Friday, September 29, 2017

Interview with Lawrence Katz: Inequality, Mobility, and More

Douglas Clement has a characteristically excellent "Interview with Lawrence Katz" in The Region, from the Federal Reserve Bank of Minneapolis, published September 25, 2017. The subheading reads: "Harvard economist on the gender pay gap, fissuring workplaces and the importance of moving to a good neighborhood early in a child’s life." The interview offers lots to chew on. Here, I'll just pass along some of Katz's thoughts on a couple of points. 

One issue is about the causes of rising inequality of wages. Katz (together with co-author Claudia Goldin) have argued that the most important reason behind rising wage inequality is that in the race between rising demand for skilled labor and rising supply of skilled labor, demand has surged ahead. The implication is that the appropriate long-run response to address inequality would be to aim at a dramatic increase in the share of Americans receiving higher education. However, others have emphasized other issues that might relate to the role of inequality, like effects from increased international competition, or have pointed out that the rising incomes of top corporate executives doesn't seem to have an obvious link to a shortage of skilled college-educated workers. In this interview, Katz argues that rising demand for skilled labor remains the primary cause of greater wage inequality, and that a substantial increase in the share of Americans receiving quality higher education from the public sector is the appropriate answer.  He says:
"In 2016, the college premium continued to grow. This other chart is sort of my favorite. This tries to put it all together for 200 years. ...

"There is no systematic survey or data prior to the 1940s census on wages. But there was one very important employer in the United States who kept good records going back to the 1820s: the U.S. military when they were hiring civilian employees at forts all around the country. They would hire a blacksmith or clerk or day laborer and write down the employee’s characteristics and the wage they were paid and look at the gap between, for example, a clerical worker and a production worker working at a fort.
"The chart starting with the 19th century skill premium data based on the military wage series put together by the economic historian Bob Margo shows that there was a rising skill premium (that fed into the high school movement) prior to the big decline in the early 20th century. Our best evidence suggests that with the shift from the artisanal shop to larger and more mechanized factories, there was a big demand—with the growth of scale of individual enterprises—for engineers, managers, accountants and clerks relative to craft workers and even relative to other production workers. ...
 
"If you look at the past 30 years, there’s been a big increase in inequality, which you can measure by the Gini coefficient. You can do this for individuals, for family households. A big part of that increase is at the very upper end—which I won’t claim is largely the education return. But even if we exclude the top 0.5 percent [of the income distribution], there are very large differences; we estimate that, as recently as 2013, about two-thirds of that [increase in inequality] is due to the growth of the educational wage premium. Almost all of that is post-secondary college and post-college. So, if you’d kept the college premium at the 1980 level, you would’ve seen only a third as much of the growth of U.S. earnings inequality. ...
"What the government has done—in the ’50s and ’60s, even into the ’70s—is invested heavily in high-quality colleges. Think of University of California campuses or Florida State. But since then, there’s been very little investment in expanding quality higher education. There’s increased crowding at community colleges and state universities, and states have greatly cut back on appropriations for higher education, particularly in the Great Recession.
"The federal government has continued to have an important role, but it’s done it with flexible support through Pell grants targeted to low-income students. The problem is that we’ve had a surge of really low-quality colleges, and the worst of that is the for-profit sector, which Claudia, David Deming and I have studied. Particularly from the late ’90s to 2011 with this very large wage premium and funneling more federal funding into loans and Pell grants, a big part of that marginal growth—particularly for disadvantaged individuals—was at for-profit institutions for both associate’s degrees and bachelor’s degrees.
"It’s been a bit of a disaster. Even though these for-profit institutions have tried to be up to date, very flexible, with high-quality online instruction, we have repeatedly found very little economic return to degree programs at for-profit institutions; instead, it’s become a massive debt trap. I think there is something to be said for the quality and capabilities, the faculty, the peer effects of a traditional public or private nonprofit university.
"So, rather than what would’ve been the equivalent of the high school movement—developing more University of California campuses or more Florida public universities, so we weren’t rationing access to quality public colleges—we allowed the for-profit private sector to come in both as a nimble creative but also as an agile predator."
A second subject involves a classic social science question: to what extent are people with low-incomes held back by living in a primarily low-income community? If instead low-income people were integrated into a middle-income or high-income community, with the schools and public services of that community, as well as the peer effects and social connections, how much difference would it make. The Moving to Opportunity study is based on a social experiment in which some low-income people had the incentive and possibility to relocate to a higher-income community. Katz is Katz is principal investigator for this experiment, and he describes it in this way:

"The demonstration program ran in five cities: Boston, Baltimore, Chicago, New York and Los Angeles. And the families eligible were living in public housing in the highest-poverty census tracts with an over 40 percent poverty rate.
"Through a lottery, MTO allowed people to jump the queue to get housing vouchers providing housing support to live in an area of their choice. There were three groups: an experimental group that had to move to a low-poverty area if they wanted to move and received mobility counseling assistance as well, a second treatment group that received vouchers but wasn’t required to live in a low-poverty area, and a third group, the control group, who stayed where they were and kept their regular housing support. And we’ve been tracking them for 20 years.
"The first question was whether we could actually get a significant number of families to move to low-poverty neighborhoods even with these vouchers. We thought that about a third of them would be able to make such moves. But, in fact, through the hard work of these families and the counselors, almost half of the experimental group families leased up in a low-poverty area. ...
"And what we find for Moving to Opportunity is in the short run it clearly made the adults in the families happier and healthier. Measures of well-being and safety improved. There were big reductions in exposure to violence. But, economically, nothing much changed for the parents.
"But it’s significant that at the time of the moves, at baseline, we asked people why they wanted to move, and very few said they were moving because they were looking for a better job. Almost always it was safety, wanting to get out of more violent areas, worried for their kids. Or it was trying to improve their housing conditions.
"We saw huge improvements one year out, five years out, 10 to 15 years out in adult health. Large reductions in obesity, in depression, in diabetes and in biomarker indicators for long-term stress. This is sort of the equivalent of your best antidepressant and your best exercise and diet program in terms of long-run improvements in adult health and mental health!
"So until we did the latest study on long-term effects on the kids, the message from MTO has been that there were huge benefits to a family’s well-being and health, but not much economically, and we weren’t finding much on test scores for kids. We had a little bit of evidence of positive things for girls and, if anything, the boys were looking a bit negative. ...
"When we looked at those same kids who didn’t look that different in their school performance but were less involved in violent crime as adolescents, they seemed to be assimilating types of social capital that may not show up on standardized tests, a sort of savvy of living in a different type of neighborhood. As Chetty, Hendren and I observe, the younger MTO children becoming adults, we see them more likely to go to and persist in college, and we see them much more likely to work, and an almost 40 percent impact in their mid-20s on earnings. It really looks like a powerful impact."
An intriguing and important lesson here is that getting the full story can often involve looking at a broad array of outcome measures. For example, the parents of the low-income families didn't achieve higher incomes, but their stress levels and health improved dramatically. The children from the low-income families didn't have gains on school test scores, but they were less likely to be  involved in crime, more likely to attend college, and more likely to work. For the typical young person, often it's not that their test scores and academic achievement need to be sky-high; after all, there's no escaping the arithmetical reality that the average student is going to have average scores! But young people who have social connectedness, a sense of possibility, and persistence to keep learning and growing can take their average academic performance and turn it into a successful adult life.