive externality to a whole country working more? Even if it's worse for each individual person in terms of life satisfaction, we get some positive externality. We can attract capital. The most innovative companies are formed in the most productive societies. We lead interesting lives by being a productive society, right?

Dan: All that does is make us monetarily richer. I'm not saying cut back like crazy. But if we cut work time by, let's say, 20%, my guess is it would cut the total amount we produce by about 10%, one for two. We'd still have real incomes higher than they were 15 or 20 years ago, assuming 1% real growth per year. I don't think people felt poor 20 years ago. Indeed, my guess is they feel poorer today, because, remember, America's national pastime is complaining, and people are complaining more now, even though the average person is better off than she or he was 20 years ago, and much better off than his or her grandparents were.

It's also interesting which dimensions we work more on than other people. The average workweek is the same here as in other rich countries, about 38 hours a week. The difference is that we don't get vacations the way they do in other rich countries. The Australians have this wonderful four-month sabbatical after 10 years of work. So while the workweek is the same, we're working more weeks per year, and more days per week, than other people. This is not a way to live.

04:03 Is the line between work and leisure blurring?

Ben: The other thing I want to get your take on: one way to think about it is that we've got work, we've got leisure, and there are hours of work. But it seems like, for you and me at least, and I wonder if this is happening in a more widespread way, the boundary between work and leisure is blurring a little bit. We're not working in sweatshops to the same extent we used to. Work is getting social, at least white-collar work. We have networks, we have friends.

Dan: You're concentrating on people like you and me, who are a rather small fraction of the workforce. The average clerk at Starbucks or any other retailer does not enjoy that kind of life at all. The average person in an office, even in a manufacturing firm, the average healthcare worker, a nurse: that's not the same as what the average professor or attorney or doctor does. I think the people living the kinds of lifestyles and work styles you're talking about represent well below 20%. When I see all this stuff in the papers, in the New York Times, for example, raving about it, they're catering to a rich audience, and that's not the average person.

05:16 Greedy work or a rat race?

Ben: I'm sure you're familiar with Claudia Goldin's work on greedy work: the idea that there are some types of jobs where the returns to the 60th hour of the week are much higher than the returns to the 30th hour. There's a huge economic advantage to being on all the time, to dreaming about work when you go to sleep.

Dan: I think Claudia is quite wrong. I think what she's missing is that basically it's a rat race.

Ben: You think so?

Dan: I'm working as a demonstration to my superiors, to demonstrate that I'm the best person to be promoted and get a big raise. Or simply that I enjoy it more than other people. But I think she's quite wrong on that. There are very few jobs, even including professor, where working the 60th hour is more valuable than the 40th.

Here's a question: what would you do with your time if you had it? Keynes talked about how people would be doing all kinds of wonderfully beneficial things. We actually have data on two countries where they essentially cut back hours of work, Japan and Korea. And in Korea, the men spent about half the extra time grooming themselves.

Ben: I saw that result, and I didn't really know what to make of it.

Dan: A lot of time grooming, putting facial products on. The Japanese spent more of the time, and this makes more sense, watching television. We know from recessions that TV watching goes up and sleep goes up. Sleep and television watching are the default activities. We know, for example, that if you have a disability and have to cut back on work, a large fraction of the time is taken up sitting in front of the tube or sleeping. And I don't think even Keynes would view those as the most morally desirable activities, certainly not television.

Ben: You could make the case it's even worse. There are a lot of negative outcomes from being heavily engaged on social media. If someone's just doomscrolling on TikTok or Instagram, that could be harmful to their mental health. If that's how we spend our leisure, that sounds pretty bad.

Dan: It does. But on the other hand, I don't think that if they weren't doing that, they'd be doing such wonderful things. Most of them would not be out doing volunteer work, delivering food to poor people, or writing poetry, which is what Keynes might have had in mind. So I'm not that bothered about the 15 hours, but I sure would like to see us cut back somewhat. We also know that having leisure bunched together, spending time off at the same time other people have leisure, is really important. That's the kind of positive externality you're talking about.

Ben: Yeah, right.

Dan: Again, it's not just cutting hours, which I'm all for. It's also how we do it. The way I have in mind is not cutting the workweek, which isn't that long, but rather allowing more paid vacation. Most countries have huge amounts of paid vacation and paid holidays. In this country, half the people have no paid vacation at all, which is just crazy.

08:23 Will policy or AI bring more leisure?

Ben: So do you think that will change? And if it changes, will it be a result of policy or of AI technology?

Dan: I think if anything, it'll have to be policy. When a lot of jobs were destroyed by the introduction of PCs, I don't think we have evidence that it led to people suddenly getting more leisure. They wound up doing other things. Secretarial jobs fell by about 30% between '85 and now. Are those people out there now working much less, spending their time writing poetry, or even watching TV? Heck no, not at all. So no, I'd go with policy. Indeed, that's what happened in Europe. As I said, they were the same as us in the 70s, and the governments took it upon themselves to start mandating paid vacations and paid holidays in substantial numbers. That's what caused it. And unless we can get our act together, which certainly isn't going to happen in my lifetime in this country, we're not going to see any difference here. We're on the gerbil wheel, and we're going to keep on running, though we might switch to other gerbil wheels because of AI.

09:34 Why leisure is a coordination problem

Ben: Why do you think there's such a big coordination failure? Why wouldn't some firm offer mandated leave as a compensating differential, where it could pay people lower wages in exchange? Why doesn't that happen organically? Why do you think there needs to be a policy that solves this coordination failure?

Dan: Let me give you an example. There was a case in a fancy suburb in New Jersey where they wanted to stop having parents spend all their time taking their kids to soccer, tennis, baseball, and so on. They got together and agreed to do this. And of course, some of the mothers started saying no, we want our kids to be the best at violin, the best at soccer. It's very hard to get an agreement when one or two people will benefit from violating it. It's the same as, I don't know if you're familiar with Al Roth's work on jumping the gun.

Ben: I've got his new book here, "Moral Economics." It's on repugnant markets.

Dan: That's a slightly different point of his, but it's another example. Somebody is always going to cheat because it benefits him or her, even though it hurts society as a whole. And the only way to get something that benefits society as a whole, when some people benefit from cheating, is to mandate it. In this country, since our government at the federal level doesn't do anything, as far as I can tell, that's not going to happen.

11:10 Compensating differentials at work

Ben: Do you think this extends to other things people don't like about work? Is the hedonic wage model a little less useful than it's presented in Econ 101? Things like risk at work, danger, and so on. That's another thing that's bad for people.

Dan: It's compensated for. It's just not compensated very well. We show in one of our papers that with a four-day week, same number of hours, people do benefit from it, and they pay for it in the form of lower wages. All these compensating differentials, extra pay for doing dirty work or less pay for doing very desirable work, are there. They just aren't very large. I think the reason is that there's often some overhang of unemployment. Marx wasn't totally wrong talking about the reserve army of the unemployed. I don't like to use Marxian terms, and it's certainly not as stark as Marx made it out to be, but it's there. These wage differences do arise. They just aren't very big, that's all.

12:22 Time diaries and what people do at work

Ben: Another thing I wanted to ask you about: I love what you've done with time use surveys and time diaries. And it got me thinking, as we're all thinking about AI: if you were running OpenAI, would you be collecting time diaries? I feel like if I were running an AI company, I'd want to see how people spend their time working. I'd want to know it minute by minute.

Dan: I would love that. I would love to get my grubby paws on data like that. The problem, and this is what's quite amazing, is that it's very hard to get data from these surveys on what people are doing while they're working.

Ben: Why is that?

Dan: You're asking people to record minute by minute what they're doing, and if you're recording minute by minute, it's very hard to do anything else. Here's another example. One other method of getting time diaries is the so-called beeper method. You give beepers to a large number of people and randomly beep them to write down what they're doing. The problem is, if you did this for everything, can you imagine being beeped during sex? That would certainly spoil the act. And the same goes for work. It would spoil it there too. So I don't see any great way of doing this. We used to have time-and-motion studies. That was in the 1920s, so-called scientific management, where they would watch what people were doing and try to streamline it. That's fine for a mechanical job, or even a job in a retail store. I don't think it works for the 20 or 30% of people in white-collar jobs.

Ben: Good to know. There are companies out there, like HumRRO and APTMetrics, that track what people do: observe them, take little notes, sit in the corner. I guess they try to understand the components of jobs. I think they were the ones that did this for O*NET. But it's not very good, not very detailed. We don't really know what people do in their jobs, which seems like a big gap in labor economics.

Dan: It's not just labor. Come on, it's everybody studying work. If you ask anybody what a lawyer does in her job, maybe they have a clue. Ask what a professor does in her job, and they'd say, well, they teach 12 hours a week, or six hours, or three hours, whatever it is. That's all students ever thought about me. And yet the actual teaching, and maybe holding a few office hours, is a small fraction of the time I spent working. It's very hard to quantify this unless the work is quite mechanical. For me, the most important part of my work is getting a new idea, solving a problem. When do I do that? My best time is when I'm running in the park.

15:13 Predicting how AI will change work

Ben: So if this is really so difficult, how should we think about our ability to understand how AI will affect work? It seems like we'd have to know the components of the job, what people do. If we don't know that, are we just in the dark? How do you think about it?

Dan: Yeah, we're in the dark, and we can trace it out later on and see what happened. I'm giving a keynote at a conference on AI next week. Since I've worked on the issue of what time of day people work, I'm talking about those results, but I also have to talk about AI. So what am I going to do? The answer is, I look at occupations that I think might become AI-intensive. There's a very nice quarterly survey on what workers say about their AI use. I take those occupations and look at how they would fare over time, compared to other occupations, if AI came in.

Ben: But this is total speculation.

Dan: Well, there's good speculation and bad speculation.

Ben: And I would think you have a valuable perspective, as someone who's thought about work for a long time. So what's your speculation?

Dan: My speculation is that in the end, it'll be very little different from the introduction of the PC. That changed a lot of our lives and cut the demand for certain clerical skills. It certainly made the people who use PCs a lot more productive. It should have improved the quality of writing, because it's much easier to revise things than it was before. On the other hand, it's the same as seatbelts. There's an incentive to become less careful, because you figure something will catch your mistakes more readily. I'm quite sure the net effect of seatbelts is lives saved, and the net here is better-written work, but it's not a 100% gain, that's for sure.

Ben: I hadn't thought of that example, seatbelts and AI, but I think that's a rather astute way of thinking about it. It's going to make us sloppy.

17:13 Will AI make us sloppy?

Ben: And are you concerned about the pathway to learning and development? If people use AI as a crutch, are they going to develop less and learn less?

Dan: Did the development of the hand calculator ruin our ability to do arithmetic? My sister, who's a math instructor, says no, it saves you from doing the dirty work. You can do better work by thinking about the mathematical concepts. I don't see why AI should be any different. People get bored. I get bored quicker than most, and people want to do different, new things. I think that will get around your concerns about AI routinizing work activities. I expect it to routinize some things and give us more time. I'm a real optimist about this, as I am about most things.

18:02 Remote work and the beauty premium

Ben: So let me bring up another hot topic for you: remote work. What do you think remote work does to the beauty premium?

Dan: Well, one very neat result in our new paper is that during COVID, remote work time actually got more concentrated in the main part of the day, despite what you might think. You'd think people would want to work any old time, given the diversity of their interests. But remember, if others are working at a particular time, then unless you work totally in a vacuum, independently of anybody else, like a novelist who can work whenever, you may well have to work at the same times as before, because they are. And that being the case, if I'm working with them, my face is still on the screen, as we are here. So I don't expect it to make much difference in the role of beauty.

Ben: Well, height isn't observable anymore. That's one dimension-

Dan: That's another aspect. Height is not the same as beauty at all, and weight doesn't matter very much unless you're really obese or utterly emaciated. For a normal person, with a body mass index of, let's say, 18 to 30, it doesn't make any difference. So of all the effects of remote work, I don't think that's going to be a biggie.

19:35 What beauty really measures

Ben: I would think there's always a concern about whether the beauty premium really reflects beauty, or whether it's picking up other information. What non-beauty information is contained within beauty? Is that a big part of it?

Dan: I've thought about this an awful lot. I've thought about it right this week, because even though I keep vowing "stop me before I do any more beauty research," I think I'm getting involved in another one. There are many methods of measuring beauty in the literature. One is face to face: if we're done with this, I give you a rating of five, four, three, two, or one. Or I could look at a picture of you. Or what some people do is try to measure it with very detailed facial measurements, like symmetry. But none of these is perfect. The picture has a problem because it's affected by how you look in that picture. With symmetry, there's more to beauty than that. There's a very well-known economist who I would say is average looking, even though her face is very symmetric. It lacks a certain je ne sais quoi.

And the problem with the face-to-face rating is that it's very hard not to be affected by the milieu in which you see the person. In other words, I see that New Yorker picture on your wall, and I think, gee, this guy is pretty good looking, because he has this wonderful picture of Central Park, looking at the pond, right near my house, by the way.

Ben: Right, our background is part of it.

Dan: I've used pictures, and I've tried to separate each picture from the others being rated, so there's no intervening milieu that might be skewing things. But as I've said since I started doing beauty research in '92, beauty is in the eye of the beholder, and yet there are tremendous correlations across people when you ask them to rate others. I think if there weren't, there would be no effect of beauty, because it wouldn't be a definable concept.

Ben: Sure. But let's say all of that is totally true, and there's an agreed-upon standard for assessing someone's beauty-

Dan: Tends to be agreed upon.

Ben: Yeah. There's still potentially more information within that. Let's say you have some theory that beautiful people have access to more elite groups and successful people, hang out with successful people, and learn from them. Or there's sorting in marriage, what have you. So you're maybe picking up on socioeconomic status or education or something.

Dan: Of course that's the case. Looks affect every aspect of your life, including things like hanging around with people who can refer you to better jobs, and so on. But again, in all the research on this, I and others try to abstract from those other things: the fact that I might get more education, that teachers may pay more attention to me, that I may be more self-confident. Then the question is simply what effect looks per se have on the outcome of interest, be it the probability of marriage, earnings, promotion, and so on.

Ben: So I'm curious. Let's say there are two components to the beauty premium. One is causal, directly attributable to pure attractiveness, and the other is picking up a bunch of covariates and other things associated with it. My guess is that in a remote work world, we have more ability to present ourselves in a nice little frame and manipulate how we're perceived a little more. So the pure beauty effect would be lessened to some degree, but the covariate stuff would still be there to basically the same degree. Do you think that's plausible?

Dan: I'm not so sure. Remember, those covariate things have all happened already. I'm sitting here, and all you see is me in our second bedroom, which doubles as my office. But all the things that make me such a superstar, which I'm not, have already been affected by my beauty or lack thereof. So I don't think it's possible to abstract from those things even there. It's already happened. It's embedded in us before we get on the screen to talk with our coworkers or anybody else.

Ben: Yeah, so it's baked in.

Dan: Baked in. That's a great term. I like that, thank you.

24:27 A curmudgeon's view of labor economics

Ben: I want to get your reflections on economics, and labor economics in general, as a field and where it's going. I know you've thought a lot about this.

Dan: I have, a lot. I'm a curmudgeon, okay? One of my co-authors in Germany is now called a senior professor. I wrote back to him and made up a German compound word for somebody my age who's out of it: a brain-dead professor. That's my new title for senior people. Anyway, I'm a senior curmudgeon. And I don't think I'm atypical of other senior people. I have a lot of friends and co-authors I work with, and they seem the same.

I find, and I'm happy to be quoted on this, that the typical paper is: find something exogenous, show that X precedes Y. Isn't that fun? It often has a title like "Blah, blah, blah: Evidence from...," which is the most mindless thing I've ever seen. Even in those studies, I'd like to know what's going on that leads X to affect Y. There was a nice little paper presented last week where someone showed that in one country, the spread of cell phone coverage affected some outcome. There could be several channels for that, and he had not thought about trying to infer which of them was causing the outcome. How can you have an effect on the world, beyond showing that X precedes Y, if you can't figure out why X affects Y? So I think all we see is whether X precedes Y, and very little thinking about what's going on. In other words, there's very little economic thinking here.

26:28 Is price theory dead?

Ben: A few weeks ago, Tyler Cowen wrote a sort of obituary for price theory. Do you think price theory is dead?

Dan: Good grief, no.

Ben: Okay.

Dan: Not at all. You've asked the wrong guy. Look, I was a Chicago undergrad, and I took grad classes. I took Milton Friedman's graduate price theory class in my third year of college. I absolutely believe in it. I wouldn't be talking about compensating wage differentials, which we discussed, and which is a pure price theory concept. I wouldn't be talking about why weekend work and night work are done more by people with low incomes if I didn't believe in price theory.

Ben: No, no, I'm not saying we shouldn't believe in price theory.

Dan: You mean we don't pay attention to it anymore.

Ben: Yeah. What I really mean is, do you think the field has forgotten about price theory, in a damaging way?

Dan: Absolutely.

Ben: It seems like it's in decline in academic economics.

Dan: I've said it all along: if you write a paper that could be done by a sociologist who's really good at using Stata or R, you shouldn't be doing it. You're not doing economics. We have something better to offer, namely our analytical tools, of which price theory is number one by far. If we're not using price theory, we might as well call ourselves sociologists and not bother teaching this stuff in grad school. Leave people alone and let them run their regressions; they'll be happier. If that's all we're doing, I think it'd be a disaster, not just for the economics profession but for the world as a whole. Because the fundamental notion of price theory is that prices affect behavior, even in cases where there's no dollar amount directly attached. Ignoring that means we don't contribute very much to society.

Sorry to rant about this.

Ben: No, no.

Dan: I've thought about this a lot. It drives me nuts, but I'm by no means atypical among more senior people. Indeed, there's one person, I won't say who, who's been very much behind the push to look at causation and make sure X affects Y. I've talked to him a couple of times about this over the last 25 years, and he says he'd like to see more theory in this stuff. The problem is that doing theory is hard. Getting a data set, finding a natural experiment, running a regression of Y on X and a bunch of covariates, and looking at what happened: that's easy to do. A sociologist, pardon me, this is nasty, can do that too. But they can't think about the role of prices in behavior. That's what we do disproportionately.

Ben: I think another point in support of what you're saying is that it allows us to answer some big questions. Our mutual friend Mike Grossman likes to say it's better to be first and wrong than last and right.

Dan: Absolutely. Or rather, it's better to be first and wrong than first and right on something so narrow that all you can say is that X precedes Y.

29:58 Causal identification and the replication crisis

Ben: I think that's really what he means, even though it's a bit tongue in cheek. The real point is to take your best shot at something big and important rather than getting fine-grained precision on a small question. But just to take the counterargument, even though I agree with you for the record: fields that haven't gone through the so-called causal identification credibility revolution have had big replication crises. Could it be that an obsession with very clear identification of causal parameters has spared us the replication crisis that has plagued psychology and other fields?

Dan: Really? If I show that a program to improve the quality of water in wells in southern Uganda unequivocally has a causal effect, have I learned very much? This is the issue called external validity. Are we going to go do water wells in every section of every African country, and will people agree with it? People are not lab rats, okay? And as long as people are not lab rats, the only issue is whether you've cooked up something interesting that others can replicate elsewhere if they want.

I find the notion of replicability another red herring in a lot of ways.

Ben: Interesting.

Dan: Psychology has the same problem, and they're not using lab rats either. For example, I was just going over a paper that should have been published long ago. People measure changes in inequality, and how you measure various things affects the outcome. Does that mean it's a replicability issue? No, these are different ways of defining the problem, and by defining it in a lot of different ways, we might get a general consensus. So no, I think that's quite wrong too. It's somewhat of a waste of time. If an idea is really important and wrong, people will follow it up, and other examples will show that. Take the Reinhart and Rogoff paper, 15 years ago, on government spending and debt.

Ben: Oh yeah, the effect of debt-to-GDP on growth.

Dan: A grad student nobody had heard of tried to replicate it and showed they'd made a very simple, stupid mistake.

Ben: Yeah.

Dan: Another example, long before you were born, made it into People magazine in 1980. Marty Feldstein, may he rest in peace, had shown that Social Security displaces private saving one for one, in a paper in the Journal of Political Economy. At the time, some obscure economist nobody had ever heard of pointed out that he'd made a coding error, and in fact the effect was below a half. And I think Marty was written up in People as loser of the week after that came out.

Ben: Of course.

Dan: But again, the only reason the guy tried to check it is that it was an important result, which happened to get a lot of publicity as well. That's why I'm not worried at all about replicability. If something's really important, people will replicate it, catch errors in it, and show whether or not it's correct. And even there, you can be correct in one example, but if it's not correct in 10 other examples, it's not externally valid, to use the current phrase, and you haven't shown very much.

Ben: I think you're right that our concept of replicability really has to mean external validity. It has to mean external replicability.

Dan: It's not sufficient otherwise. Narrow replicability is fine for physical experiments, maybe. It's fine if you have genetically pure strains of rats in a biology laboratory. But that's not going to happen in the social sciences, and I don't expect it to.

34:06 Why economics publishing is so slow

Ben: Since you brought up long publication times and wanting to get a paper out: in economics, there's a much longer lag in getting a paper published.

Dan: I had a paper on that last year. You might have seen it.

Ben: Is this a way to protect incumbents?

Dan: Perhaps it's a way to let incumbents choose who will become incumbents along with them. But given that we have tenure, at least in some places we still do, why should I as an incumbent worry about protecting myself? My job is safe, or was. I was tenured early on and never worried about it. So the question is why it takes so long. It's a different sort of low-level externality. For some reason, our profession has gotten the idea that referees should go over papers again and again and again, and that authors should spend huge amounts of time responding to every single referee question. Most of the time is taken up by the back-and-forth, with authors writing responses to every single point the referees made. This is a disaster. I know full well that responses to referees are often longer than the resubmitted papers themselves.

That's one problem. People worry about this. There was an international commission of editors that said we should do this or that. They were basically very namby-pamby recommendations that I'm quite certain will not make any difference. The one thing that does make it faster is that in economics, journals have gotten into the habit of rejecting most things out of hand. If you're only going to publish 5%, I can't see wasting referees' time on 100% of submissions. That helps a bit. But at some point you get a paper that's refereed, and you go back and forth, back and forth, back and forth.

The particular paper I mentioned, we sent to one journal and got four referee reports. They said no. We sent it to another one and got two reports. They both seemed to like it. We wrote a response. It sat with one referee for six months. Eventually she said we hadn't cited X, Y, and Z, two of which were unpublished papers by the same author, who in my mind must have been the referee.

Ben: Yeah, that's clear.

Dan: We got it back in three months: make these changes, it looks pretty good. We sent it back. It took five months to get it back again, which was two weeks ago.

Ben: Is this a stable equilibrium, or do you think it will change?

Dan: It's been stable for the last 30 years. I don't know how long stability can last, but I don't see it changing. There are a few good things about being really old and not really needing to deal with journals, although I do constantly. The situation these days is terrible.

My favorite story on this is my first paper in a major journal, the Quarterly Journal of Economics, a so-called top five journal. I sent it in October of '69. I got it back in January: make a few changes. I sent it back in February, and it was accepted. It was in print in August of '70. That's 10 months from initial submission to print. Now, print doesn't matter very much, because everything is put on the web in final form right away. If that happened today, it would have been available to everybody in final form in March. That's not bad: five months from initial submission. These days that doesn't happen unless you're in the sciences. I submitted a paper to the Proceedings of the National Academy of Sciences in September. It had two go-rounds, and it appeared in a February issue of PNAS. That's the way to do business. The sciences do that. We don't at all. In economics, we are the worst of all the social sciences by far, and we are killing ourselves.

Ben: Why do you think economics is different in that way? Why aren't other fields doing the same?

Dan: I've had friends who are editors of journals in other fields, psychology in particular. What they say is: if we want something, we'll say make these changes and send it back. If you make the changes satisfactorily, we'll accept it. If not, you're done. Whereas here we'll go through three, four, occasionally, very rarely, even five rounds. And sometimes the fourth round will kill it. Everybody thinks he has to be a co-author. How we got to that ethos, I don't know. It didn't used to be that way, as I indicated. And I don't know how to get off the hamster wheel unless you can convince enough people to jump off, and we haven't been able to do that. There's the hamster wheel again, a nice analogy for bad things.

39:17 What Dan would research next

Ben: Totally. I've got one more question for you, and thanks so much for your time. If you had another 82 years of research, what would you be most excited to work on?

Dan: Hell, I'll take 10 more at this point. I'd like to do the stuff we talked about, time use research where we actually see what people are doing. I think that would be very, very valuable, for the AI questions and other things. The problem is that you have to either get some company to agree to it or go out and get grants. And at age 82, I certainly don't have the energy to do that. Indeed, I haven't applied for a grant since I was 65, because my wife told me then, "Daniel, you often get grants that require you to do things you don't want to do. We don't need the money. Stop it." And I've been married 60 years, so I listen to what she has to say. [laughs] Anyway, there's your answer.

Ben: All right, fair enough. Dan, thanks so much for a fascinating, wide-ranging conversation. It was a pleasure having you on the show.

Dan: Thank you, Ben. Thanks for having me.

Ben: The Economics of Work is brought to you by Revelio Labs, workforce data for research and benchmarking. If you enjoyed this episode, please remember to rate, review, and subscribe. It'll help us reach other curious listeners. I'm your host, Ben Zweig. Our producer is Cole Wagner. Thanks for listening.