What can be done to counter inequality in the United States? What policies help people rise from poverty? What is the role of education, and how have the labor markets responded to various policies? David Deming discusses what Harvard Kennedy School’s Malcolm Wiener Center for Social Policy is doing to address the pressing issue of income and wealth disparities in the United States today.
Wiener Conference Calls recognize Malcolm Wiener’s role in proposing and supporting this series as well as the Wiener Center for Social Policy at Harvard Kennedy School.
Mari Megias:
Good day, everyone. I’m Mari Megias from the Office of Alumni Relations and Resource Development at Harvard Kennedy School and I’m delighted to welcome you to this Wiener Conference Call. Today we are joined by David Deming, who is a professor of public policy at Harvard Kennedy School and director of the school of Malcolm Wiener Center for Social Policy. He is also a research associate at the National Bureau of Economic Research. David’s research focuses on the economics of skill development, education, and labor markets. He’s one of the principal investigators at the CLIMB initiative, an organization that seeks to study and improve the role of higher education in social mobility. He also studies the future of work focusing in particular on how technology changes the labor market. We’re very fortunate that he has chosen to share his expertise today with the Kennedy School’s alumni and friends. David.
David Deming:
Hi everybody. It’s a delight to be with you today. During this holiday time, I know everyone’s busy and I’m looking forward to our conversation. I thought I would... of course I’m going to talk a little bit about my own research, but given that this is the Wiener Conference Call and I am the director of the Malcolm Wiener Center for Social Policy, I thought I would spend just a little bit of time telling you about all of the social policy activities that are happening at the Kennedy School, at the Wiener Center, and elsewhere and to tell you... the reason I want to tell you about this is because I’m really excited about it.
In the past year or so, we have started a major effort to really rebuild the social policy faculty and research activities at the Kennedy School. Part of that is generational because we’ve had a number of prominent faculty like William Julius Wilson, Sandy Jencks, David Ellwood, all nearing retirement and we sort of need to bring in the next generation of social policy faculty and I’ve been a big part of that. I led a search last year where we successfully hired three senior economists who study social policy issues ranging from criminal justice to health care to immigration and globalization, international trade, education, and a variety of other issues, and we’re also undergoing a search right now for sociologists who study similar issues under the broad aegis of social policy and so that’s I think optimistically around half a dozen new hires at the Kennedy School all related to social policy and then we’ve also, I’ve brought in a number of folks from other centers at the Kennedy School and then also some from around Harvard.
We have a bunch of people who are studying really cutting-edge issues in social policy who are all gathered together at the Malcolm Wiener Center community. Since I started as director in July I’ve engaged in a number of what I would call community-building activities. We have a weekly coffee where everyone gets together, the faculty and staff and students, to talk about issues and just kind of share what’s going on Monday mornings, and we also have a weekly social policy seminar where we bring in distinguished outside speakers and I’ve also organized a bunch of center activities. Like we’re going to go visit a homeless shelter, a local homeless shelter and volunteer in the spring and then actually we’re having the Wiener Center holiday party tonight in my house.
Broadly speaking, my first year has really been about building this community and working together because a number of us work on projects together and so we’re kind all getting to know each other better and thinking about the work forward, and then the next phase of this work is really going to be to increase the external footprint of the Malcolm Wiener Center to kind of let the world know about all the great research we’re doing, and all the great work we’re doing in policy and practice related to social policy, and so I’m very excited about that. I’m happy to answer questions about it if you guys want to ask more but at the interest of time I thought I would then pivot to talk a little bit about research.
Before I do that though. I do just want to say since this is the Wiener Conference Call and I am the director of Wiener Center, I want to say thank you to Malcolm Wiener and the Wiener family for all of your support of this work. I see it firsthand as a director of this center how much of a difference it makes and I think this is some of the most important work that’s happening in the country right now, and I want to thank you guys for supporting it on behalf of myself and behalf of the entire Wiener Center community, so thank you.
Okay, so my work. I am an economist and I study, as Mari mentioned, education, skills and labor markets and about... I mean it’s work on a variety of things, but what I’m doing right now is I would say about half of it is this project with the CLIMB initiative that she mentioned, which is related to the role of higher education in creating social mobility in America, and the other half of my work is on what I would broadly call future of work type stuff, AI and technology. First I want to talk a little about colleges and then I’ll move on to the future of work stuff.
The CLIMB initiative, to be a little more specific. is an organization that I started along with Raj Chetty at the Harvard econ department and John Friedman who is at Brown, and it is a partnership between us and about 450 U.S. colleges and universities around the country plus the college board, which administers the SAT and the PSAT and the ACT. We’ve obtained from them all of their data going back to the mid 1990s on admissions and attendance at all these colleges and then also a national population of college test takers, and we’ve merged that into U.S. tax records so that we can study the role of colleges in intergenerational ability, because of course when people apply to colleges, they’re applying as children and they’re claimed by their parents on their tax returns and so we can link people at the point of application to their parents and then we can observe them in adulthood and see what happens after they graduate, and so we’re using these data to paint a really rich picture of what’s happening with American colleges.
We have an ongoing project now that I ... I can’t say too much about, but it’s very exciting. It’s related to admissions at selective colleges. It looks that it’ll be coming out pretty soon. And then another larger project looking at what I would call the kind of policy implications of investments in higher education for the broad population. Of course, something like 70 percent of American college goers go to public institutions and so that’s where a lot of the policy action is and so we’ve got sort of a big project looking at admissions for the more selective, mostly private school side and then another project looking at the broad landscape.
Part of what we’re doing with CLIMB is not just trying to understand what’s happening for research purposes, but also to use this network to try to understand what’s working. Are there colleges out there that are doing a particularly great job at graduating students, particularly low-income students, first-generation college students? Are there certain programs that they’re running or certain features of these schools that we can learn from and they can learn from each other in our network and sort of build in policy solutions that help increase graduation rates and ultimately build the skills of the next generation of workers, which I think is an urgent national priority and important for a long run economic growth.
That’s ... and I’m doing a bunch of projects just sort of looking at policy solutions for higher education, increasing access and using higher ed as a tool to reduce economic inequality. That’s one thing I do. Another thing I’m doing is as I mentioned about technology and skills and the future of work, and maybe that’s ... what some of the questions I see are going to be about and that’s great. I have done some work looking kind of in detail at what happens when jobs change. What are the new skills that are required by employers and what are the things that are becoming obsolete and kind of what are the overall big picture trends.
I had some pretty recent work looking at the growing importance of social interaction and teamwork in the labor market. I think this was related to technological change pretty fundamentally because there are a set of skills that we as humans are actually quite good at even though we do these things unconsciously and without very little effort, they’re actually quite cognitively complex. One example I’d give you is all of you on this conference call to varying degrees can have a two-minute unstructured conversation with another human being without much problem and you don’t think about that as being cognitively complex, but it definitely is and we know it is because it’s actually impossible for anybody to program a machine to do that with any reliable accuracy of course, then it turns out there’s a prize called The Loebner Prize and it’s given out every year to people who can write a chat bot program that can fool a human judge into thinking that it’s a human and nobody can do it. That’s because of the open-ended nature of conversation in social interaction. It’s not just an algorithm that’s necessary to try to predict what people are going to say next. You also have to have what psychologists called theory of mind, which is essentially the ability to walk a mile in somebody else’s shoes, to imagine the mental state of another person and anticipate their desires and needs. And you see people do this effortlessly all the time and we don’t actually know how we do it. When I teach my classes I’m constantly scanning the room, looking at the faces of my students trying to figure out are they understanding what I’m saying? Are they bored? Are they checking their phone when I’m not looking? Did I just offend them with the joke I made? I can tell, I can read the room and get a sense of how things are going and I’m sure you guys have had this experience too. I mean in your own life, but you don’t know how you’re doing it. You can’t write down a set of instructions for how a machine could figure out what people are thinking and so that’s the type of thing that’s very hard to program in a machine or an AI, some type of algorithm because we don’t actually consciously understand the process. It’s much more complex than we realize. On the other hand, we tend to think of things like the ability to play chess at grand master level or the ability to do complex mathematical problems as a sign of high intelligence in humans, and in some sense it is. On the other hand, it’s actually a pretty trivial programming problem.
Those are things that in formal logic, formal mathematical operations are actually pretty new and so the cognitive processes behind them is actually a lot more shallow than you think. It’s hard for humans but easy for machines. Whereas social interaction is easy for humans and hard for machines and so that’s very much a trend that we see in the labor market. Jobs that are in more intensive in social interaction and teamwork are growing as a share of all jobs in the U.S. economy. That’s been true for about 25 years. They’re paying relatively higher wages over time and you see growth in jobs like managerial positions, teaching, health jobs, college instructors as well. Economists, I’m pretty well covered. If you look at job categories that are growing.
In a lot of categories of jobs that are shrinking more than you might think are STEM jobs that are not related to computers. Things like engineering. It’s not that those jobs are shrinking or they’re bad jobs. They’re still very good jobs, but they’re becoming less important in part because we’re increasingly developing technologies that can automate work that’s not people-intensive and so when you look at the big picture, I think the future of human work is going to become more people-oriented and not less, and so that’s kind of one element of the work I’m doing and then I have another set of projects looking at the returns to college majors over time.
I just started writing a regular column for The New York Times, and I wrote a column recently about returns to college majors over time. Essentially, it’s quite a controversial in a good way, but essentially it’s showing or demonstrating that if you graduate from college, right after college people who major in engineering or computer science earn about 40 percent more on average than people who major in social science or history, economics, political science, sociology and history, but if you look at those same people at age 40, there’s almost no earnings difference between them and it’s not because STEM majors earn less, it’s because the other majors catch up and they tend to for two reasons. One is that people who major in technical subjects tend to leave them over time, so the share of STEM majors working in STEM occupations declined pretty rapidly after the first decade of working life, and on the other end, people who major in social science and history tend to move into a lot of different occupations, including some that have quite high upper-tail earnings, in business and law and medicine and things like that. When you paint the picture overall, what you see is that these technical majors, while they’re very important for giving people some skills that matter a lot to employers, in the short run they have a shelf life, and so ultimately what colleges are teaching you is something deeper. It’s problem solving skills, critical thinking, learning how to learn. These are all kind of general future-proof skills that can survive whatever’s coming around the bend in terms of automation because they’re inflexible and transferable.
That’s kind of the type of work I’m doing. I’m writing another column right now about AI in particular looking at patterns of job growth and decline in that are quite surprising. For example, everyone talks about the future of driverless cars, but it turns out that truck driving is actually one of the fastest growing occupations in the US right now and one of the reasons for that is actually directly related to automation. It’s because everybody’s ordering things online now and so the jobs that are in demand are short-haul truck driving because Amazon can automate most aspects of the process of the... kind of process when you order something on your phone and when it shows up at your door, but they can’t automate that last mile of delivering the package to your door, and so you’ve seen a huge increase in hiring, especially in big cities and truck driving.
On the other hand, you see declines in retail jobs because people aren’t going to malls and so if you look 10 years ago, you’d see a lot of articles written about the retail apocalypse and how automation was going to destroy retail jobs and that was true. That’s true to some extent, but you saw no articles at all written about how automation was going to increase truck driving jobs because that second follow-on of automation, which is it creates new job opportunities, it’s much harder to foresee than the direct impact that replaces jobs, and so people tend to paint these doom-and-gloom scenarios about automation that are focused on the proximate impact, which is, what is the job that is going to be directly replaced, but it’s very hard to anticipate the following impacts, which are often much better.
That’s what I’m sort of what I’m writing about. Last I just want to say, even though I focus on education and skills, I teach a large class at the Kennedy School about economic inequality. It’s called “The Causes and Consequences of Inequality” and it’s terrific. About 125 students from all around the university and it’s a lot of fun and so I make it my business to stay informed on a broad range of topics related to inequality that are not necessarily my narrow area of interest. I teach about globalization and trade. We talked about the wealth taxation plans of some of the Democratic presidential candidates. We talked about politics and mass media and a variety of other subjects. I’m happy to take questions on a range of things and I’ll answer them to the best of my ability.
Q: I’d like to start things off with getting a little bit more into the whole question of artificial intelligence and machine learning and their impact on value chains in virtually every industry. Can you discuss a little bit about how we can address the expectations for short-term displacements versus longer-term benefits?
Yeah, that’s a terrific question. I think that is the key challenge because in the long run, automation tends to be a force for good, right? If you look at... think about the Luddites, okay. Many years ago, the Luddites rioted because they were being replaced by a technology called the Jacquard loom and in the short run they were put out of work and it created a lot of social unrest. In the long run their descendants have become knowledge workers doing a variety of jobs that their ancestors couldn’t have possibly anticipated, and we all buy our clothes at a fraction of the handmade cost, and so I think the long-run prognosis for automation is positive. You see that everywhere. You see that in the home. We already have robots in our homes. They’re called dishwashers and microwave and ovens, and that’s great because it frees us up to do all kinds of other things that are much more interesting than washing clothes and washing dishes.
The challenge is always about short-run disruption because of the people whose jobs are replaced by technology, so in my example earlier, the retail workers and the truck drivers who get new jobs as a result, they’re not the same people, right, and you can see it’s more extreme examples of that all the time. You look at the growth in software developer occupancies and one of the fastest growing occupations, those are people that are very high skilled that tend to be better off and some of the jobs that get replaced, that’s less true of those people. The challenge is always how to manage this transition and so I think... I don’t really have an answer that has to do with economics. I think that’s a political question, which is if we can see automation coming down the pike and we know it’s going to hurt some people and help others, I think it’s our responsibility as a society to anticipate that and I think one of the issues we have in the U.S. is that we haven’t done a very good job of managing that transition.
I don’t necessarily think that has much to do with policy that’s directly about technology. It has to do with policies that kind of allow for more flexible labor market arrangements that help people pivot easily to different jobs when they lose their existing jobs or when they want to get retrained and give some measure of economic security to workers who need to make costly and uncertain investments in skills to set themselves up for future success. I think that’s really the main issue. I think there’s some particular things about AI that are foreseeable, which I’m happy to talk about, but there’s a lot that isn’t. We don’t know exactly what the follow-on innovations from these quite general developments in machine learning and AI are going to lead to.
I don’t think anybody really knows and so part of it is we kind of have to develop policy for an uncertain future rather than saying, we know exactly what’s going to happen and here’s how we need to respond. I think that’s very hard to do.
Q: I’m calling this time from Toronto. First of all, my compliments as you have advanced what I consider to be very salient important questions meriting reflective response, but my direct question is grounded in the element of civics. Is our democratic system of governance now capable of affecting or ameliorating the transformative shifts in terms of values, because in terms of short-term transactional governance, the idea is grounded in that of being reelected versus being what is in the best interests of greater society. Whereas a transformative, you tend to be imperiled by leaping too far forward and how do you ground that, because Robert Putnam and Chris Hedges have lamented in terms of the fractionalization of our society between those who have and those who have not, and the despondency among the undercapitalized in terms of youth and also the fear of speaking truth to power because of that element of homogenization and being what I term politically correct, which unfortunately is not enhancing our values, whether epistemic or values in terms of doing least harm. I look to you in terms of your response and I thank you for your insights.
Thank you for your question. It’s a tough question. There’s a lot there to tackle. I don’t know if I’m going to give you a persuasive answer to every element of it, but I’ll try. I think you’re right that... I mean I share your concern over whether our country is prepared to help manage this transition that’s coming and by the way, I think that’s larger than just AI. I think the question is there a lot of things coming down the pike and I think everybody has a sense that things might look a lot different than the U.S. in 20 years than they do now and I think there’s a lot of concern. I share that concern in how are we going to manage that politically and civically and how are we going to come together.
I do think that it is hard to develop an answer that addresses a lot of the short-run issues we have, and I think part of what policy and researchers focused on policy should do is actually to deliberately focus on the long run because I think if you look at the history, certainly the U.S. and in many other countries, most of the best policy decisions that were made were ones that took a long-run perspective and I think that’s because the constituency for things like investments in basic research and investments in the basic skills of the population that constituencies are much less than they should be because many of the beneficiaries haven’t been born yet, and so I think it’s incumbent on people like me and places like the Kennedy School to advocate for things that are maybe not addressing the problems of tomorrow, but are long-run wins for sure even if we can’t visualize exactly how they’re going to play out.
I’ll just give you one example. If you look back a hundred years ago, at the turn of the 20th century, there was only one country on Earth that had begun the transition to mass secondary school education, so high school education, and that was the U.S. And that turned out to be an incredibly fortuitous decision for our country, because the European model was a focus on elite education, sort of narrowing the pipeline as you go higher, and in the U.S. we really did mass produce high school degrees and that turns out to unlock a century of economic growth that was basically training mass numbers of people to give them basic literacy and numeracy skills so they could work in the factories during the U.S. industrial revolution, and so I think a hundred years later, high school education is near universal in the U.S. The graduation rate is over 90 percent now.
And I think... we hear a lot of talk today about investments in college and the importance of it, but I don’t think there’s any question in the long run. We need to greatly increase the fraction of American youth who have more education. I don’t know if it needs to be a four-year degree for everybody or a two-year degree or whatever. I’m a little bit agnostic about how, but I think there’s no law of nature that says the natural amount of education ought to be X number of years and I think it makes sense that as people are living longer and work is becoming more complex and more knowledge-based. The knowledge frontier is continuing to move outward in terms of what we expect of people to be productive members of the working world, and so I think there’s no question that we need to make basic, large-scale investments in the skills of the population. That’s not the kind of thing that’s on the political agenda today always but I think it’s incredibly important for setting ourselves up for a long-run success, and I think by the way, there’s a lot of evidence that education increases civic participation as well.
Q: What lessons have been learned from other countries that have successfully managed to narrow the inequality gap that can be applied to the U.S. context?
Yeah. I’ll talk again about education, not just because it’s my expertise but because I actually think if you look across the landscape of what we know about what social policies work, actually the return to education is one of the most durable results in social science, and so I think what you see and continuing what I said earlier about investments in education is that the U.S., a generation ago, was a world leader in tertiary education in college in the kind of baby boomer generation, and since then we’ve made a little bit of progress, but not very much, and other countries have rapidly increased investments in tertiary education. Countries like South Korea, many European countries, actually have caught up to the U.S. or even surpassed the U.S. so that now we’re sort of middle of the pack in terms of share of young people getting a tertiary education, and you do see a very strong relationship between countries that have made investments in tertiary education and measures of inequality, and I think that’s because the price of a college degree that is what... how much more employers are willing to pay it to somebody who has a degree versus those that don’t, is a function of the demand for college degrees, but also the supply of them, right, and so in a country like the U.S. where many people... many more people than they used to are going to school, but the graduation rate is low and it’s lower at schools that are under-resourced than at elite private institutions like Harvard.
The fact that we’ve essentially, through funding, narrowed the pipeline to a high-quality college degree means that the return of that degree is very high, much higher than it is in other countries. The returns that college premium is much higher in the U.S. than it is elsewhere, and about 50 percent of the total increase in economic inequality since the 1970s can be mechanically explained by the increase in the return to a college degree. It doesn’t do much to explain increasing wealth inequality at the very, very top, but it does a whole lot to explain the difference between people at the 90th percentile of the income distribution and people at the 50th percentile—and that’s a big gap. That gap has grown a ton and a lot of that has to do with the high return to a college degree and the fact that the pipeline to getting a degree is quite narrow.
I think big picture, if you look what can we learn from other countries and then we can learn that if you make these investments, if you increase access to college, that’s going to devalue the college degree and I always say, well, that’s exactly the point. It shouldn’t be the case, but the only way to enter the middle class of America is to get a college degree from a fancy school that very few people can afford. It ought to be that the degree is less... not what’s necessary is that it’s less of an advantage relative to kind of the massive workers out there. We need to democratize access to this important social policy lever, mobility lever.
Q: What are the differences in policies that may produce better or worse results in urban, suburban, and rural areas and across geographic regions such as the Midwest, Northeast and South?
Oh, that’s a great question. There has been sort of a revival in policy circles and interest in place-based policies, and I think for good reason because the issues faced by some parts of the country are very different than others. I think a lot of the policies that have worked quite well in urban areas and education and other fields actually implicitly rely on a reserve army of young people who want to live in cities, who are willing to work for relatively less than what they could earn elsewhere in kind of do-gooder type jobs and they often kind of cycle through those jobs after a few years and so if you’re starting a charter school in New York City, it’s easy to find pretty talented people to staff the school and work 80 hours a week for not much pay. That’s a lot harder to do in Cleveland, Ohio, and it’s even harder to do in the exurbs of other cities or in rural areas.
I think the policy solutions really are quite distinct. I think that for non-urban areas it’s very hard to attract talent and so you have to think very deliberately about a strategy for doing that in a way that’s just not true in urban areas, and I also think in some places the mix of employers is really quite concentrated. If you look at, for example, one of the big things that happened in the U.S. over the past 15 years was the opening of trade with China and that had a really disproportionate impact on communities where there were factories that were producing products that competed with China, right, and so actually one of our new faculty members, Gordon Hanson, was one of the authors of this series of papers about what’s colloquially called... the kind of colloquially called the China shock showing the large impact on some of these communities and the amount of economic distress that was created by the dislocation and the closing of these factories. I think it makes sense to have a set of policies that are regionally focused that are targeted for some places because the issues are quite distinct.
Q: Can you address how changes in unionization may relate to income inequality in the United States?
Sure. It’s a topical question because the Harvard Graduate Students Union is on strike right now. Unionization is on the minds of people on the Harvard campus. I picked a bad semester by the way to teach a 125 person classes because I now have 125 final exams to grade all by myself but that’s okay, you didn’t want to hear about my troubles. Unionization, it’s a very interesting topic because it’s one of the many things that it’s kind of coincidence with the rise in economic inequality since the 1970s.
You see, one of the things we talked about in my class is that you’ve seen there’s enormous increase in economic inequality in the past 35 years. At the same time you’ve seen a bunch of different things happen. The rising returns to education, the slow in growth of tertiary education, decline of unionization, globalization. All these things together—and the challenge is how do you take one fact pattern and figure out what are the different causes and how do they all add up, and we talk a lot about that in class.
I would say my summary sense is that the decline of labor unions has had a big impact on what I would call 50-10 wage inequality, so is the growing gap between the median income and the bottom income because unions traditionally played a very large role for less-educated workers, especially men, propping their wages up to kind of living wage level allowing you to support a family and that had a big impact in sort of, it’s kind of... it’s tough to know about, sort of, to disentangle the role of unions because unionization declined for political reasons at the same time globalization has put enormous pressure on U.S. manufacturers. Manufacturing in the U.S. has been a long-term secular decline, that decline started before the decline of unions and so it’s very hard to know how much of this is really about the changing composition of unions in the U.S. It used to be mostly manufacturing. It’s now mostly public-sector unions and service-sector workers and teachers and health care professionals. Who is unionized has really changed a lot in the U.S. for long-term structural reasons, but I do think it’s something that’s on the tip of everybody’s tongues and I do think we’ll see a lot of policy proposals related to unionization and sort of broadly speaking how to get some of the benefits of unions propping up the floor for workers in the middle of the income distribution as a way to address inequality.
Q: Good afternoon. In the discussion of economic inequality that is certainly dominating lots of conversations and has been mentioned here, and whether it’s inequality of wealth or income. One of the concerns I’m trying to understand better is that two economist from Berkeley, Emmanuel Saez and Gabriel Zucman, have written widely and have had a lot of attention describing very starkly the growth in economic inequality in United States and elsewhere. Recently there have been a number of studies, however, that suggest that there are flaws in the Saez, Zucman work and that they... and there are different approaches but the issue is that the Saez work overstates pretty dramatically the issue that the level of inequality that has developed in the last few decades, the cover article in The Economist in the last week had comments on that. The economist Laurence Kotlikoff at Boston University has published on the topic as well on the shortcomings of the Saez at work. I think that the problem is that we want to be sure we’re measuring the right thing and looking at the right thing before we create policy and if we’re not looking at the problem the right way we’re not likely to get the right policy. My question really is, do you have any headlines you can offer about the Saez work and the various critiques that have emerged on it?
All right. Thank you for your question. It’s a great question and I’m glad you read The Economist article. We actually did talk about that article in my class yesterday. It’s very topical. I also want mention that Emmanuel Saez came and gave the Stone Lecture here at the Wiener Center a couple of weeks ago and so we had a long, extended discussion about some of these methodological issues. Actually, I don’t want to get too much into the weeds, but I do feel like I can give you some sense of the nature of the disagreement and give you my general take on how I think you should interpret it.
The Economist article and the work by Kotlikoff and others, I think point to the very real issue of the assumptions one needs to make when calculating income shares for a very small group of people who is the top 1 percent of the top 0.1 percent, and there are legitimate disagreements and... I think legitimate ambiguity about how one should count things like... a lot of the distinction between the earlier work by Piketty, Saez, and Zucman in the recent article in The New York Times and their book is about some things like... for technical things like how to treat the incidence of the corporate tax. When corporate tax rates declined with the passage of the 2018 Republican tax bill, the question is how much of that decline in the tax passes on to workers in the form of higher wages and how much of it stays with shareholders? And so there’s a whole academic literature. I’m trying to estimate tax incidents and I think what happened was Piketty, Saez and Zucman made one set of assumptions in an earlier set of work and another set of assumptions in their book, which isn’t ideal but I also think it points to the fact that these things are actually really hard to calculate.
My summary sense of how you should take this is that I think The Economist article you kind of kicked up some dust about the substance... like there are some important stuff to this conclusion to draw for whether the top income shares in the 1 percent and above are increasing a lot, increasing a little or only have increased modestly since the mid 1990s, but they’ve definitely increased since the 1980s, which is when the real exposure to inequality started and I think it also... a lot of those issues are much less important for even like the 95th and 99th percentile or the 90th and 95th and you see just the kind of widening up the rungs of the ladder of the income distribution all over the U.S. and in all over other countries at much lower levels of income, and so all those things have nothing... That’s sort of a separate issue from what the debate is about, the numbers that Saez and Zucman used.
I do think stepping back and thinking about policy, the big picture for me, which is that economic inequality, pretax inequality is growing for reasons that are complex but some can be addressed by policy. I think that’s still true and it doesn’t change my conclusions about that. I also think it’s very hard to design policy for a group of potentially 2,500 incredibly wealthy people. I think we should be focusing our policy on a broad range of... basically I think there’s... my broad view about this is I think we need to increase social spending in a lot of areas, especially education to other things like health care, and we need to raise revenue to do that and we should be, as policymakers and as a public, be openminded about ways to do that.
I think actually for me the priority is not so much who we should tax, but it’s what we should spend our money, on and then I would work backwards to say how do we need to raise revenue and what is a fair societal way to do that? That’s how I look at this issue. I think much more on what we need to spend money on rather than who we need to raise the revenue from.
Q: Hi. I’m really fascinated by the research that you’ve done around the importance of social skills as compared to technical fields. As you know, there’s a lot of emphasis and interest particularly from philanthropic organizations in investing in STEM education and corporate foundations investing in STEM programs. I also find that a lot of large corporations are, especially at the entry level, developing algorithms and programs to screen initial job applicants, presumably looking for things like technical and hard skills. This seems to suggest to me that we’re looking at an impending gap in understanding and elevating the importance of the social skills that you call out. Can you talk a little bit about what the implications of that might be and how we can better support educating students on these really important social skills?
That’s a terrific question. Thank you for asking. I want to say pretty clearly that even though I could see how it would be interpreted as such, I don’t think of the work I’m doing as being anti STEM because... and the reason is that when you think about how to develop social skills or socio-emotional skills as people sometimes call them. It’s not really so much about what you’re teaching it’s not like, okay, it’s sixth period, now we’re going to have social skills time. It’s more about the way you teach and the way you design school settings and in some sense you want to design them more like the workplace, like the ways that people interact in every other non-school setting, so you think about like the pedagogical model that I still have where 125 students sit in the classroom and listen to me talk and sometimes ask questions and then turn in individual assignments. It’s just nothing like what they’re being asked to do in the rest of their lives.
That was always true to some extent, but it’s even more true now than it used to be because so much of work is project-based. It’s open-ended, ambiguous, it’s team-based... it involves an uncertain set of goals and so I just think that we need to... I don’t think it’s so much about like, you should major in X rather than Y. I think it’s more that we need to rethink the way we educate in a way that corresponds to the demands of the modern workplace. I mean, that’s just something that I’m working on pretty actively. I have this group at the Malcolm Wiener Center that we’re currently calling the Project for Workforce Development at Harvard. We don’t have a great name yet, but we’ll basically... I’m working with folks at the School of Education and the Business School, and we’re doing a bunch of work studying the different ways that institutions and employers connect and help train people for jobs and help kind of measure what skills people have and think about nontraditional pathways to getting to employment.
I think this is a very... there’s not a lot of academic work in this area. I’m trying to do more of it. I think it’s incredibly important and I think when you... part of it too is just engagement. It’s very clear that when you design educational models that are more interactive, that are more work-based, that are more team-oriented, people just like it more and they stay engaged and so I think that’s a huge part of the story.
Q: Fascinating discussion. I wonder if I could ask you to drill down a little bit more or maybe just real broader, a little bit more on the question of social mobility and the apparent decline, the obvious decline in mobility and thinking that a better understanding of the causes of that, a better understanding of some of the consequences and then interventions that you would have us thinking of that?
Yeah. Great question. I mean, I think mobility is such an important a thing to focus on precisely because it gets at the root of some of the normative questions about inequality that are so important, which is about opportunity, right? It’s possible to look at a static distribution of income in a society and it could reflect very high intergenerational mobility or a very low intergenerational mobility so everyone, let’s say you... as we currently have in the U.S., a society where there are some people who are very well off and some people who are not, that could alternately result from a world where everybody who’s born poor stays poor and everyone who’s born rich stays rich, so very low mobility, or it could be a lot of shuffling around intergenerationally, and I just think normatively and philosophically we think a lot differently about those two scenarios.
I spend a lot of time thinking about mobility in the context of college access, primarily because one of the very striking things we see is if you just compare where people follow the income distribution based on their own parents’ incomes, so you’d say like, if I was born poor, how likely am I to be rich? The U.S. does have, as you say, pretty low levels of mobility even compared to other countries. We’re not really the land of opportunity compared to, let’s say Denmark or other countries in Europe where they have good data to measure this. However, if you look among the set of low-income kids who go to a particular college and high-income kids, then you see much more mobility. If I tell you I’ve got a first-generation college student who grew up in a poor family and goes to Harvard, or not just Harvard, but goes to SUNY Stony Brook, and then I’ve got a wealthy kid who goes to SUNY Stony Brook or Harvard so take the... compare two kids of different family incomes who go to the same college, their income is very similar and so what that tells you is all of a sudden if I condition on the set of people who have managed to get into a particular school and receive all of the education that happens there, then I can no longer tell you who’s rich and who’s poor.
We’ve gone from a low-mobility society to a high-mobility society by virtue of granting access to an institution and so part of that is who are the kids who are born talented, who are able to CLIMB the income ladder? And so one of the things we’re doing with the CLIMB initiative is trying to understand the true impact of colleges on social mobility, and that work is still ongoing but I can tell you it’s very high from what I can see.
I really think... you then work backwards and say, well, what is the actual income distribution at colleges? And it turns out it’s mostly rich kids at colleges that are well resourced. I think that’s one, not the only policy but one where I think we know we have a lever, which is we can change the way we admit students and make the distribution of income at colleges more equitable and we know that’s going to translate into a reduction in inequality because of all the research around the benefits of higher education. I think there are lots of potential tools out there, but to me this is a very obvious one and one like, it’s not a magic formula. I mean it is, once people get to college where we know if we just admit more students, we can do a lot of good and so we at the CLIMB initiative are very focused on this right now but I don’t think it’s only about colleges for mobility, but I do think that’s an important avenue.
Q: Hello, I’m mid-career 1997 and I guess I want to go at this from the side of 3.7 percent unemployment in our country. Presumably 96 out of every 100 people that want jobs have jobs right now in the country, and yet here in East Harlem where I’m working there are so many people that don’t have jobs and especially young people and I think that’s not just East Harlem. I think there’s just something skewed about that number 3.7 percent, when really it’s quite a bit higher than that I think. I’m just wondering on data, on issue of data. I mean, what’s a fix on the fact that the data’s wrong on unemployment? The data’s wrong on our political polling because they’re not reaching all the factors. The data’s going to be wrong on our census. I mean, is there some kind of big thinking going on about how to sort of really recalibrate how to get the right data so we can make the right decisions.
Good question. First off I would defend the data a little bit in part because I am on the technical advisory committee of the Bureau of Labor Statistics and so I know all the efforts that the Department of Labor is undertaking to try to collect accurate data. I would say it differently, I would say, not that it’s wrong, but that it’s incomplete. I think a lot of what you’re describing as I hear it as more that the definition of who counts as unemployed is quite fluid because we have a lot of underemployment. People who are working, let’s say as an Uber driver, but would really prefer to have a full-time job, or people who say... there’s this third category, so unemployed technically means unemployed and not seeking work, right, and so a lot of people say they’re not in the labor force because they’re not seeking work and it could be because they have a disability. It could be because they’ve given up. It could be because they’re retired and so that category has grown too, and of course if you add up people not in labor force and unemployment together, you get a much bigger number than 3.7%. But I do think there’s a larger issue here about... to the extent that the workplace is changing and I do think even though there’s a lot more changes to come, you see a lot of, for example, the gig economy is a big issue and there are quite a lot of our institutions and our data collection strategies that are not equipped to manage these changes.
One of the things we’re doing at the Bureau of Labor Statistics is actually trying to developways to measure, for example, the extent of gig work and to measure things like people who do work... there’s a thing called Mechanical Turk, which is a platform run by Amazon where you can essentially hire people on the spot market to do just a crazy variety of tasks. I could give people 10,000 images and say I’m going to pay you 5 cents an image to tell me which one is a chair. Which one of the images is a chair and so companies actually use these to kind of do the last human mile of a lot of processes that appear from our perspective to be fully automated like when you are on Facebook and you’re scrolling through videos, you think, how do they know what I want to see? And a lot of times there’s a human hired through a platform like Mechanical Turk screening those images, making sure none of them are too explicit, making sure they are what the machine learning algorithm says they are.
That kind of work is really happening under the surface of the economy and we aren’t measuring it well or aren’t measuring it all because it’s not really... when you think about driving an Uber, like are you unemployed? No. Technically because you’re driving your car... would you prefer to have a different type of job? Would you prefer to have more hours? Would you prefer not to work in this gig economy? Would you like a few more security? Yes, and so I think that’s not necessarily that the data are wrong, it is that the data are incomplete and the paradigm for collecting those data is just outdated. That’s one of many policy challenges we face.
Q: Yes. I’m a 1996 MPA. I grew up at a working-poor home filled with domestic violence, so I’ve had PTSD for 60 years, but I’ve also gotten treatment for it for much of that time. I guess my question is, having grown up in a working poor family. Some of the problem isn’t about any of the things you’re talking about. Some of the problem is people have their own interests and care about what they want to do in the world. You’re talking about public spending. The Massachusetts Rehabilitation Commission back in 1969 declared me to be emotionally disordered, which allowed me to get college tuition. Now, you might think that two things don’t work together but for me they were fabulous. I got to Keene State College and I got a degree in teaching, which is what I wanted to do. Now, I didn’t have a long career in teaching, but I had a 40-year career in human services where I had a lot of important jobs and I just had a great life. I think part of what you’ve got to look at is also how to help people figure out what it is they want to do for work and then how to pay to get them the training they need to make that happen.
Thank you for your question and thank you for sharing. I think that what you say is so important and I think it speaks to, I guess a larger philosophy, I would say, I have about policy, which is I think we need to be a society that creates second chances for people and gives people a cushion when they want to take risks and sometimes the nature of risk is that it doesn’t pay off, and what you’re describing is a policy that allowed you to have a second chance or maybe a firsthand or a third chance, I don’t know, but another chance at opportunity and I think what we have too often in this country is that some people grow up in a family or live in a community where they’re allowed to fall down and they’re allowed to fail and they’re not stigmatized by that and they have other second chances, third chances, fourth chances and some people don’t.
Some people have to walk on ice and be perfect to succeed in society, and so for me it’s almost... I mean this is definitely a normative statement, but I think we ought to be one. I think of policy as hopefully providing tools for people, not giving people a career, but giving them a safety net so that they can explore and find things that really drive them, and if they have a disability or if they experience a hardship, it’s not game over for them. They have another chance. I think that’s an imperative for us if we want to be the type of country that I think we should be and certainly the way we talk about America and what it means is the land of opportunity, I think. That’s what we’re talking about. Thank you for sharing.
Q: I have a question kind of back to the nature of the work that’s going to be developed over the course of time. I’m working on an initiative in an inner city where I believe the high school population is not college eligible and not solvable in the next five to 10 years, and so I have an aim at having those kids look harder at jobs which do not require a college degree and you could go all the way to your truck driver’s story. I always use the examples of electricians or plumbers and things like that because I believe most of those jobs today have living wage opportunities for people. As you look at the future of work, am I directing people in a place where those opportunities are likely to increase or will they go the way of the retail jobs? That’s my fundamental question, not that I’m giving up on college as a path, but half of those kids are not going to make it that way and so I’m trying to redirect. I just want to... I’m curious whether my redirection really has long-term sustainability for numbers of jobs.
Right. I understand. It’s a great question. I mean, I think while I say in the long run, we ought to be pushing towards more education for everyone... I understand that but that’s not realistic for everybody given the stage they’re in. Let me give you one specific suggestion. I think that health care jobs are very robot proof and future-proof and the reason is because if you look at something like agriculture, a hundred years ago, 40 percent of all jobs were in agriculture roughly, and now it’s like 2 percent and the reason is we develop these technologies that make agricultural productivity much greater, hybrid seeds and irrigation techniques and machines that can farm more efficiently, and yet we only need so much food as a society and so when we get more productive, we just spend less time, less human effort on those things and that’s great for us, but then there are some things that when we get richer, we want more of it, right.
Everybody wants to live a happier, longer, healthier life and so that’s the kind of thing that’s not going to disappear. If we actually ever do get more productive in health care we’re going to want more of it and so I see no future at all where jobs are eliminated in health care. I just don’t think it’s possible because I think if we develop technologies that extend life and extend health, we just need more workers to help staff that process and so I think health care jobs are recession proof. I think they’re future proof and there are quite a lot of jobs in what I call like allied health professions, so not doctors and nurses but health technicians, health therapists, that are growing, that are in constant demand and the hospitals are the major employer in many non-urban areas, so it also addresses non-urban areas as well, in many hospitals. My wife works at Boston Children’s Hospital and the way they hire is often they will hire people to do very basic jobs like front-office type jobs or back-office jobs and then if they like people and the people are doing a good job, they will pay or subsidize for training so they can learn how to operate the MRI machine or other things like that, and so people can get trained up while they’re working on the job. They kind of gradually move up the ladder and many of these jobs pay a living wage. I would suggest the health care industry is a place where young people can look and get family-sustaining wages without earning a college degree.
Q: Hi, I’m MPP 1983. There’s been growing discussion about reparations for slavery, and I was just wondering whether that’s an issue you have given any thought to as to if something like that were ever to be pursued or enacted. How it might be used, the various ways that could be used to address social inequality and the other issues you’ve talked about.
Yeah. It’s a terrific question and a tough one, not just for sort of economic reasons but political reasons as well. I think that there’s no question in my mind that it is not... that the legacy of racial injustice in this country has very much stayed with us and a lot of that as I was talking about in my class is because of the history. I don’t think we have a lot of policies that are explicitly race-based in this country anymore. They’re outlawed but many things... it’s surprising if you read Richard Rothstein’s book, for example, The Color of Law and think about the history of neighborhood redlining in this country. I mean there were a lot of policies that were very explicitly race-based that restricted the availability of housing and mortgages to white families. Levittown was a government-funded development that was restricted to white people only, and so if you were a white veteran, you could get a highly subsidized mortgage, but if you were black, you were not able to get that. I don’t have either the expertise or a strong view about exactly how that should happen, but I think the idea of reparation, meaning restitution for past wrongs, is absolutely necessary. I don’t know what that means in terms of exact policies but I don’t think... I mean, if you look at other countries that have perpetuated atrocities on some groups of their people, the ones that have successfully navigated that or more successfully navigated it are the ones that have acknowledged it actively, and so I think as long as we continue to pretend that this was all over, we will never solve the problem. I mean, I don’t... what that means for policy, I’ll leave it to other people who are more expert than me, but I think it’s something that we should seriously consider.
Mari Megias:
Great. Well, thank you for that question and answer and thank you to all of our callers and to our speaker, David Deming, at today’s Wiener Conference Call.
David Deming:
Thank you all for having me. Have a wonderful holiday season.