Jason Feifer: This is Build for Tomorrow, a podcast about the smartest solutions to our most misunderstood problems. I’m Jason Feifer, and in each episode, I take something that seems concerning or confusing today and figure out where it came from, what important things were missing, and how we can create more opportunity tomorrow. I have a very dorky annual year end tradition, and it goes like this. As we are about to enter a new year, I start to wonder: What did people from 100 years ago think this year would be like? So for example, as we are to enter the year 2023, I start to wonder: What did people of 1923 think 2023 would be like? And as it turns out, this is not actually a very hard question to answer because of newspaper archives.

Every year, I just go to newspapers.com, which is the archive that I happen to use, and I go to newspapers from 100 years ago. So for example, in this case, I go to newspapers from 1923, and then I just search the term, the year 2023. What comes up is always an amusing smattering of predictions because people from 100 years ago loved predicting what 100 years later would look like. It’s generally a mixture of experts and school children, for some reason. And they say all sorts of fascinating stuff. Some if it is, well, incorrect. For example, there was one year where I saw them predict that Mexico would be the world’s leading power in 100 years, and also that flying cars would be so common that there would no longer be doors in the front of buildings down on the ground level because nobody would be entering buildings from the ground anymore, so all the entrance doors would be at the very top.

But like I said, there were also very real predictions, for example, that in 100 years, people would be able to control the temperatures throughout their home, which of course we can do right now. And then also that people would be able to listen to an author read their own book, but you wouldn’t have to be in front of the author to hear it. It would come to you by some kind of magic wire, which is more or less exactly what you are experiencing right now. And I’m always fascinated by this mixture because what you’re seeing is people who are … They’re looking into the future through a pinhole. They see the new developments in their own lives, which back then was driven by the development of electricity and new forms of transportation. And then they were, well, extrapolating it out. They were looking at what limited information they had about the things, the forces, the inventions that would shape the future, and they were trying to envision what would actually happen, and they’d get it a little right, and they’d get it a lot wrong.

And that is more or less the best that we can hope for today as we have all these debates about how social media, or Web3, or shifting cultural norms, or whatever, are going to shape the future. We don’t really know. Point is that it’s pretty fascinating looking back because it really is a great reminder, one, of how people try to understand how their world will lead to the next world, and therefore, the limitations that we have in trying to anticipate the future ourselves. So I went back to the year 1923 to see what they were predicting about the year 2023, and this year, I have to say was different from all other years that I have done this because in the past as I’ve looked at newspapers in 1920 to see what 2020’s going to be like, and 1921, and 1922, and so on, there’s always a real mix, like I said, experts and school children, just tons and tons of predictions.

But in 1923, there was one, one prediction, one prediction that absolutely dominated the conversation, one prediction from one man that led to a national debate, and that prediction was this. In the year 2023, electricity will have made the world so efficient and so prosperous that people will only need to work four hours a day.

Audio Clip (Voice Actress): And while he didn’t come right out and say so, he left us to infer that ice cream and pie would be served with each meal and everybody would have an iron doe on the front lawn.

Jason Feifer: That sums up a lot of the reaction to this prediction, which was that it was just so ridiculous and fanciful. That particular line was written by a syndicated columnist who ran in newspapers nationwide named Joseph Van Raalte. And in that same column back in November of 1923, Joseph Van Raalte also captured another very common reaction.

Audio Clip (Voice Actress): While there’s a poor farm or a jail dotting the landscape, why talk of the wonders that science is going to perform in the year 2023?

Jason Feifer: In other words, we got a lot of problems right now, so it’s pretty cold comfort to hear that 100 years from now, when everybody today is dead, we will have figured things out. Why don’t we instead focus on how to solve our problems right now? And those two comments basically sum up how a lot of people felt about this prediction back in 1923. It was either, that sounds ridiculous, or meh, maybe that’s true, but it doesn’t help us now. But those weren’t the only reactions. This prediction from 1923 sparked a huge debate about what it means to work and what it means to be human, and how much technology can really change us. And that is an incredibly relevant conversation because we’re having it again right now.

I mean think about it, back then in 1923, Charles Steinmetz saw the rise of electricity as a thing that would ultimately replace a lot of the work that we do. And he thought that would be a good thing because it would just free us up to have more leisure time in our lives. Today, we see AI and automation as again something that is going to take work away from us, but we see that as a bad thing because we imagine it will just mean that there’s less work to go around, and that therefore, people won’t have enough money to live. And what’s going to happen? Well, we already know humans are pretty bad at predicting the future, but I wondered: Is there something about what happened in the last 100 years that can inform what happens in the next 100? Because here we are, it’s been 100 years since Charles Steinmetz died, and we know because are alive right now, we know what happened. We know how his predictions turned out. But here’s the thing, it’s actually a little more complicated than it seems.

Jason Crawford: So the first thing I think you have to understand I think is that it actually did happen, it just didn’t go as far as they predicted in certain dimensions.

Jason Feifer: This is Jason Crawford. He writes this blog that’s amazing called The Roots of Progress. It explores the history of technology and how progress happens.

Jason Crawford: We do work less than we used to, and work is more comfortable than it used to be.

Jason Feifer: And the very idea of work, how we understand work today is completely different than what it was 100 years ago. By looking back at what Charles got right and what he got wrong, and why the world didn’t quite evolve the way in which he predicted it would, can tell us perhaps a lot about exactly what is coming next, so that is what we are going to do on this episode of Build for Tomorrow. We’re taking a glimpse into our future by looking at, well, someone else’s future, the future of 100 years ago, which just happens to be the world we live in right now, coming up after the break.

All right, we’re back. So we’re talking here about a prediction from 1923 that said, “In 100 years,” aka our time right now, “people would be working a four-hour workday.” And although that never came true, it’s kind of funny that today the idea of a shortened workweek, and for that matter, a shortened workweek involving the number four is very much in the zeitgeist. I mean, Tim Ferriss created a whole empire off of a book called The Four Hour Workweek. And lots of companies these days are transitioning to a four-day workweek. And if you rewind 100 years back to 1923, well, the idea of a shortened workweek didn’t come from nowhere, it was also in the zeitgeist.

Jason Crawford: I picked up the book, I think it’s not very well-known today, called Economy of Abundance.

Jason Feifer: Not an Oprah bestseller.

Jason Crawford: Well, and it’s published in I think the ’30s. Stuart Chase is the author.

Jason Feifer: Fun fact, Stuart Chase was an economist who was credited with coining the term new deal. And Jason had a copy of his book in hand that he was flipping through.

Jason Crawford: The beginning of chapter two, he’s quoting a whole bunch of different people who are all making these similar predictions. In the first paragraphs of this chapter, he quotes Steinmetz, or refers to Steinmetz, says, “Who saw a two-hour working day on the horizon.” He quotes, let’s see, Fred Henderson in a book Economic Consequences of Power Production, said, “It would not be a question of an eight-hour day or a six-day week, but more probably of a six months working year,” so I guess we would only have to work half the time, “which,” says Henderson, “Is already the rule for university dons.”

Jason Feifer: And then it just keeps going on like this. You got somebody saying that it will be a one-day workweek, maybe even just one hour of work on that day. You got people saying it’ll be a 24-hour workweek.

Jason Crawford: And this goes on for a few more pages. I won’t bore you with all of the stuff.

Jason Feifer: But the intriguing question is: Why? What was happening at the time that was driving all these people to make these bold predictions about how little we’d be working?

Jason Crawford: A lot of progressive social reformers didn’t like the way that the first industrial revolution had played out because it had given us these grimy cities and all this coal smoke, and people stuffed together in dense population centers. And so they really thought with electricity, “Wow, we have a chance to remake the world a second time, and to do it right this time because we’ll have social considerations.”

Jason Feifer: Which is crazy because that’s not that dissimilar to the way that we talk now. Back then, they were talking about electricity changing everything, and now we’re talking about digital tools changing everything, that because we can all connect by video chat and because we have all these asynchronous platforms that we don’t need to be in person together anymore, which means that we can work wherever we want, and maybe even whenever we want, and also that we can make work more efficient so we don’t have to work as often as we used to. And therefore, we can address things like burnout. And did you catch that word I just used there, efficiency? Efficiency is a big part of our solution today. Digital tools make work more efficient.

Efficiency can therefore mean we work fewer hours and accomplish the same amount. When a company moves to a four-day workweek and says it has maintained the same level of productivity that it had during a five-day workweek, that is because of efficiency. But also of course, efficiency can cut the other way. It can produce, well, cheap products but at a human cost. The reason Amazon is able to offer low prices on products and deliver something to your door within a day is because its warehouses are built for maximum efficiency, which means that its warehouse workers must move around like robots with every step tracked for maximum efficiency. And wouldn’t you know it, 100 years ago as they were imagining how work would transform for the better.

Jason Crawford: This was an era that was really obsessed with efficiency and especially the efficiency that could come from better organization and management, because they had just begin to glimpse this. So think about Henry Ford and the assembly line. Right? That was a system where a better organization of the factory led to this high efficiency, high labor productivity, and low cost and higher wages. There was also this guy, who you may have heard of, Frederick W. Taylor, who was a quote, unquote, efficiency expert. And he would go into factories and he would do what are known as time and motion studies, and he would do a lot of other things. And he would sort of study labor, and he would give this advice on how to get more work out of people, which the laborers didn’t particularly like, but hey.

Jason Feifer: Pretty interesting. Right? I mean, here we have two eras 100 years apart, the 1920s and our decade now, and we are both witnessing a technological change to how we work and an obsession with efficiency, and we’re predicting massive changes as a result, which makes me wonder. Well, okay, why didn’t things turn out the way they predicted in the 1920s? Because I don’t know about you, but I’m not working a four-hour workday despite living in the future compared to 1923. Now earlier, you heard Jason Crawford say that work became easier than it was in 1923, so in a way, what they predicted kind of came true. And that is true, but it doesn’t account for why we still work the number of hours we do. And Jason had a couple really interesting ways of understanding this, so let’s take them one by one. First, he said it’s not really meaningful to track change based on the number of hours someone works in a day. Instead, try looking at the number of hours they work in a lifetime.

Jason Crawford: It was not uncommon to have something like 80-hour workweeks in the past, to have six-day workweeks, and are 12 hours a day, especially for factory workers. This was a huge issue in the ’30s, and the two-day weekend, by the way, was not always a standard. It used to be quite common for people to work day, day and a half on weekends. Having paid vacation is a thing that wasn’t always a benefit that people got. By the way, we not only reduced working hours in those ways, but also in terms of sort of working years. So we eliminated child labor, so we freed up children to no longer have to work, and they can go to school instead. We also introduced retirement, so retirement didn’t used to be a thing. You pretty much just worked until you dropped, or worked until you died, or you worked until you couldn’t.

Jason Feifer: The result of this is borne out in some pretty stark numbers, so check this out. According to the economic historians, Michael Huberman and Chris Minns, back in 1913, the average American worker worked 2,900 hours a year. That averages out to be about eight hours a day, seven days a week. In 1929, the average American worker was working 2,316 hours, so nearly a reduction of 600 hours. And in 2017, which was the latest data available, the average American worker was working 1,757 hours, so that is nearly, well, let’s do the math here, it is exactly 1,143 fewer hours per year that the American worker is working compared to 1913. Now let’s look at some other big changes.

Jason Crawford: Here’s the other key thing. Lifetimes themselves increased, so it used to be that the retirement age was close to the life expectancy. Now what’s happened is we’re living longer, and so you actually now can expect to have 10, 20, or more years after you retire.

Jason Feifer: Which is to say that now the average number of working hours per lifetime is going down because we’re adding non work years to a life. And what does that look like? Well, I found a 1995 study in the journal, Technological Forecasting and Social Change, which found that the total life hours worked shrank for the average British worker from 124,000 hours in 1856 to 69,000 hours in 1981. So that’s a big reduction in the number of working hours in a lifetime, and it’s being spread across more years in a lifetime because people are living longer, which now leads me to a different question I guess, which is, okay, when people in 1923 predicted that we’d be working less, they were actually pretty correct. But why are we then still working so hard? I mean, fine, academically speaking, we’re working fewer hours than we did 100 years ago, but the average worker today sure isn’t feeling that. We still work a lot. Our days are long and full, and often very hard. So why didn’t that change? Well, intriguingly, I think that we can start to answer that question by going back to 1923 and looking at the counter predictions, all the people who said back then, “No, no, no. This is impossible. People will never be working a four-hour workday.” For example, here’s what someone wrote in the New York World in 1923.

Audio Clip (Voice Actress): To meet the requirements of a generation to come with its workday halved and its playtime doubled, we’ll necessitate the invention of new forms of recreation to occupy the leisure of those who have no taste for study and self improvement, if it is not to pall upon them.

Jason Feifer: Now I think that person in 1923 was writing that kind of sarcastically like, “Well, if people are going to have a lot more time on their hands, someone’s going to have to come up with something for them to do.” But of course, that’s kind of exactly what happened, but with an unexpected result.

Jason Crawford: To my mind, that’s probably the most important factor here. I think all these predictions of 10, 20 hour workweeks, what they kind of got wrong was what economists called elasticity. So if your labor productivity doubles, you could do anything from work half as much to earn the same amount of money, you could work the same amount to earn twice as much money, or anything in between. Right? You could decide to work three quarters as much and earn one and a half times as much money, and get a little more leisure and a little more money both at the same time. So I think they failed to predict all of the stuff that you would want to spend more money on.

Jason Feifer: There are of course infinite new forms of leisure, all sorts of ways that we can treat ourselves, and enjoy, and entertain ourselves and travel, and whatever. And then there’s all the stuff that we can just buy for our homes that makes our lives easier, refrigerators and vacuum cleaners and all that stuff. It’s an interesting failure of imagination in a way because Charles Steinmetz in predicting that we would be working a four-hour workday back in 1923 was sort of imagining that parts of our world would radically change, all this new innovation. But then parts of our lives would stay totally static, that we would have the same expectations for our quality of life as we did back in 1923. I mean, if we wanted to maintain the exact same standard of living that we had in 1923 today, well then yeah, we wouldn’t need to earn nearly as much money as we do. We probably could work four hours a day. But of course, that’s not what we want.

Jason Crawford: Yeah. I agree that it’s a failure of imagination. It reminds me of another failure of imagination that was pointed out in the book, Where is My Flying Car? by J. Storrs Hall. He points out that the futurists of the, I guess not 1920s, but just a decade or two earlier, say very early 20th century. They predicted all kinds of amazing technology, and one thing that they missed was the automobile. Now they definitely envisioned progress in transportation, but all the forms of the transportation that they envisioned were public, were mass transit like, frankly, what most transit had been up until that time, if you think of large ships and locomotives and so forth. Right? And what they missed was the need and the desire for a personal vehicle.

Jason Feifer: Isn’t that interesting? Because their world was one in which transportation happened collectively. They imagined innovations in which transportation would just become better at collective transportation. But they couldn’t imagine just a total shift in the way somebody would use this thing. That’s kind of the failure of imagination that we often get boxed into when we think about the future, is that some things will change, but we can’t quite imagine how all things will change. Anyway, before we move on to what the last 100 years can teach us about the next 100 years, I want to look at one more form of complain from 1923. This was one other way in which people of the time said the idea of a four-hour workday is ridiculous, and this one comes from The Wooster Telegram.

Audio Clip (Voice Actress): The spirit of competition will have to go before the four-hour day comes.

Jason Feifer: And here is a similar concern from The Indianapolis News.

Audio Clip (Voice Actress): Where Dr. Steinmetz errs is in believing that the average person regards the work by which he lives as unpleasant. There are hundreds of thousands of persons who are much interested in what they do.

Jason Feifer: All of which is to say that people were saying, “Wait a second. People don’t want to be liberated from work. That would be kind of a hell.”

Jason Crawford: Yeah, I think that’s absolutely true. I think there is a psychological need to have something to do. People need a goal. They need a project. They need something to occupy them beyond just leisure. We don’t need infinite leisure. We need a way to exercise our faculties, our creativity, our ingenuity, our whatever it is that makes us come alive and feel alive. I think it was Paul Romer, economist Paul Romer, who pointed out you look at Burning Man. And what do people do? They’ve got a week off, and they get together in the desert? What do they do? They work. They create projects. Right? They’re building things. And if you look at what people do in retirement as well, I think it seems like it’s common advice that one of the things to have a good retirement is to get busy. Pick something to do. Don’t just sit around.

Jason Feifer: It’s almost as if the chief goal isn’t to not work, but rather to choose the work you do.

Jason Crawford: Yeah.

Jason Feifer: After that conversation, something occurred to me. This whole time, we’ve been talking about how to measure and define work. Is it work hours per day or work hours per lifetime? Is work easier or harder than it was before? But what if we also need to talk about how to measure and define leisure? Because that was the other half of the prediction. Remember? The prediction from 1923 was that by the year 2023, people would work four hours a day and then enjoy leisure the rest of their time. And I started to think I’m one of the lucky ones.

I have mostly chosen what to do for a living, and I really enjoy it, which is a luxury. Maybe for people like me, work is a kind of luxury, a kind of leisure. Maybe we are fulfilling exactly that prediction from 1923 because, I don’t know, maybe if I tallied it all up, I’m doing four hours a day of work, work I don’t want to do, of which there is plenty, believe me. And the rest of it, well, it’s my version of luxury. Maybe Charles Steinmetz was right all along, that is, at least for people on one side of the equation.

David Autor: Computers were invented in the 1950s, or even the 1940s. But when they became cheap and prevalent was in 1980s, and they have dramatically changed work, and in non uniform ways. Not everyone has benefited.

Jason Feifer: This is David.

David Autor: David Autor, I’m a Ford Professor of Economics at MIT. I’m a labor economist.

Jason Feifer: And when I brought this whole question to David, and I showed him these predictions from 1923, and I asked what he made of it, he said, “Well, yes, all of the explanations that you have just heard throughout this episode are very valid explanations for why the world didn’t quite turn out the way that Charles Steinmetz predicted.” But if we’re going to draw lessons about what happened in the last 100 years and apply them to what could happen in the next 100 years, we have to take a very serious look at one of the consequences of electricity, which was of course the development of the computer and the way that the computer radically altered work for everyone.

David Autor: So if you do creative, analytical, interpersonal work, having access to information processing and things that carry out all your routine tasks is terrific. If you are a production worker, or clerical worker, or administrative worker, a lot of that work has been automated. And it’s not been beneficial for people without college degrees. Many of them have been pushed downward into relatively low paid personal services.

Jason Feifer: Now to be clear, David is not all doom and gloom. He actually thinks that AI and automation and new technologies can play a really productive role in creating more opportunity for all, but it’s going to require us to be really, really thoughtful about the development and the use of those technologies. And that, he says, is going to be the big question that defines the next 100 years. So what does that look like? How do we take what just happened from 1923 coming up to 2023, and then apply it to what’s coming next? That is what we will get into after the break.

All right, we’re back. So now it’s time to turn to the future to see what we can learn from the last 100 years and how it can help us anticipate or shape the next 100. And the reason I called David Autor of MIT is because, well, a few years ago I heard him on the Planet Money Podcast talking about a concept that feels very relevant to this question. It is a concept called the lump of labor fallacy. And if that sounds familiar, it might be because I also referenced it a few episodes on this podcast using audio from David when he was on Planet Money. But now I’ve got him myself. The lump of labor fallacy is the belief that there is a fixed amount of work to be done. There is a limited amount of work to be shared among all available workers. And if that’s the case, then when a machine does the work that a human once did, or if you add more workers because of something like immigration, then you’re just putting people out of work permanently because now there are too many people for the fixed amount of work.

But this isn’t true for a couple reasons, among them, when one industry shrinks, another grows. In 1900 for example, 41% of the US workforce worked in agriculture, but then machines automated a lot of that work, and a century later, only 2% of people worked in agriculture. But here’s the thing, total employment in the US did not go down. Why? Because those people didn’t lose their jobs on the farm and then never find another job. The automation instead freed people up to do other work, work that was created in part by the automation. Now there was, for example, a growing farm equipment industry, and also new industries that served these workers who maybe now made more money in less time, and wanted to enjoy themselves with that free time, and therefore were interested in new forms of leisure, like we talked about earlier.

And this is one of the reasons why when you look across time you see that technology had driven a growth in total income and a rising standard of living, even as it has repeatedly replaced people’s jobs. So that’s why I wanted to talk to David, because he has a good nuanced understanding of how these shifts impact people. And when we started talking, he said, “You need to appreciate the context in which all of this change happens because it’s not a straightforward story.” Yes, national income has risen significantly in the last 100 years, thanks to electricity and everything that followed.

David Autor: Well, it’s a question of distribution. They do raise national income. Almost all of these technology raise national income. They allow us to produce more output per hour. But there’s been a really strong decoupling between the growth rate of productivity and the growth rate of the median earner. And so total earnings have gone up a lot, but they’ve been so upwardly distributed that a lot of people have not benefited much or at all. And that’s the concern about technology, or at least the way we use technology. There’s no question it will make us in net wealthier, but it doesn’t mean it’ll make everybody in the country wealthier.

Jason Feifer: And this gets us to what you heard David say just a few minutes ago, that computers in automating tasks that were once done by hand made some workers more efficient, and made other workers obsolete. This meant that some workers moved up the economic ladder and others, typically lower educated workers, were pushed downward into low paying service jobs.

David Autor: And so that’s the real concern about technology and work. It’s not that we should worry per se about the jobs being done by machines. It’s great when machines reduce our burden. It’s questionable. How are many of us going to have a claim on resources and income if our labor is no longer valuable?

Jason Feifer: And how do you solve this? Well, some people have said the solution is redistribution of wealth. If some people can earn, then those people should be distributing some of that earning to those who cannot. But David said that is just not realistic because people don’t like redistribution.

David Autor: The people who are being taxed don’t like it. Most people don’t like receiving it. Right? Most people would rather feel like they earned a living based on their own contributions.

Jason Feifer: So what do we do instead? Well, one of the other arguments would be we have to stop the technology, or at least put severe limits on the technology because now the technology is poised to harm us. But David says, “No, that’s also not the right way to look at it.”

David Autor: The technology is not developing itself, at least not yet. We are the masters of that, and we can focus on how we want to use it, better or worse. And there are many examples where technologies that we invent are highly complementary to us, and make us better at what we do, increase the value of the skills that we have. There are many other examples where it does the opposite, where it just replaces us. And so the question is whether we can use it better to get