Is tech dividing America?
When Americans consider how technology has changed their lives, they tend to focus on how the internet and smartphones have altered how they watch TV, connect with friends, or how they shop. But those changes pale in comparison to how technology has already restructured the economy, shaking up the workforce and shifting opportunity to tech-centric urban hubs. As artificial intelligence quickly moves from fiction to daily reality, that revolution will arguably become much more consequential.
Economists broadly agree that technology will continue to be an engine of economic growth. But it also will upend old certainties about who benefits. Already, we can see a growing inequality gap, with winners and losers by region and workplace. The next wave of changes, handled badly, could make this gap even more extreme.
MIT researcher David Autor has been at the center of that conversation for two decades now. One of the world’s premier labor economists, Autor has helped drive a reconsideration of how Americans are really coping with the changes transforming their workplaces. And he’s trying to take the conversation beyond the ivory tower: His 2016 TED talk about the surprising impact of automation, “Why Are There Still So Many Jobs?” has been viewed more than 1.3 million times.
Autor’s interest comes from seeing these changes at the ground level: Fresh out of Tufts University with a degree in psychology, he ended up running a Silicon Valley-sponsored computer-training program for at-risk children and adults at San Francisco’s Glide Memorial Church, a counterculture hot spot. When he headed back to Harvard’s John F. Kennedy School of Government for an M.A. and then Ph.D. in public policy, he brought a newly keen interest in figuring out how the technologies being pumped in to the labor market would shape what it means to be a worker in the United States.
Autor says that both the optimists and the pessimists are wrong when it comes to America’s ability to deal with technology’s rise. On the plus side, robots won’t put us all out of work. But we really do need to worry: Just because we’ve made it through similar shifts before doesn’t mean it will work out this time. National policy has a huge impact on how countries weather these changes, and there’s reason for us to worry about how our leaders are facing the next wave.
We should take a lesson, says Autor, from how the discovery of crude remade Saudi Arabia and Norway. Oil revenue brought new prosperity to those countries, just as automation has been a boon to U.S. gross domestic product. But Saudi Arabia and Norway responded by adopting “different institutional arrangements,” ones that left Saudi Arabia rich, unequal and unhappy, and Norway rich, egalitarian and happy. The choices facing the United States, he says, are similar—whether the wealth will be segregated or shared. So far, he suggests, “I don’t think the U.S. is dealing with it especially well.”
Autor spoke to POLITICO just before attending the World Economic Forum in Davos, Switzerland.
(This interview has been edited and condensed for clarity.)
We’ve just started to think seriously as a nation about who wins and who doesn’t as the American workplace automates. In 1998, you co-wrote a paper that showed the rise of technology in the workplace was actually proving to be good for higher-skilled workers. Is that a fair read?
What that paper suggested was that it’s definitely the case that automation is raising the demand for skilled labor. And the work that I’ve done since has been about what set of activities are complemented by automation and which set of activities is displaced, pointing out that on the one hand, there were tasks that were creative and analytical, and on the other, tasks that required dexterity and flexibility, which were very difficult to automate. So the middle of the skill distribution, where there are well understood rules and procedures, is actually much more susceptible to automation.
So, there’s a hollowing out of middle-class jobs, but high-skilled, high-wage workers and the low-skilled low-wage workers remain? Is that what we’re seeing play out right now in the U.S.?
That polarization of jobs definitely reduced the set of opportunities for people who don’t have a college degree. People who have a high school or lower degree, it used to be they were in manufacturing, in clerical and administrative support. Now, increasingly, they’re in cleaning, home health, security, etc. Ironically, we’ve automated some of the stuff that was more interesting for us, and we’re left with some of the stuff that is less interesting.
And that’s still true, but it’s probably not any longer the frontier. In the last five years, computer science has made an end run around that problem because the nature of artificial intelligence and machine learning is that you can get machines to do things that people don’t understand. We don’t have to explain, “These are the attributes of a chair.” You just show a bunch of pictures of “chair,” and have the computer figure it out. I think there’s always going to be huge demand for human expertise, judgment and creativity. Those traits will be scarce at least for the foreseeable future. What’s becoming less scarce, what is now not scarce anymore, is people who can do repetitive cognitive or physical tasks, and I think there will be a lot more service tasks done by machinery over time. So, I do think that it’s going to be more and more challenging for people without significant training to do high-paid work.
“Automation anxiety" is overblown, you’ve said. How anxious should American workers be?
People are talking about how robots are going to take all the jobs, but we’re in a time of very dramatic employment growth and have been for a decade. Job growth is robust throughout western Europe, as well. So, we’re certainly not in a period where there’s any outward sign that work is coming to an end. We have had two centuries of people worrying very vocally about how automation will make us superfluous. I don’t think it’s made us superfluous, and I don’t think it’s on the verge of making us superfluous.
The greater concern is not about the number of jobs but whether those jobs will pay decent wages and people will have sufficient skills to do them. That’s the great challenge. It’s never been a better time to be a highly educated worker in the western world. But there hasn’t been a worse time to be a high school dropout or high school graduate.
In recent years, we’ve seen pushes to spread Silicon Valley-style technological prosperity to other parts of the country. Does that have potential?
I hope so. People are talking about it, which is good, because they’re not going to do anything about it if they’re not talking about it. Whether talking about it will generate results is highly uncertain.
I do think it’s absolutely the case that a lot of what we see—a lot of the political dissatisfaction, as well—comes from the fact that as average wealth and income in the U.S. have risen, it’s a very, very geographically concentrated phenomenon. Most of that is basically New York, San Francisco, Los Angeles, San Jose, Houston, Boston, and a couple other places. It’s not broadly shared prosperity. A lot of the country is actually kind of downwardly mobile.
So, it’s very misleading to just say, “Oh, well, average incomes are up.” The economic geography is much spikier than it was 30 years ago. The hills and valleys are much, much steeper.
In 2001, you wrote about the idea of geographic arbitrage, or how companies were going to take advantage of people being able to live and work anywhere, thanks to the internet. My sense is that hasn’t happened. If that’s right, why hasn’t it?
I agree. You could have imagined a world where we all have Skype and mobile phones and broadband, and no one commutes anywhere, and we all live in our remote hilltop houses overlooking the water, and we’d have no reason to travel. But that doesn’t appear to be the case at all.
It appears that remote and in-person communications are complements, not substitutes. Somehow the force of people wanting to clump together has actually, seemingly, if anything gotten stronger. And when we talk about these geographic inequalities, that’s what we’re seeing.
But you’ve also written about “superstar firms” like Google, Facebook and Apple that are the powerful centers of gravity in Silicon Valley right now. Is part of the geographic concentration we’re seeing that—at least when it comes to tech—there just aren’t that many big companies to go around?
That’s true, and all the more so because they don’t have that many employees.
But if you look at the top 20 candidate cities for Amazon to locate its second headquarters, it’s Atlanta, Austin, Boston, Chicago. … None of these is Gary, Indiana, right? These are not the down-and-out of cities. I’m not saying I blame them for that. That’s a perfectly reasonable thing to do. But I think some people had it in mind that Amazon would go into some blighted metropolis and turn it around. That’s clearly not what these choices suggest.
Say you’re Jeff Bezos. Would you pick Gary, Indiana?
No. I wouldn’t. But I’m sure he felt some pressure to do that, and it must have occurred to him that that would be a great thing to do if he could do it. I assume that they went through the process and said, ‘Well, can we do this in Gary, Indiana? That place has been down and out since they made ‘The Music Man.’”
But they must have looked into it and said, ‘That’s not going to fly. There are not enough educated people there. And we’re not going to get them to move there because it’s not a place they want to be.’”
Boston, where I live now, is growing explosively: Unemployment is incredibly low, real estate prices are really high, the city’s getting much more crowded. So, you would think that companies would be running away. But Amazon, Microsoft and Google all have campuses within two blocks of my office at MIT. There’s an agglomeration force that causes companies to want to move where other companies are. And talented people move where other talented people are.
If you’re a local elected official on the other side of the desk, do you unambiguously want Amazon’s headquarters?
Yes. Yes. Yeah. I don’t see any downside to that.
Sure, it’s going to create congestion and drive up real estate prices. But it’s going to bring in talent, create jobs, create opportunity, raise wages—and not just for people with Ph.D.s. It’s going to raise wages for people who work at fast-food restaurants, drive for Uber, clean houses. So, these cities are not crazy to want the headquarters.
In just the past year, Silicon Valley as an industry has developed a good and evil reputation. It’s cutting-edge and pays well, but it sometimes disrupts the world without seeming to care too much about the consequences, a la Uber. Which is it?
I don’t think it’s either. It creates a lot of benefits as well as creating real challenges: It’s definitely the case that it’s raising total GDP, but has been very dis-equalizing. It is up to our institutions to deal well with that or not.
Some countries have done a much better job at sharing the gains and making sure that everybody’s bought in. Others have been much more social Darwinists about it, and the U.S. is very much at the extreme of that among industrialized economies, of going, ‘rah rah,’ to the winners and ‘too bad for you,’ to the losers.
How, exactly, are other countries good at it?
Countries that, I think, are doing really well with this—Norway, Sweden, Denmark, Germany, Switzerland, Austria—have very good educational systems that prepare people not just for highly educated, Ph.D.-level jobs, but also very good vocational, technical education systems.
But there’s also the notion that there are multiple stakeholders in the economy, not just shareholders. Workers have more voice, and that makes people less apprehensive about these changes because they expect that if they are gains, they’ll get a piece of them, where in the U.S. a lot of people think, ‘Well, there might be a gain, but I’ll be worse off.’ And they’re probably right.
Other countries have made it a lot easier for people to feel comfortable about the changes they’re bringing on themselves. I think that’s one diagnosis of the current U.S. political system.
You’ve noted in your work that LBJ created a “Commission on Technology, Automation, and Economic Progress” way back in 1964. I didn’t know whether to be encouraged by that or saddened by that—that we’ve been talking about these questions for a long time and don’t seem to have any better answers.
There are two schools of thought that you hear often. One is, ‘the sky is falling, the robots are coming for our jobs, we’re all screwed because we’ve made ourselves obsolete.’ The other version you also hear a lot is, ‘We’ve been through things like this in the past, it’s all worked out fine, it took care of itself, don’t worry.’ And I think both of these are really wrong.
I’ve already indicated why I think the first view is wrong. The reason I think the second view is wrong is because I don’t think it took care of itself. Countries have very different levels of quality of life, institutional quality, of democracy, of liberty and opportunity, and those are not because they have different markets or different technologies. It’s because they’ve made different institutional arrangements. Look at the example of Norway and Saudi Arabia, two oil-rich countries. Norway is a very happy place. It’s economically mobile with high rates of labor force participation, high rates of education, good civil society. And Saudi Arabia is an absolute monarchy that has high standards of living, but it’s not a very happy place because they’ve stifled innovation and individual freedom. Those are two examples of taking the same technology, which is oil wealth, and either squandering it or investing it successfully.
I think the right lesson from history is that this is an opportunity. Things that raise GDP and make us more productive, they definitely create aggregate wealth. The question is, how do we use that wealth well to have a society that’s mobile, that’s prosperous, that’s open? Or do we use it to basically make some people very wealthy and keep everyone else quiet? So, I think we are at an important juncture, and I don’t think the U.S. is dealing with it especially well. Our institutions are very much under threat at a time when they’re arguably most needed.
In the time we’ve been talking, Amazon’s probably released a new AI product. As an academic, do you worry about keeping up with the pace of technology enough to usefully analyze it?
Nancy Scola is a senior technology reporter for POLITICO Pro.
via The Agenda
February 7, 2018 at 08:20AM