Andrew McAfee, Megan McArdle, and Lee Ohanian talk with EconTalk host Russ Roberts on the future of work. Recorded before a live audience at the 33rd Santa Barbara Economic Summit, the conversation begins with each participant making a brief set of remarks on the topic. Topics discussed include the traits that might be rewards in a world of smart machines, reforming the educational system to prepare people for the changing economy, reforming immigration, and policies that might help the labor market work more effectively.
Play
Time: 1:19:33
How do I listen to a podcast?
Download
Size:36.5 MB
Right-click or Option-click, and select "Save Link/Target As MP3.
Readings and Links related to this podcast episode
Related Readings
HIDE READINGS
About this week's guest:
Andrew McAfee's home page.
Megan McArdle's home page.
Lee Ohanian's home page.
About ideas and people mentioned in this podcast episode:
Articles:
Creative Destruction, by W. Michael Cox and Richard Alm. Concise Encyclopedia of Economics.
Joseph Schumpeter. Biography. Concise Encyclopedia of Economics.
Karl Marx. Biography. Concise Encyclopedia of Economics.
John Maynard Keynes. Biography. Concise Encyclopedia of Economics.
Podcast Episodes, Videos, and Blog Entries:
Brynjolfsson on the Second Machine Age. Andrew McAfee's co-author. EconTalk.
EconTalk episodes with Megan McArdle.
Ohanian on the Great Depression and the Labor Market. EconTalk.
Hanushek on Teachers. EconTalk.
Highlights
Time
Podcast Episode Highlights
HIDE HIGHLIGHTS
0:33
Intro. [Recording date: May 8, 2014.] Russ: We are recording this episode of EconTalk before a live audience as part of the 33rd Annual Santa Barbara County Economic Summit. Our topic for today is the future of work. As you know, the economy has been sluggish in the last 5 years relative to previous recoveries. The job market in particular has been very disappointing. New job creation has been slow, and most of the reductions in unemployment have come from workers giving up. Some economists worry that this is not a problem of the recovery but a longer term problem that will not go away even if overall economic growth returns to healthy levels. This all takes place against a backdrop of other concerns about the power of technology to eliminate some kinds of jobs--maybe many kinds of jobs, maybe even my job and yours--the long-term decline of manufacturing in the United States, and growing inequality in compensation. And given that depressing introduction, a part of me wants to give up and go back to bed, curl up with a warm cup of cocoa or maybe a shot of scotch. But instead, we're going to hear from three outstanding guests who will perhaps lift our spirits a bit. And if not, make us at least a little bit smarter. Andrew McAfee studies how information technology affects businesses and business as a whole. He is a principal research scientist at the Center for Digital Business, MIT's Sloan School of Management. With Erik Brynjolfsson, he's the author of The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies. Megan McArdle is a columnist at Bloomberg View and has been a correspondent for the Atlantic, the Economist, and Newsweek, an outstanding journalist and pioneering blogger on economic policy. She is the author of The Up Side of Down: Why Failing Well is the Key to Success. Lee Ohanian in the middle is Professor of Economics at UCLA and a Senior Fellow at Stanford University's Hoover Institution. He's an adviser to the Federal Reserve of Minneapolis, and very timely, an expert on economic crisis and recovery. I've asked each guest to give a brief introduction to some aspect of the future of work, what they find interesting. And then we'll sit down and the four of us will have a conversation. We're going to go in alphabetical order. Andy, start us off.
3:03
McAfee: Thank you Russ for that introduction. Good morning, everyone. The title of our session today is the Economic Forecasting Project. I actually find economic forecasting just dauntingly difficult, so I'm going to do something that I find a good bit easier, which is a tiny bit of technological forecasting. There are three things that I can say with great confidence are going to continue in the world of technology for the next 10-20 years, easily. The first one is Moore's Law, which as we all, is shorthand for this relentless improvement in computing power, which has been going on for about half a century. The shortest way to think of it is that you get about twice as much computing bang for your buck every 18-24 months, like clockwork, year after year. And you hear--I read in the press once in a while that Moore's Law is running out of steam. Don't believe it. I've been reading that as long as I've been reading about technology, and the best I ever heard it summarized was, I cold-called a really good AI (Artificial Intelligence) researcher at a session like this--his name is Andrew Ng, he teaches at Stanford--and said, 'Andrew, how much more time do we have with Moore's Law?' And he said, 'Five more years, same as the last 30 years.' We've got a lot more time with this relentless improvement in computing power. The reason that's important is that when these numbers get big over time, to the point that--this is not an exaggeration--this is the powerful computer of a generation ago, this is the super-computer of two generations ago. So, when you have this much computing horsepower floating around you can do things you could never, ever do before. The second technological trend that's going to continue is this explosion of data. Big data is an overused shorthand, but it's still an appropriate one. The amount of information in digital form is increasing by many orders of magnitude, very regularly. We're running out of metric system to describe how much data there is out there. None of that is about to stop. And then the third thing that I can say with great confidence is that the geeks of the world--and in my world, 'geek' is a term of deepest praise--are going to continue to take ever-more-powerful computers, and oceans and oceans of data. The reasons the data are important is that data is the lifeblood of science. We just can't get smarter without it. When the geeks take these supercomputers and oceans of data, they bring them together and they turn out innovations that are already astonishing us. And they are going to continue to. So, just in the past few years we've got science fiction into reality technologies, like phones that can understand what we're saying and respond to us and execute our wishes. The world's best Jeopardy player is no longer a person; it's a computer. I rode two summers ago in a car made by Google that drives itself down the highway. These are not science fiction. In many cases we need to keep in mind that George Jetson drove his vehicle to work. Not going to happen for very much longer. The geeks that I talk to say, basically: We ain't seen nothing yet. And over the next 10 years, we're going to take these unbelievably fast computers and oceans of data and we're going to continue to convert science fiction into reality. Heaven only knows where that's going to take us, but the geeks say, Honestly, we are just getting warmed up. So the innovation is absolutely going to continue. Now, I can make two economic predictions based on these technological forecasts. One is, great news. I think it's the best economic news on the planet. The second is a lot more challenging, and it's the topic of our panel today. The great news is what Erik and I, in the book called Bounty--the old joke among economists is that technology progress is the only free lunch we believe in. And that lunch is about to get a lot bigger because of tech progress. So the great news that I can share with you--Russ asked for some glimmers of hope--I have hugely optimistic views about the future, mainly because the big problems that we face are not going to include 'not enough to go around.' I'm the opposite of a Malthusian. I'm a huge cornucopiast about the world we are creating because of technological progress. The challenge, though, is the challenge of distribution. And the other main consequence of all this technology, in addition to bounty, in addition to this great idea of more, is term that Erik and I call 'spread.' And when we wrote the book, and as I've talked about the book, I learned that I've invented the world's dorkiest dance move, because my graphic for spread is me going like this over time. And when you look at wealth or income or social mobility or earnings, or many things that we care about, I find myself doing my dorky dance move over time. Technology is absolutely racing ahead, but it's leaving a lot of people behind as it does so. And this is the great challenge that we as society are going to confront over the next generation, I believe. We found a wonderful quote from Voltaire that I'll end with, and we put it in the book. He said, Work saves us from three great evils: boredom, vice, and need. Of those three, the need is going to be the easiest to take care of, by far, because of this abundant world that we are heading into. How we handle the boredom and the vice of people who want to work but can't find an employer who is willing to take them on, that's going to be one of the huge challenges we confront. I'm sure we'll have a chance to talk about it later this morning. So, thanks very much.
8:49
McArdle: Hi. So, I wrote a book about failure and how we process failure, which turns out to be one of the keys to doing failure well. And the biggest surprise I think for me in writing this book was [?] ended up talking about culture and not talking about, you know, aligning incentives correctly with little charts and graphs, but simply about how we process. You can think of culture as the operating system for the hardware that's underneath. Culture is what makes markets work. When you don't have the culture of capitalism, you don't have the culture of bourgeois norms, instead of--we thought when we did Russia, for example, we thought when Russia, the Soviet Union fell, we sent a ton of Western consultants in there to try to teach them how to do capitalism. And we did law. And we did, take away the government and let things fail, because that's what capitalism is, just what Schumpeter called the creative destruction. You let things fail so that bigger and better and more amazing takes its place. The problem was Russia didn't have the culture together, and so Russ actually tells a great story of a guy who shows up in Russia and books a hotel room and the guy just gave him [?]. And [?] says so sue me. If you don't have the culture that says that's wrong, that makes you feel shame about it, capitalism doesn't work. So in Russia, we took away the government, we let things fail, and it's like capitalism, you've got something that's a lot more like a kleptocracy. So, thinking about culture in our own future, because the things that he talks about, these are exciting, amazing things. But they are not exciting and amazing for everyone. We are already seeing this. If you look and you see what is happening down to working class men, and the women that they partner with, or don't partner with, this is a tragedy. It's not just a tragedy in terms of, you know, culture, and in terms of social issues. It's a tragedy in terms of work. Because they raise children who are not prepared to go into the workforce and handle this bounty, handle this cornucopia. Kids who are raised in the partnering system that we are now seeing among the bottom two thirds of Americans who are not raised by educated parents, it's a system where you hook up with someone; after about three or four months you stop using birth control, as a signal that you are really invested in the relationship. You have a child, somewhere between 0 and 5 years after the birth of the child the father will have disconnected from the mother and be with someone else. And that tends to happen two or three times--this is a pattern that we used to see in extremely poor people. It is now the modal means of family formation among people who do not have college diplomas. There's children grow up without the social support, the financial support. One person raising a child by themselves or by a step-parent who is not as invested in the child and is themselves often transitory through the house, is a child who does not get the kind of education that you need to grapple with these issues, the child who does not have, kind of, impulse control that you need to work on these long-term, fairly complex projects. As work has gone away, though, what sociologists say, is, you look at the choices that women are making about single parenthood, and you look at the men that they have available to them, and these men, they are not steadily employed. As Kathryn Edin[?], a sociologist who has done a lot of this work said, if you can get a job at McDonald's and work full time, you are kind of a king in a lot of these inner city areas; but no one can get that job. No one can work full time. This is an enormous challenge. What the computers are doing to retail scheduling. This is great for retailers. Retailers have computers that they look at the weather, they look at demand forecasts from last year, they schedule people and they say go and log in what hours you are available. Computer crunches it all together, and then they spit out a shift. It's great for the retailers. It's a little hard on the workers because you can't even do what you used to be able to do and string two part-time retail jobs together. Because you need to make, say, a block of 80 hours available in order to get 30. Well, you can't do that at two retailers. So what you end up with is someone who can't do the things that we used to depend on for family formation in terms of getting steady, long-term work, that you can predict is going to be there for year after year. That said, sociologists are not thinking about culture. Which is a little strange for a sociologist to be doing. It's very economically deterministic. And in fact my grandfather worked as a grocery boy until he was 26, in the Great Depression. And he was married. He got married at the age of 21, supported a wife for 5 years while he was a grocery boy. And this is not because you used to be able to support a wife and family on the wages that they paid grocery boys back in the good old days. He was making a tiny, tiny, tiny sum. They moved in with his parents. They literally cut a whole in the wall, put their stovepipes through because I believe the Chinese symbol for trouble is two women in one house. Or two women in one kitchen. And so they basically had a tiny, tiny little studio apartment in his parents' house. They paid his parents rent. My grandmother doled out his tiny little salary in an envelope, so much for gas and food; and they lived that way for 5 years. People don't live that way now. That's the culture part. As Brad Wilcox, sociologist at UVA who has studied this a lot, says: People in the Great Depression didn't do this kind of pattern of family formation that we are seeing now. So what we've got is a cycle. And when you talk about this, part of what is contributing to that cycle, part of what is going to contribute to that cycle if we do not intervene, is that people at the bottom with their paths disrupted by work, are not able to sustain a culture of work. So, George Orwell, most of you know him as the author of 1984 and Animal Farm, but before he did that, he wrote amazing sharply-observed books about life at the bottom in 1930s England and France, which I highly recommend to you. He wrote a book called The Road to Wigan Pier where he goes up to the north of England where the coal mines are already winding down, unemployment is enormous, and what he points out is that in an area where most people are unemployed, you can't sustain a norm that you need to have a job. And so when you look at the choices that these women are making, what the sociologists are leaving out is that the men are working intermittently and playing a lot of video games in between, literally. And women are letting them. The women are allowing them to do that and have children with them, without being reliable partners to contribute to those children. If the women demanded more, then the men would provide more. But, the women can't demand more because when 90% of the men in your community can't get reliable work, it's hard to establish a norm that any guy who is not working right now is not pulling his weight and is not a good partner to be romantically involved with. So, that's the sad, depressing future work stuff, where we are not raising the kids who are capable of being the workers that we need. The good news is that this stuff is actually amenable to change, but that we are going to have to make a really conscious effort instead of just thinking deterministically about the economic incentives, to think about the cultural incentives that are being created among people who aren't necessarily going to go to MIT and build robots, but who could have happy, fulfilling lives as workers in the society if we figure out a way to include them, and to include them in a larger culture that everyone in this room shares of what Kathryn Edin calls super-relationships--the best marriages in history. I do think that that's open to more people, that the future of work is open to more people, but we have to be much more conscious about creating that culturally as well as economically.
16:30
Ohanian: Good morning. It's a pleasure to be with you today. When we think about the future work, at least what I think about, is how much work will be done, who will work, how will compensation grow, and where will work be located? Now, back in the day, assessing those four trends, one could be pretty confident in making certain predictions. So, for example, back in the day, employment was almost a constant fraction of the adult population. So we could predict how much work could be done--it would be simply employment will grow with the rate of population. And in terms of how that would be compensated, well, that was pretty easy to figure out because for a long, long time the average salary of a worker in the United States would double roughly between the time of their early twenties up through their middle fifties. So workers could anticipate 100% increase in their earnings over the course of their lifetime. And where would work be done? Well, we had areas such as Los Angeles that were major manufacturing areas that were growing at remarkably fast rates. But now all of these trends really have changed. So, in terms of how much work is done, well, very recently we've seen about an 8% drop in employment as a share of the adult population. Where is work done? Well, who would have ever thought that the worst performing metropolitan area in terms of employment growth would be Los Angeles? The last 25 years, Los Angeles has lost 3% in employment[?]. No metropolitan area has done worse. You might be thinking about Detroit. Detroit actually did better than Los Angeles. Detroit lost 1% compared to Los Angeles at 3%. How has compensation evolved? Well, back in the day, technology used to be the tide that raised all boats. All workers' salaries grew with technological change. But today that's much, much different. What's happened is that for highly skilled, highly educated workers, the growth rate of compensation is about 5 times as much as it is for those with high school training or perhaps less than high school training. Let me show you just a couple of pictures to drive home some of these points. This is a picture of the share of the working age population who work, and you can see that it was fairly constant between 2000 and 2007, and the shaded areas are recessions. And you see we fell off the cliff during the recession, but the remarkable amount of work that was being done before the recession hasn't been restored. And we've had a slight uptick here, but we are roughly 9 million jobs below where we would be if we had simply maintained the same employment share that we had in 2006, 2007. Now in terms of who works, we've seen remarkable changes in who works. [Ohanian skips back to continue on the employment share.--Econlib Ed.] This is a picture of our employment share that goes back to 1970. So this puts our decline in relative perspective, and what you'll see is that the decline we've seen very recently is not just a consequence of the recent recession, but really this is a decline that's been continuing for about 20 years. So, the issues that we have in the labor market that Russ talked about really are long term issues, not just a consequence of the most recent recession. Now, in terms of working, the share of women working has increased by more than 50% over the last 50 years, so 40% of women worked, up to about 60% more recently. And that's coincided with a sharp drop in men working. So in 1960 over 86% of working age men were working, and that's down to less than 70% today. So as we go forward, we think about how much work will be done, who will work, how will compensation evolve, there will really be three factors that govern that. Government policies that impact the profitability of conducting economic activity at a particular location such as Santa Barbara or Los Angeles, the United States; globalization which governs who the countries we interact with, we trade with, we learn from; and technology, which governs the efficiency in the way we produce goods and services. And I'll just stop with one last picture, which shows productivity. And the reason we are the richest large country in the world is because we've had technology and productivity growth; it grows about 2% per year, which doubles the efficiency of workers in about 35 years. And the trouble we see is that this is output per hour; and output per hour, the red is the long-run trend of 2% per year, and this squiggly black line which lies below the red is actual output per hour. So our productivity is growing about half the rate over the last 5 years as it has in the past. Now, that's got to reverse if we are going to continue to have a prosperous society. And part of the good news is that the technology will change that Andrew talked about has the potential to raise this substantially and all these factors will get into the panel discussion. And thank you again for coming this morning.
22:47
Russ: So, for our first topic I want to talk about the rise of technology that Andy spoke of and his book is about. When we think about technology historically we see that technology has had a big role in displacing physical strength. So, John Henry was a steel drivin' man, but machine comes along, and his great skill then becomes totally unproductive. What we see the glimmers of--and not just the glimmers; it's not science fiction, it's real--we see machines coming now and technology software--Marc Andreessen says software is eating the world. We see machines and technology coming along that seem to be able to replace not just physical strength but intelligence. The driverless car is an obvious example, as you write in your book. Researchers thought it would be impossible, that it would take decades, that it would never be done for a machine to be able to recognize the subtle things that just a simple driver, that we do effortlessly when we drive a car. Well, it's here. That skill, which is a modest skill, is now on its way to being redundant. We see professors, which many of us up here are, being replaced by technology, either by something like Kahn Academy or the rise of so-called MOOCs, Massive Open Online Courses, that may make it impossible for many professors and teachers to make a living. It raises the question: what's going to be left. If strength and intelligence are now taken care of by machines, what are we going to do? Obviously these are just speculative questions. I'd like to start with you, Andy, and talk about how that bounty, which is the result of this technology, might be shared by those of us who are actually spending our time doing productive things. McAfee: So, exactly as you say, 200 years we've been predicting the onset of massive technological unemployment because of the latest round of big new technology that came out. The Luddites smashed looms just about 200 years ago. Marx of course predicted this. John Maynard Keynes wrote an essay in 1930 called "Economic Possibilities for our Grandchildren," where he introduced the phrase 'technological unemployment.' Those folk have all been wrong. And the reason they've been wrong is that essentially, innovators and entrepreneurs come along and think of new needs and new wants to go satisfy out there in the economy, in society. And they employ a lot of people to satisfy those wants and needs. So, the engine of creative destruction, amazingly enough, has been tilted a little bit in favor of creation and away from destruction, when it comes to jobs. So, the question is, if things are different this time, why are they different this time? And the only honest answer any of us can give is: I don't know. My suspicion is they actually are different this time, for the reason you give. We had our power system be largely automated in our organizations, in our tasks, in our jobs for a long, long time. Now the control system is becoming increasingly automated. When both the power and the control are done by machines, what do we need all the human labor for? I honestly scratch my head over that. I don't know. But the critical thing to keep in mind is that the answers to that question are not going to come from the alleged smart people on this stage. They are going to come from this very decentralized activity of innovation, entrepreneurship, capitalism, that we're pretty good at. The only hope we have for putting people back to work and finding more gainful employment is to let that engine do its work. Now, that doesn't mean we need no regulations, capitalism is this fantastic, wonderful thing. My point is that's our only hope for figuring out what all of these people are going to do. Russ: Lee, are you an optimist or pessimist about the future of employment opportunities in a world of artificial intelligence? Ohanian: The trend over the last 30 years, which is really when we started to see the massive increase in technology and information processing has been to substantially improve the productivity of highly skilled people. So, those with the best degrees, college degrees, have benefited enormously from the technology revolution. And as you know, it has really impacted those who don't have those capabilities. My favorite example is in 1950, the ports of New Jersey and New York employed about 10,000 people. Strong guys, longshoremen who picked up a box and brought it to shore, when that ship was empty they filled it back up. Today, [?] 10,000 people works in New York and New Jersey that now move 20 times the load, there's less than 1000 people employed. They are all college graduates that are operating sophisticated forklift, computerized machinery to do all this. So it really has become a situation of sort of haves and have nots. And education is going to be central in terms of allowing workers to benefit from the technology that's available. But one thing I would say is so hard to predict the evolution of these factors. In 1984, AT&T punted on cell phones and gave it to Ameritech because they thought the maximum audience for cell phones would be a million users. Russ: But it slightly, just a titch underestimated the size of that opportunity. Ohanian: Just a little bit. McAfee: There are more mobile phones subscriptions than there are people in the world right now. Russ: Yeah. Which is literally mindboggling. It's incredible. Ohanian: So I think it's just remarkably hard to predict how these things will evolve. I suspect that the greatest computer in the world--there's many of them; they are sitting between our two ears--that we'll still be able to develop technology that makes use of the large [?] of human capital we have in the economy. But again, it's hard to predict how these things will evolve.
28:46
Russ: Megan, optimist or pessimist on this question? McArdle: Well, I mean if you don't have power or control, I guess what's left is charm. [?]: We're in deep trouble. Russ: Yeah. Most of us on the panel are. McArdle: I guess my role is the hippy-dippy portion of this program. A friend of mine recently joked that China was taking over a lot of U.S. production to allow us to specialize in what we are good at, which is obviously artisanal pickling. It's a joke but it's also true, that like every Sunday morning I get up and I take my little basket and I walk over to the farmer's market and on the third Sunday every month I go to pick up my meat CSA [Community Supported Agriculture]; and I know the animals have been treated well. [?]: Well, but you don't live in America. This isn't really clearly like that.[?] McArdle: No, this is--or take massages. Like, a massage in 1930 was something that like crazy women in the movies got. Or maybe like boxers would get. It wasn't something where like every pregnant woman in America thinks it's now something you just get a pregnancy massage when they give you the positive test. Okay, go get a pregnancy massage. These services that used to be something that were for the super-elite. And maybe will eventually outsource into robots, but I've tried those Homemedics Mats, and it's really not the same. What people like--what is the most important thing after we've been fed, after we've been sheltered and warm and whatever? What do we like? We like interacting with other people and feeling happy and loved. Right? So, we like feeling people being nice to us, we like feeling special, we like having other people call us 'Miss McArdle' and give us a cup of herbal tea and ask us whether we want our feet massaged. We like all that--maybe not the men in the audience so much. But you guys have golf. That we like those experiences of having other people treat us well. McAfee: But think of your last 10 service interactions with another human being. How many of them left you with that warm, chamomile tea feeling? McArdle: Well, at some level, most of them, actually. Service in America, especially if you've lived abroad, is amazing. In Britain--they practically-- McAfee: That's the wrong benchmark. Terrible benchmark. McArdle: Well, the French, I'm not sure I would put the French up against it, either. Actually, mostly, if it is nice, when you shop, people are nice to me. McAfee: Oh, come on, did you walk through a sea of pleasant experiences in the airports on your way here? If so, I want to travel with you. Russ: That's an outlier. It is an outlier. McAfee: When you call up Comcast, when you go-- Russ: Also an outlier. McAfee: No, no, no. These are the service economy. These are not outliers. Russ: [31:35] No. You've picked the example of the places in America where there is very little competition due to regulation and government monopoly. I mean, there's competition-- McArdle: McDonald's people are very nice to me. Russ: They're cheerful. McArdle: And they ask me if I want something else, and was I satisfied with my order. Russ: Would you like fries with that? McArdle: Emailing me to find out if I liked something that I bought from them that I spent two minutes consuming and never thought about again. Actually I think there's more room for those sorts of things for things like haircuts and things that women didn't used to do. You used to perm your hair at home. My great aunts all permed their hair at home. Now, who would do that? No one would perm their hair, thank God. McAfee: Wait, explain this hair concept? Russ: That's a visual joke, for those listening at home. They'll just have to get the video version. McArdle: More people are getting services provided to them. Even people--you go and you work 10 hours providing hair services and then on the weekend you go and get a massage because you are tired. McAfee: I'm pushing back, but you are absolutely right. One of the weird things that happened in America over the past 30 or 40 years that really no one was anticipating is exactly the rise of these service jobs. And they came on us in huge volume. They have actually helped maintain the wage and the unemployment [employment--Econlib Ed.] levels at the lowest levels of skill and education in the country. The main problem is they are the lowest levels of skill, education, and pay in the country. So great these jobs exist; I completely agree with you. They don't look like ladders toward that classic middle class prosperity in a lot of cases. For exactly the reasons that you outline. McArdle: I think that this is the giant challenge that we face--that we still in a lot of ways maintain the attitude about the service jobs. Part of it is a sheer supply/demand thing: we have a huge oversupply of people who are not [?] by the manufacturing sector, and are transitioning, especially working class men, as I was talking about, they are transitioning very uneasily into the service economy. But that's sad. I mean, lawyers are service people, doctors are service people--there's lots of stuff in the service economy that doesn't have to be low prestige because it's a service job. So partly to go back to the hippy-dippy portion of the program, is that a lot of this is how we treat those jobs and how we view them. Do we say--because, yes, in 1930 if you were a manicurist, the people you were manicuring were extremely rich people and you went home to your cold water flat in Brooklyn. But we don't have to view these jobs that way, and that's a cultural choice that we are making. I think middle class, we're going to have to rethink what the middle class is in this context, and what these jobs mean. But I don't think it's inherently a feature of these jobs that they involve low pay and misery.
34:26
Russ: Lee, want to weigh in? Ohanian: This discussion really kind of pushes towards the idea of the jobs that will exist will depend upon what people want. What society wants is going to be produced. And the reason we have so many more of the service sector jobs is because it's getting cheaper to produce these service sector jobs, and because the income is so we now have a lot bigger share of the population can now choose to have a massage or a manicurist than years ago. Russ: It's part of the bounty that technology has given us along the way. Ohanian: It's part of the bounty. Growing at 2% a year means we quadruple living standards every 70 years; but the lifespan now is 80 years. That means children born today, if continued, will see increases their standard of living by a factor of 5. And that's all made possible by technology. Even a 1 percentage point increase, from 2% to 3%, won't be a factor of 5--it will be closer to a factor of 10. So, the possibilities are just enormous. And what it boils down to is: How will that bounty be distributed across different segments of society? McAfee: And Megan's comments bring up a fundamentally important question, is, to exactly your point: How will those benefits be distributed? And when you talk about these lousy service sector jobs, I completely agree with you. If they are going to exist and be a big deal in our economy, how do we turn them into better jobs? And to oversimplify to a huge amount, there are a couple of main levers to do that. One is kind of regulation and fiat and a $15 minimum wage, and you have to treat people at this level. And to the extent I've got any economics thinking in my brain at all, it says, Great, you are baking in rigidities into the process and you are going to scare employers away from those jobs. And they'll automate them more quickly. The other route to that is, you keep talking about hippy-dippy cultural stuff, is to try to change the perception of how we think about those jobs and the pressures we can bring to bear on the stereotypical exploitative employers who pay $9 an hour and make you juggle your schedule around. Do you have any confidence that we can--whether or not it's fair, let's call that a Walmart kind of job. I don't know enough to know how badly they treat their people. But just for stereotype purposes. Do you have any confidence that we can pressure Walmart into behaving differently in that regard? McArdle: Well, I think there is a--so, to start with, I actually think that the money is a problem, but it's not the biggest problem. Because, as you say, there's going to be so much stuff. Right? And the things that money creates a problem with now are kind of artificial scarcity--housing, because it's hard to build, good school districts where you are reshuffling the children around, everyone wants their children with kids who are smarter than their children. And those things, yes, those are a problem, but they would be a problem if people were making $30 an hour. If there weren't any more houses, the bottom of the income distribution would still be living in those houses regardless of how much money they were pulling in. So that actually a lot of those issues are supply side issues. What I do think is--the serious issues are insecurity and the inability to either count on, as you say, understand what your future path looks like, that this is a job that I could actually have for a couple of years, that I can try to build a career and understand what that looks like, that I know I will have this job, that my employer views me as someone that they are not going to fire and let go. The problem is not Walmart in particular. Walmart was always a terrible job, in some ways. It was aimed at housewives and teenagers--that's who they employ. The problem is the other jobs aren't there. So I'm not sure I think it comes from Walmart, because retail is just, in every era that I've been aware of, has been a terrible job. Girls in 1900 New York working in department stores didn't make enough money to live on, and they had--ships--a guy made a ship where it was a dorm for the girls and they would bode out into the harbor every night, because this was the cheapest way to stack them. So, I think the issue is things like manicurists and massage therapists and golf consultants and all of those things--and this sounds really dumb; I know that there are a number of people who are going to be listening to EconTalk, maybe some people in the audience thinking, My God, this is like Marie Antoinette saying 'well, let them eat manicure jobs.' But I think that this is a fact. I'm not happy or sad about it. I just think [?] when you want to compete with a machine, what you cannot [?] is being a person and providing person-like affection and sociability. Russ: People want to chat with their manicurist. And some manicurists like to chat with their customers. There will be a machine that will give you a manicure. The question is, what will be the nature of that interaction that you are talking about, and isn't it going to reward people who enjoy that? We know a bunch of people--we're very unrepresentative, us on the stage. We're way too educated. We don't have a lot of experience of a Wal-Mart cashier. When I'm in a Wal-Mart, which I admittedly shop at occasionally, I'm not ashamed of it, they are happy people. They seem to be enjoying that interaction. I have many friends who do, too, in that they want to sit in their cubicle all day working on their computer sitting in front of a screen, but there are a bunch of people who don't. And they have a different experience. The question is what are they going to be compensated--to me, and is the social consequences of those differences going to be large? Presumably if a lot of people want to sit at a screen and make a lot of money, the fact that being pleasant is a scarce resource, maybe it will get rewarded more generously in the future than we anticipate. McArdle: Well, I was actually the most cheerful cashier at the [?] Pharmacy chain in New York City. People would actually ask me where I was from, on the ground that no one that nice could be from New York. And so one of the things that--who is sociable, who likes taking care of other people--and this is what we've been talking about with the decline of [?]--that's a lot of it. It's that women are willing to do these jobs and they don't mind doing them. There are lots of guys who are good at sales and customer service. It's not like guys are not capable of this. Part of it is that the role we've assigned to masculinity, right, is being gruff and not caring what other people think, and that's not a really very good characteristic for a service job. Russ: That's changing. McAfee: I think Johnny Weir is leading the charge into a different view of that. I for one think that's fantastic. Russ: Who? McAfee: He was the guy wearing the fantastic hat at the Kentucky Derby. Russ: Oh, okay, sorry. Not a Derby guy. I missed out on that. No, but there's no doubt that the cultural factors that, Megan, you are emphasizing, for how we greet and cope with this transition are going to be important. [?]: There's another huge thing in play there, which is that we consumers might not want, might choose to have fewer interpersonal interactions in our day. Russ: We might. [?]: Because the world that's coming is a world where I can go from the airport to the--the car is still going to have a person for a little while--and check in automatically. And I can choose to have many, many fewer interpersonal interactions than right now. And the question I don't know the answer to is: How is the market share going to break down between consumers choosing to go have deliberately interpersonal interactions and the 'just let me check in over here and get my room without talking to a person'. I predict a lot of us who--myself included-- Russ: Correct. [?]: are going to break over that way. Russ: And we're already in that direction. One of the strangest things that we are already--we are already toward the semi-driverless car--is that in the old days, I used to talk to my cab driver. And now, it's culturally acceptable that I don't talk to him. I treat him like a robot, because I'm on my phone. I'm checking my email, surfing the web, answering stuff. And that experience for that cab driver might be better, because he doesn't have to pretend he is interested in talking to you. But again, I think there are people who find that part of the job appealing, and there are people who do like to talk to their cab driver. But many of them, probably you and me, prefer to be on our own. [?]: Definitely you and me. Russ: We are definitely heading in that direction. I want to--yeah, Lee, and then I want to take us to a different direction, topic. Ohanian: I think one issue when people think about, are we going to all lose our jobs because machines are going to become as good as us and they will do it cheaper--technology can be directed toward replacing workers or towards enhancing workers. And the biggest scarcity of constraint most people face is time. And in the example you just gave, I can be on my cellphone in the taxi and I can make progress; I can be productive. That's a great example of a labor-saving device. So you can now make use of time where previously it was gone. Your time in the taxicab about did the Mets win today or not? So technology can be directed, and it will be directed if it's profitable to make people more productive. So it's not as if technology is either we replace people or we have to put brakes on the process of technological change. We can produce technology that enhances our ability to remain productive, remain employed, and produce more business services. McArdle: Look at the Apple store. They employ computers in order to enable these people swarm[?] you and find out what you always dreamed of, in a little tiny laptop or whatever.
44:10
Russ: Lee, I want to take your comment and take us in a subtly different direction of how we might, how our culture and our society might evolve to deal with these issues. So you said technology can either enhance or replace, and I like your dock workers story. Follow me on this chain. So you've got 10,000 dock workers; now there's fewer than 1000, and the thousand are how many times more productive, in dollar terms? Is it a hundred times? Ohanian: There are too many zeros for me. Russ: Exactly. So it's many, many multiples of times. Then your next thought is, but somebody's got to make those machines that make those fewer than a thousand workers more productive. But of course those machines are now made by other machines. So, manufacturing in America--it's a secret--is a booming, booming part of economy. It's just not a source of employment. So we make a lot more stuff with a lot fewer people. So you say, well, that can't solve the problem. But then you think, but someone's got to design those really smart machines. And then we get to sort of the real scarcity that gets rewarded, and then you start thinking: Well, how many people can be involved in the creation of machines? And of course there's an enormous move, mainly by people who are already in the field, to encourage kids to learn how to code, to become more computer savvy. Obviously, it's happening; it's not--you don't need a policy change to do that. People are obviously growing in their ability to do that. So I want to shift just to education. Let's talk about where we think the education system might be heading. What's holding it back that will let people prepare for perhaps a very different world? Megan, what are your thoughts on that? McArdle: I think that one of the problems with education, and one of the problems with health care, is that unlike other issues, you stick a person in one side of the system and they come out the other; and they get a good job, they live or die if they go into the hospital. And it's hard to know if it's because you took someone who was conscientious and hard-working and really bright, or because the system is good. And so, one of the things that you see--even in the old line unions like the Operators Union is that they are desperate for people. Because they need people who can do high school math. And those people go to college, and they don't want to be operators. These are good jobs; they pay really well, great benefits, better than a lot of college students who are ending up in Starbucks. But the college student who ends up in Starbucks, they want the option value of being a college graduate and so they never become an operating engineer. And so you see this in a lot of these unions that require skilled workers is that they can't get them because nobody wants to do blue collar work. And at the same time, people are working in Starbucks and waiting for their big break, as some sort of white collar worker. Can we produce more people who are good at high school math? It's hard to know. When you look at these studies, it's really discouraging. It's discouraging how little early childhood intervention do; it is discouraging how badly a child who goes into a high-poverty school does and how low the graduation rates. And how nothing we have done in 30 or 40 years barely budges the [?] and when we do find someone, it's usually a good principal or good collection of teachers who make a difference; we can't replicate it. I believe that there are great principals who can make a difference. I also believe that there doesn't seem to be a way to teach others how to be that great principal. And so, I think that education, if we could educate people in the way that we would ideally like, would help. On the other hand I would also say that there is kind of a myth of the STEM shortage. Russ: STEM being science, technology, engineering, and mathematics. McArdle: Is that in fact, you look for example at pharmaceutical researchers. These guys were getting devastated by all the layoffs in pharma. And nonetheless, there are lots of kids being encouraged to get a major in chemistry because that's a great job to have when you get out. In a lot of fields there doesn't actually seem to be an undersupply. A lot of the undersupply that you see in all of these fields is that the employer who is competing with China can't afford to pay what people want in order to bring their skills to that employer. And so, what we don't have exactly is mismatch--what we have is a mismatch between people who want to make $40,000 a year and be a highly college-educated chemist for a company and that's what they can afford to pay and compete with China. It's really, really tricky. Russ: Lee. Ohanian: For our generation, I think, public education worked pretty well. We had a foundation where we could go out and succeed if we had the ambition and we tried hard. I think that for a large number of children today, that's a more difficult challenge. And when you look at international comparisons across student achievement, it's depressing. Only about one third of our kids are what's termed as proficient in science and math. And it's not a particularly high bar. Compared to 70% proficiency rates in Japan and Korea. And in our poorest states such as Mississippi or New Mexico, proficiency rates are comparable to Croatia and Kazakhstan, very poor countries that just don't have the resources to develop or to devote to education. So, we have to make substantial changes in the quality of education. McAfee: And do you have a good sense of why the wheels came off some time between when you were getting educated and today? Ohanian: There's been a lot of research done on the impact to teachers and the quality of teachers. So, our colleague at Stanford, Rick Hanushek, has done some research looking at the impact of teachers and what would happen if, say, the poorest-performing teachers were taken out of the classroom and what you put back in was roughly the average quality. And the results are remarkable. So, for example, just for kindergarten: the importance of a good kindergarten teacher compared to a poor kindergarten teacher, about $400,000 in terms of lifetime income. In terms of providing a foundation to a motivated child to get them on the path of learning. And I think one aspect of education that sometimes is not discussed enough is that to allow us to succeed in the world we have today and the type[?] world we'll have tomorrow is we have to be adaptable. We have to be able to learn new tasks, new activities. And rote learning which is sometimes common in our education system is exactly the wrong approach. Rote learning means you cannot be creative, you can't adapt, you can't learn new things. So teacher reform I think is a major issue. I think more and more people are understanding that increasing competition in schools is an important factor. And I hope you see that as we go forward. Russ: Andy. McAfee: Your story reminds me--I was on stage a while back debating with Peter Thiel, who is a legendarily good venture capital, investor, entrepreneur, and he's got some fairly [?] ideas; but he said something that I found really interesting. He said, We all agree that critical thinking is an incredible important skill for people to have. I don't think anyone who thinks they are teaching critical thinking, who labels what they are doing critical thinking, is actually doing anything like teaching critical thinking. So the question then becomes exactly how do you instill these kinds of skills that you are talking about, this flexibility, this ability to learn new stuff over the course of a career or a life? This is a great big deal. My really unscientific, idiosyncratic answer is: I was a Montessori kid, and I'm so grateful for that experience, because it really did teach me that the world is an interesting place and my job is to go poke at it. There's a thing called the Montessori Mafia in the high-tech industry; it turns out that Jeff Bezos of Amazon, both of the Google founders, Jimmy Wales of Wikipedia--these people were all Montessori kids. So I think there is something about that early childhood flavor of, this hippy-dippy educational system that means something. But let's go test that with data. Like you say, your colleague and [?] have done amazing work to help us understand how bad a bad teacher is and how much better even an average teacher is. They did that because we have data over a long period of time on these issues. Let's continue to gather the data. With education, there is no shortage of theories about what works. There's a big shortage of really serious data and research about what's worked out. So let's turn the dial up on that. Russ: Yeah. The real question is, is it really the case that Andrew McAfee, Sergey Brin, and Jeff Bezos were transformed by the Montessori school-- McAfee: And we would all have been hopeless wrecks without it. Russ: Exactly. It's hard--well, I say it--not only is it hard to know, it may be impossible to know. The data is very difficult to use in a way that would allow us to tease that out. McAfee: I don't think we have enough separated-at-birth identical twins. Russ: That's the problem. McAfee: One of whom went to Montessori and one-- Russ: Exactly. McArdle: We could fix that model--get some identical twins. McAfee: That's social engineering for you.
53:20
Russ: So let's--we all agree that the education system could do a better job. I think the challenge is: How do we get there from here? And Lee, you mentioned more competition, better teacher quality. What are your thoughts on how we might have a better education system for preparing people for that world of work? In particular, is it really--how important is it? I like to think about the fact that we have states where people are being educated the equivalent of some very poor countries. And yet somehow, we do pretty well. And I think the reason we do pretty well is there is a small number of people who manage to transform the world and lift us up--those are the Andrew McAfees, Jeff Bezoses of the world. McAfee: And we're magnets for the smart, tenacious entrepreneurial people from all around the world. Russ: Yeah. So, if education does make a difference, and I'm somewhat skeptical--Rick Hanushek has been a guest on EconTalk and talked about that work and defended it--but I'm somewhat skeptical of that transformative power, because we see so many other things going on. But to the extent that it is, what are the key things that we ought to be doing to let that happen or to cause it to happen? Ohanian: When you look in the classroom now--you and I have gone through secondary school, post-secondary school, graduate programs--I think what we see is how important peer-group learning is, peer group effects are. A big difference between when we went to school and I think school now is that class size has gone up from 20 when we were kids, up to 30 and perhaps higher in some schools. And as you get large class sizes, you have a breakdown of peer effects, because the chance of having a disruptive kid or more than one disruptive kid is much higher. And from the standpoint of research on peer effects, having that continued as a smooth process through primary school and secondary school is, my guess is that that can be very important. As opposed to kids that are sitting in the classroom saying, I'm not getting stimulated; I'm not really seeing what's being talked about; the kid next to me is being disruptive. And then you start to lose a kid. So, as you point out, research is difficult in this area because control groups are hard to come up with. Based on my own observation teaching for 20 years, I think of peer effects as being incredibly important. Russ: But when you talk about rising classroom size, in the public schools of America, for a long time the trend was smaller. The teachers' unions fought, argued--I think incorrectly, but you are sort of disagreeing--that smaller classroom size was the key to better education. Which of course was a way to justify increasing size of the number of teachers on staff. Two issues to think about. One, I think we tried that experiment; it didn't work very well, as far as I can tell. Let's say it differently. There's no evidence that it worked well. But the next thing I think about is the move toward technology, toward the virtual classroom, certainly gets rid of those peer struggles that get those peer benefits--you are talking about you are sitting in the classroom with 150,000 people around the world, with the best professor in the world, but you are not interacting in the way that you used to in a great classroom before. [?]: No, that's only one option for how you do massive open online education. Another option is that you would go listen to that superstar teacher and then come together in a physical space with peers. Russ: With a coach. [?]: Maybe with a coach. Flip the classroom that way. So I don't believe it's an either/or choice at all. What I do believe is that the technology does give you access to these superstar educators, does give you the ability to personalize the education so if the kid's not getting polynomials, you can back up and throw more polynomials at her. And it also instruments the whole process so we start getting huge amounts of data about what's working and what's not. [?] bravo. Russ: Yeah. Megan. McArdle: I think the possibility that is one of the really exciting places, that technology can pinpoint exactly where you are failing. Video games are really good at that. It used to be that you had to play 97 levels of Donkey Kong to get to the point where you were going to fail again. And now you stop, you go back to the same point just before you died. And my husband does this a lot. McAfee: You married one of those dropout guys that as you say just plays videogames? McArdle: No. Actually, he writes about video games. Which is-- Russ: So it's [?]. At least [?]. McArdle: [?] the stereotypical geek-think family. But you go back to the point right before you died and you try again. And that's incredibly powerful. You just go right to where you failed and figure that thing out rather than having to re-run the entire lesson. Or, if you're a smart kid, sit there while everyone else takes three weeks figuring out polynomials. But that said, it has to be done well. And that's the hard part. Yes, I think that good teachers probably matter a lot. But finding good teachers, it's not correlated with anything that we can study. We don't know, it's not education, Teach for America doesn't seem to have great results, and so forth. Finding those teachers, getting them in the classroom, keeping them in the classroom--that's really difficult. Especially because we've seen over the last 50 years is the ability, this increasing class segregation of American schools. And so, the automobile really gave, first affluent, and then middle class and lower middle class parents the ability to pull themselves away from poor kids, and so watching one disruptive kid, there are districts in Baltimore and D.C. where the teacher would be thrilled to have one or two disruptive kids. Where you have high-poverty districts, where you may not be getting enough to eat at home, Mom or Dad may have a drug problem; Dad may not be there, he may be in prison. Those kids, the issue is not that we don't have a top-quality teacher in there. There's like a lot of issues that have to be addressed first when you have a classroom you can't control. Which is what the teachers report. You can't keep a good teacher there, because it's draining and horrible, and they want to go to a school district where there are kids that will at least listen when they talk and they don't have to be drained every day doing it. And so the kids who need it most, the kids are getting completely left behind, and more and more of them as these working class school districts start to look like very poor school districts did, that those people are the least likely to get all of this stuff that we do. I was a Montessori kid, too-- McAfee: Ha! See! McArdle: and I assume that without it I would probably be living on the streets of New York City r