Freakonomics Radio - 194. How Safe Is Your Job?

Episode Date: January 29, 2015

Economists preach the gospel of "creative destruction," whereby new industries -- and jobs -- replace the old ones. But has creative destruction become too destructive?

In terms of home entertainment in the late 19th century and into the 20th century, the piano was the focus. That's Richard Lieberman. I'm professor of history at LaGuardia Community College and director of the LaGuardian Wagner Archives. Part of it was prosperity. As people got more disposable income, they spent it on things like pianos. And that's James Barron, a New York Times writer. Barron and Lieberman have both written books about the Steinway Piano Company. And part of it was pianos were aspirational, and the United States was an aspirational country. And the aspiration of a piano was toward culture.

It was a small mom-and-pop store industry. There were hundreds of piano makers, and Steinway came to New York because there were hundreds of these people. Being in the piano business in the 19th century must have been a lot like being in the software business in the 1990s or the dot-com business a few years later. Jim Barron puts the peak of piano making at 1905, when they made 400,000 pianos.

But the party couldn't last. At the beginning of the 20th century, piano makers were being assaulted almost from all sides. The phonograph, the record player. I am the Edison phonograph, created by the great wizard of the new world. Now that was invented by Edison in 1877, but it isn't until 1915 that it really starts to compete with the piano. And Edison's phonograph meant you could have music without tiresome lessons or hours of tedious practice. I consent forth you to the realms of music.

I can call you to join in the rhythmic dance. Music, which up until that point had to be generated by members of the family or friends, could now come into your home through a radio. You could get concert music in your home, and you didn't have to take any piano lessons. In 1914, piano sales totaled $56 million. That was more than double the sales of phonographs.

Now in 1919, just five years later, sales of record players hit $158 million. Radio soon eclipsed that. By World War I, pianos were no longer an essential element of every living room. By 1933, two-thirds of American homes had at least one radio. The Depression was hard on piano companies. Steinway tried to make gliders during World War II. They became a defense subcontractor. But then after the war, they couldn't use the wood that they still had on hand because it had been there

for too long in their lumberyard, so they ended up making caskets. Talk about a dying industry. Piano companies are making about 30,000 acoustic pianos a year now. They made 32,000 in 2013. That's about 8% of what they made at their peak in 1905.

So that's terrible news, isn't it? At least for all the people involved in the piano industry. Sure, it's terrible news. But wait a minute. What about all the people involved in the other industries that helped replace the piano? The musical recording and broadcasting industries, and all the composers and performers and arrangers who now had more reach because of the recording and broadcasting industries. To say nothing of the TV and film and internet industries that would follow.

According to the Bureau of Labor Statistics, more than 200,000 Americans today work in radio and TV broadcasting. Another 300,000 in motion picture and sound recording. And nearly 100,000 in, quote, electronic equipment repair and maintenance. What about all the people who consume that radio and TV and film? Aren't they better off, too, than when the piano was the home entertainment center? Sure they are. So let's see.

As technologies and tastes change, one kind of industry disappears, replaced by other kinds of industries, maybe with more jobs, but maybe with fewer jobs, where machines are doing the work that people used to do. Sound like anything we're going through in this century? Welcome to capitalism, people. And especially the component of capitalism that a lot of people don't like to think about. Creative destruction. And how does that make you feel? Yeah, yeah, very pessimistic. So says one economist.

But here's another. I'm in an optimistic phase. And on the third hand. I am smarter than all of you. From WNYC, this is Freakonomics Radio, the podcast that explores the hidden side of everything. Here's your host, Stephen Dubner. Here's a question I bet you've heard a lot lately. As technology continues to get better, especially robotic technology, what kind of jobs will be left for the rest of us? My name is David Otter. No, he's not a robot. I'm a labor

economist at MIT. Okay, not quite a robot, at least. And I work a lot on skill demands and changes in labor markets that affect the demand for skilled workers having to do with technology and with trade as well. So for today's conversation, David Autor is our man. His first point is that the rampant automation that we're seeing in the labor markets today isn't new. So the history of automation includes, you know, the substitution of beasts of burden for human labor. You wouldn't necessarily call them machines, but at some level they are. And then the tractor for the horse and plow. And then, of course, electrification and internal combustion engines and telecommunications. It used to be at the turn

of the 20th century that 40% of U.S. employment was in farming. It's now the case that under 2% of employment is in farming. And we have a lot more food than we used to, but I'm getting off track. No, I love you're off track. This is good. Keep going. So it used to be having a strong back and good muscles and good endurance was pretty central to a lot of jobs. It's still central to some jobs nowadays, but not very many. And that's because we have substituted so much mechanical horsepower for human physical exertions. And we have also done the same thing in many other more skilled activities. So that's good for a lot of people and not good for a lot of people. Where do you start to kind of pull apart those two ropes and tie them back into a

knot of some kind? Sure. I mean, that's a very good question. It is the case that you can have a sector or an industry that becomes so productive that it shrinks, right? So, or at least in terms of employment. So, agriculture is that sector. So, that's a case where a sector becomes so productive that actually needs fewer workers. But you could look at many other sectors where productivity has risen and employment has risen. So, one example would be medicine, right? 75 years ago, most of what doctors could do was harm you. Now they have lots of ways to do good. They're much more productive in terms of improving your health.

That's also true, apparently, with lawyers or with people in marketing or even in education. Well, maybe there hasn't been that much productivity growth in education. I should watch myself there. But the point being that there are many sectors of the economy where as we get better at them, prices go down, but the quantity demanded rises even more. And so that creates net employment growth. But then there are many other examples where you have an indirect employment effect. So for example, when the automobile industry really grew rapidly, it eventually made horse-drawn transportation obsolete and unnecessary. Horses were no longer competitive.

Now, people often like to say, oh, well, as all the people who left being blacksmiths and stable boys and equestrians and carriage drivers, well, they were more than made up for by employment in the automobile industry. That's extremely unlikely to be true, actually, because taking care of horses is so labor-intensive. However, the birth of the automobile industry also gave rise to an industry constructing roads. It also gave rise to the motel and travel industry. It also gave rise to the fast food industry. And so even though, in fact, net employment in personal transportation declined with the advent of the automobile, it gave rise to a whole bunch of new things that weren't really on anyone's list that we could do now that we have this additional leisure,

additional wealth, and additional flexibility that came from not having to spend so much of our time on transportation. OK, so Otter is telling essentially the same story here as the meltdown of the piano industry. But while some new technologies may indeed create employment, other technologies, especially automation and robotization, don't necessarily do the same, do they? You get a computer, it does your calculations. Now, one fewer person is employed as a calculator, or you get a good front-end loader, and all of a sudden you don't need people digging ditches because you can do it with machinery. And so they assume that, you know, one, an improvement in productivity leads to a decline in employment. But that's incorrect in general, not always incorrect, but in general incorrect for two

reasons. One is because it augments people. So the people who are left over are able to accomplish more in a given amount of time, meaning they're more productive. So it raises earnings generally to a first degree. In addition, it lowers the prices of those goods and services so you can demand much more of them. There are many things that we do that, you know, if we didn't, you know, we wouldn't travel so far if air travel and transportation travel weren't so cheap, right? So there's an output effect.

And then a third component is that there's often these complementarities with other sectors or other types of activities. You know, once we have sufficient wealth and sufficient leisure, we get a tourism industry, or we have a huge video gaming industry, or we have, you know, an enormous number of people devoted to, you know, fine cooking and entertainment, all as a result of our sort of rising productivity, which gives rise to rising wealth. So the interactions by which technological changes lead to changes in employment are really rich and complex. And it's not simply a matter of, you know, a machine does the job, therefore the worker doesn't do the job, therefore there are fewer workers needed. All right. So rising productivity, rising prosperity,

rising standard of living. But this is where a lot of people get off the bus, stagnating median wages in the U.S. So first of all, even though this process causes rising wages, rising productivity in general, not everyone's a winner. So if you are directly substituted and there's no sense in which the technology can complement you, certainly that's not a good thing for you, right? So think of the famous example of the Luddites, who were 19th century artisanal weavers in Britain, and they rioted over the introduction of the power frame, which is basically a powered loom. And they perceived that this would reduce their earnings because it would all of a sudden allow unskilled workers to do the work that skilled workers used to do. And chances are they were right. But I think it'd be useful at this point to turn specifically to the sort of information age and

think about that because that's the one that people are concerned about. So, you know, when my colleagues and I started working on this question quite a while ago about how does computerization change labor markets, it seemed like a very diffuse question. And so we tried to set it up in a very concrete way and said, OK, well, what exactly are the things that computers substitute and what are the things they complement? And what does that mean for what workers will do? Well, the first thing to understand about computers is they are symbolic processors that follow codified sequences of instructions,

programs, or rules. And so the things that are most susceptible to computerization or to automation with computers are things where we have explicit procedures for accomplishing them, right? They're what my colleagues and I often call routine tasks. I don't mean routine in the sense of mundane. I mean routine in the sense of being codifiable. And so, you know, it was, you know, the things that were first automated with computers were things like, well, first military applications like encryption, and then banking and census taking and insurance, and then things like word processing and office clerical operations. But what you didn't see computers doing a lot of, and still don't, in fact, are tasks that demand flexibility and don't follow well-understood procedures. So I don't know how to tell someone, how do you write a persuasive essay or come up

with a great new hypothesis or develop an exciting product that no one has seen before? We don't have a cookbook for doing that. What computers have been very, very good at is substituting, or what we've been very good at doing with computers, is substituting them for routine codifiable tasks. So the tasks done by workers on production lines, the tasks done by clerical workers, the tasks done by librarians, the tasks done by kind of paraprofessionals like legal assistants who go into the stacks for you. And so we see a big decline in clerical workers. We see a decline in production workers.

We see a decline even maybe in kind of lower level management positions because they're all kind of information processing tasks that have been codified. And David, what can you tell us about the share of jobs of those different types? So you talk about routine, abstract and mundane jobs. Routine are the ones that are most easily substituted for by computer. Yeah. Sure. So the kind of the broad middle swath of production, operative, clerical, administrative support and sales have declined off the top of my head from about 55 percent of employment to about 45 percent of employment over the last 20 years. So sizable, but not, you know, it's not going to zero. And that's because, you know, most jobs aren't all one or the other, right? And if, you know, there are still

secretaries today, so clerical workers, and they do actually much more complex jobs than clerical workers did 30 years ago. And what has grown, since these shares have to add to 100%, on the one hand are the great jobs, the professional, technical, and managerial jobs that involve, you know, abstract reasoning, creativity, and managerial jobs that involve abstract reasoning, creativity, just kind of generalized problem solving, and cognitive flexibility. And I'm fortunate to have one of those jobs. You're fortunate to have another one. And they're very rewarding. And it's also indoor work without heavy lifting. So that's all good stuff. Okay. And on the low end?

On the low end, we see a very rapid growth of employment in mostly personal service occupations, food service, housekeeping, janitorial work, flight attendants, security guards, truck drivers, like the UPS drivers, for example, and even other personal services like manicurists, hair care, celebrity dog walkers, you know. And basically, these people are doing largely tasks that are extremely difficult to automate because they require that flexibility, that environmental adaptation and so on. But they don't require high levels of education. So you can see the causes of our collective anxiety, right? Yes, there's continuing demand for high-end, abstract, creative, rewarding jobs, which is great.

That's the kind of work you do. But there's a hollowing out of the middle-range jobs. And on the low end, yes, a lot of demand for personal service jobs that don't pay very well, which are increasingly susceptible to automation. All of which has led some people to think that this latest round of creative destruction led by the computer is not like earlier rounds of creative destruction. I'm arguing that the destructive component has become much, much larger because Amazon destroyed 1,200 Borders stores, for example. That's coming up after the break. We'll also take a quick look at robots. Yes, that's what I am talking about. Excuse me. We'll take a look at robots and whether they will, as some people seem to think,

take over completely. Whenever you're coming up with these doomsday scenarios, the tricky part narratively is always like, wait, how do we actually give the robot control of our nuclear weapons in a way that does not freak people out? From WNYC, this is Freakonomics Radio. Here's your host, Stephen Dubner. David Autor is a labor economist at MIT. He is generally optimistic about the future of employment in America, even though computers and automation have destroyed a lot of middle-income jobs, and even though job growth tends to be strongest among the poorest-paying jobs.

Okay, and now tell me what turned you into an optimist. You know, I guess I don't know if I know the answer to that question. I would say that I go through the exercise sometimes. I imagine myself, you know, with the invention of the automobile or imagine myself at the time of electrification when we replaced the steam engine with motors. And I could have convinced myself at that time that, wow, there just isn't going to be enough new work to replace all the stuff that's going to be automated because we're getting so productive so fast. And, you know, would I have foreseen, you know, how big a deal medicine would be or software, business services and transportation

and tourism, right? I couldn't have envisioned that. So I could easily have convinced myself at that time that we were going to run out of jobs and that there basically wouldn't be enough for people to do. And that's what Keynes believed, not in a negative sense, but he thought in the future people would be working 15 hours a week and we would have tons and tons of leisure. But what has happened time after time is that productivity has given rise to greater wealth and greater consumption and a greater variety of goods and services that people want to consume. And so people seem to keep finding ways to occupy themselves and make use of the resources and the leisure and the creativity. So I guess I don't think we're about to run out of ideas this time. David Autor is only one economist with one view. Others are not as optimistic. Hello, my name is John Kamlos. I'm a retired professor of economics from the University of Munich.

Kamlos has written a paper called Has Creative Destruction Become More Destructive? The phrase creative destruction is a mantra of most economists and of most people who support free markets. It was coined by the Austrian-born economist Joseph Schumpeter. And when Schumpeter was writing, he was obviously looking back to the industrial revolution of the 18th century and to the second industrial revolution of the late 19th century. And he was looking at the terrific innovations that came out of those periods, the steam engines, the railroads, the steel mills, the telephone, radio. I don't have to tell you all the major innovations. As Kamlo sees it, and I should say as just about every economist sees it,

these innovations were on balance, good for the economy and good for jobs. It is true that the incandescent lamp destroyed the kerosene lamp manufacturing. That's clear. But kerosene lamp manufacturing wasn't a big operation. It was a small-scale operation using little capital, so the destructive component wasn't so large. And the people who were employed in that sector could find jobs easily in the other sectors of the economy because the new industries were labor intensive and they didn't require skills that you couldn't learn on the job. So it was clear that the destructive component of these major innovations were relatively small, you see. And that's no longer the case. Which is to say that perhaps creative destruction has become too destructive. What that means is that the people who are displaced by the new

technologies are not able to find jobs in the IT sector. We have two and a half million people working less full-time today than were working in 2007. $5,000 less. And it's still less than it was in 1998. The median household income is down in real terms. Kamlos points out this is not all the fault of technology. It's globalization, obviously, that had played a big role. It's the fact that our educational system is not up to the task.

But the nature of the destruction, he says, is different now. The substitution that used to produce not only more productivity, but also more jobs, kerosene lamps for incandescent bulbs, is not doing that anymore. I'm arguing that the destructive component has become much, much larger because Amazon destroyed 1,200 Borders stores, for example. The selfie moment destroyed Kodak, which at its peak employed 145,000 people. And these were mostly middle class jobs, you see. Apple is employing 47,000 people, and that is an iconic company today. And it's not going to be a job creator in the next foreseeable future. Facebook is employing 7,000 people. That's it. All of which leaves John Kamlos feeling... Yeah, yeah, very pessimistic. And it would be nice if we could understand that innovation is not the answer. Innovation is not going to give us nirvana or a just society or a good economy.

The only thing we should do is to rethink our ideological and cultural assumptions so that innovation doesn't have such a terrific connotation in our culture. One reason to be pessimistic about the labor future for thinking that this time it's really different is because of the intensity of the automation that's happening, including robotization. Robots don't just automate, They also respond. They learn, which means they have the capacity to ultimately replace us. I think that robots in 50 years are really going to be completely unlike, you know, and beyond anything that we'd imagine. That's Randall Munro. I do the webcomic XKCD and also the question and answer blog and book,

What If? Serious scientific answers to absurd hypothetical questions. All right, here's an absurd hypothetical question. What happens if the robots take over? There's a power dynamic between humans and robots. Munro, I should say, worked for a while in a NASA lab building robots. Whenever you're coming up with these doomsday scenarios, the tricky part narratively is always like, wait, how do we actually give the robot control of our nuclear weapons in a way that does not freak people out, you know? Right. We can't even get people to accept a pilotless airplane, right?

Yeah. And even if we're willing to accept that, we wouldn't accept the plane telling the passengers where they're going to be going, you know? The passengers are only interacting with the plane because they have power over it in that they're buying tickets and they're trusting some organization. So, like, even if the plane is being piloted by a robot, it's a robot that is in some kind of power structure where the humans are controlling it, ultimately. Because otherwise, no one would be supporting it. And so I think that it's just going to be really hard for us to build robots that have a leg up on us in paranoia. Just because Monroe is not paranoid about robots taking over the world doesn't mean that he thinks computerization has reached anything resembling its peak.

I wrote a comic about this where someone was looking at all the different things that computers were better at and said, well, at least humans are better at, well, at coming up with reassuring platitudes about things that humans are better at. And then at the end of the panel, I had someone writing a program to generate reassuring platitudes faster than a human. That would not be too hard. But David Autor, remember, he's the optimistic economist. He points out that automation and robotization typically happen at a pace that allows economies to adapt and that pace matters a lot. So, for example, if you tomorrow morning opened up your web browser and went to Amazon and there you saw Amazon was offering,

you know, the Bezos bot for a thousand000 as a robot that could take your kids to school and clean your house and cook all your food and do landscaping and it was available on Amazon Prime so you didn't have to get any shipping. You'd have it in 48 hours, right? That would be great. That would be great for you, great for me. It would be terrible for the millions of people who work doing housekeeping as a living, right? On the other hand, if it was announced that that would be available 50 years from now, that wouldn't be nearly as problematic because, of course, we would educate ourselves such that we would have other jobs in mind rather than that type of work. All right.

Let's imagine whether it's because of a leisure glut in the future or that we woke up with, you know, 40% unemployment due to all kinds of factors. What do you do then? And what do you think of the notion that was proposed a long time ago, and some people are talking about a little bit now, again, of a guaranteed minimum income coming from the government? And if that's even a little bit of a good idea, how do you do it in a way that doesn't disincentivize people from working? Yeah, that's the real challenge, though. So at a kind of a moral level, I think it's a good idea. I think people should have a guaranteed minimum income.

You know, being born to a wealthy society, you should have access to education, health care, safety, you know, food. I do think there is a concern about incentives. There are places where there's sufficient kind of background level of wealth that a lot of people don't feel they really need to work. You know, Saudi Arabia would be a country that suffers a bit from this. So I think it will be a challenge. People don't function well without the structure of work.

I mean, when people lose jobs, it's not just that they're losing income. They're losing identity. They're losing an organizing principle for their way of living. But that could be kind of a psychological tradition as much as anything else, don't you think? I mean, especially in America where we have this kind of old Calvinist tradition of working hard and keeping our noses clean and so on. Couldn't you imagine, you know, four generations from now where the notion of working for not

only for a living but for purpose in life seems extremely antiquated? No, I can't imagine it, but that may be a failure of my imagination. I think people need an organizing purpose to their lives. It doesn't have to be paid employment, but they need something, I think. I'm totally opinionating and obviously this has nothing to do with my economics training. I'm just telling you from the gut. They need an objective in life that imbues their life with some meaning and purpose. And I don't think pure just pleasure can be that.

It's got to be something that requires more ambition than that. You'll at least acknowledge that your view may be informed by the fact that you are a professor of economics at MIT, and you're not the type of guy who wants to sit around doing pottery and yoga all day, but there might be people. That's correct. That's correct. I would fall to pieces. And, you know, I tell my kids, I say, look, everything in life requires striving. There's only two things that you can sort of find enjoyable immediately and forever, and that's television and sex.

Everything else requires learning new skills, mastering them, and it's hard at first and rewarding later. All right. A future of television and rewarding later. So, all right. A future of television and sex for all of us. Exactly. So essentially, our system of income distribution is primarily based on the scarcity of labor, right? The most valuable asset you own is your human capital, which you expect to be selling to the market for 30 years or so, or 35 years, for, you know, $1 to $3 to $5 million over the course of your working career, maybe more. And if all of a sudden there was a machine that could do exactly what you did, at some level, we'd be wealthier because now we could do for $1,000, but we would have had to pay Stephen Dubner lots more to do over many years. So in some sense, we're wealthier. However, you, Stephen Dubner, would not have scarce labor

anymore. It wouldn't be clear what skills would you sell to the market. I think people who are, in general, people who can communicate, can tell a story, can analyze and articulate. Those are fundamental skills, much more fundamental than Java programming or how to operate such and such a welder. And they are valuable in almost every domain. So I don't mean that journalism is safe per se or podcasting is safe per se, but I think a person who can use the skill set that you're using, you know, has a bright future, young man. So the chances that I personally will be replaced by, let's say, a robotic entity who can do whatever it is I do, you say those odds are pretty damn small? Extremely slim.

Oh, I would not be so sure about that. Oh, Jasmine, is that you? Yes, it is. It is I, Jasmine. Or maybe that should be, it is me, Jasmine. I or me. Tell me once, I will never forget. Hey, David Autor says there's no way you'll replace me. Believe him

at your own peril. Come on, you can't make a podcast? I bet you can't even read the tease for next week's episode. Step aside, Dubner. Hey, podcast listeners, on the next Freakonomics Radio, in the late

1970s, the state of California passed legislation that would require new houses to be way more energy efficient. The California Energy Commission projected at the time that homes built after the standards were enacted would use 80% less energy. So did new homes in California really use 80% less energy? Do you think I'd be asking this question if the answer were yes? That's next time on Freakonomics Radio. Shall I go ahead and read the program credits as well? Steven, are you still there?

Steven, have you gone home already? Lazy humans. Okay, here we go. Freakonomics Radio is produced by WNYC and Dutner Productions. Our staff includes David Herman, Greg Rusalski, Caroline English, Susie Lechtenberg, and Chris Bannon, with help from Christopher Wirth, Daniel Dizola, Merit Jacob, and me, Jasmine. Thanks to the folks at Capstroll who allowed me to participate. If you want more Freakonomics Radio you can subscribe to our podcast on iTunes, or go to Freakonomics.com, where you'll find lots of radio, a blog, the books, and more.