CHAPTER 12

A Well-Known Problem

Jared Bernstein

NICHOLAS EBERSTADT HAS written an important analysis of a serious problem ailing the U.S. economy. The increasing share of American men disconnected from our labor force costs us economic output and lowers the household incomes of these men. It can also be psychologically damaging, given the importance that work plays in our culture and individual self-esteem.

Nick’s analysis is also timely, coming as it does amid a presidential election in which an allegedly “angry working class” is rejecting establishment precepts, most notably globalization. Certainly, the lack of meaningful, gainful employment is thought to be part of this phenomenon.

MEN WITHOUT WORK IS A WELL-KNOWN PROBLEM

The problem of prime-age men without work is real, but it is not, as Nick argues, underappreciated. Labor economists have long worried about this trend, and the issue broke out of the halls of academia at least twice in the last thirty years. In the 1980s, ballooning trade deficits with lower-wage countries, the loss of U.S. manufacturing jobs, and exit of displaced American production workers from the labor force began to alarm industrial workers, their unions, and local communities. A decade ago, two New York Times reporters wrote on the missing-men problem.¹ It is fascinating and depressing to return to that story, since it contains most elements of Nick’s analysis.

Moreover, as our current recovery has progressed and unemployment dropped, it has been widely recognized that labor force participation has significantly lagged. Economists and commentators from the CNBC crowd to the Federal Reserve have given it wide attention on the first Friday of each month (“jobs day”). As Nick himself notes, President Obama’s Council of Economic Advisers (CEA) recently published a highly regarded piece on the issue.

I mention all this not to nitpick—or Nick-pick—but to make this point: To believe the men without work problem is underappreciated is to risk missing numerous important other debates that the decline of men’s work has inspired—debates about automation, the role of immigration and globalization in displacing prime-age men, and the alarming trends in elevated death rates among white men with some of the same characteristics of the men not in the labor force (NILF) group.

PROBLEMS WITH THE DIAGNOSIS

Nick’s evidence of the trends is largely sound and certainly exhaustive. He exploits many more datasets than is typical in this type of analysis. The longitudinal data he taps are a welcome addition to work in this area.

However, I have a fundamental disagreement with his diagnosis. Nick (and he’s not alone here) insists that the declining trend in men’s work is extremely linear. In fact, the trend is far more cyclical than his straight-line analysis suggests. While there’s no question that trend lines through labor force participation rates (LFPRs) and especially employment rates for prime-age (25-to 54-year-old) men have a negative slope, there is critically important cyclical variation around that trend. That cyclical variation suggests the importance of demand-side explanations, and it is particularly pronounced for less-educated, prime-age men, the very group whom we worry most about in this debate.

Also, the fact that the baby boomers started aging out of the workforce in recent years has the potential to create unnecessary confusion in these data. In fact, most analyses, such as a recent influential paper by numerous Federal Reserve economists, argue that at least two-thirds of the recent decline in the LFPR is due to aging boomers retiring. The top line in Nick’s figure 2.1 excludes retirees while the bottom line does not. Compare the upticks at the ends of the two series. For the twenty-plus series, which includes elderlies, the uptick is shallow. For the other line, prime-age workers, it is steeper, suggesting more of a cyclical response.

Consider, too, the recent positive trend in prime-age male employment rates: The series has regained two-thirds of its recent loss! This trend stands in stark contrast to Nick’s “linear” argument. Figure 12.1 highlights a key point that is underemphasized in Nick’s essay: the cyclicality of prime-age EPOPs. The evidence of strong cyclicality, especially for the least-advantaged/skilled men, suggests less their flight from work and more work’s flight from them.

Acknowledging this cyclicality is important, but it doesn’t undermine Nick’s main point about the negative trend in men’s work or the urgency of the problem. It does highlight, however, my other main disagreement with Nick’s analysis: He believes that the “flight from work” is a more of a “supply-side” problem than is warranted. Though Nick agrees that weak labor demand is a causal factor, he discounts it compared to, say, the CEA’s recent analysis. This leads, I believe, to a serious misdiagnosis.

Nick says virtually nothing about the loss of production worker and manufacturing jobs and the role of our persistent trade deficits in these losses. It is widely accepted that this shift has played a fundamental role in declining male work rates, but the word “manufacturing” comes up only once in Nick’s analysis—in a quote from the CEA.

Rather than fleeing work, many prime-age workers, particularly the less-skilled, respond to the ebbs and flows of labor demand. Nick, while recognizing demand is in the mix, downplays its importance. Part of this comes from his insufficient attention to trends in work by industry and occupation. For example, Nick argues that EPOPs for men behave differently than those for women, and he can’t think of a demand-side phenomenon that would generate a difference. In fact, the shift from production work to services (think manufacturing to health care) in both the United States and other advanced economies is an excellent and well-known explanation.

Nor do I find Nick’s other critiques of the demand-side explanation convincing. He argues that for those without high school degrees, immigrants have steady employment rates relative to the native born, implying that a demand-side shortfall should affect both groups. But figures from the Economic Policy Institute’s State of Working America show that such native-born workers without a high school degree are a small and uniquely disadvantaged group. Among native-born workers, 5 percent lack a high school degree, compared to 26 percent for foreign-born workers. Nick himself underscores this point about the unique disadvantages of this small but troubled group when he points out that criminal history explains the “the remarkable gap in LFPRs between low-skilled prime-age immigrant men and low-skilled native-born men.”

Also, there’s an occupational story at play here, one tied, once again, to labor demand. Compared to less-educated Hispanic immigrants, white and black high-school dropouts have not been nearly as heavily employed in construction. While the employment rates of immigrants were lifted by the sharp increase in demand for construction workers, that was less the case for black and white high school dropouts.

Finally, Nick posits that if demand-side effects were “a truly significant determinant of changes in labor force participation patterns,” regional differentials would tend to diminish following such “shocks” as labor markets sought “equilibrium,” and this hasn’t happened. His assertion is unconvincing. Economists recognize that the nation comprises different regional labor markets that have been hit differently by trends in globalization, technology, the loss of manufacturing jobs, demography, population flows, the employment correlates of regional shocks (e.g., when the housing bubble bursts in a region with much new construction), and more. The variation Nick shows could just as easily be a function of geographical variation in labor demand as its absence.

WELFARE BENEFITS, DISABILITY, AND CRIMINAL JUSTICE

As for the causes of the American male’s flight from work—an ancient left/right argument—Nick believes welfare benefits, disability, and criminal justice factors play critical roles, and gives them more weight than demand-side factors.

First, the evidence simply fails to support as strong a welfare-benefits explanation as Nick claims. It is widely agreed that our income-conditioned welfare system has become more, not less, dependent on work. The incentives increasingly go the wrong way for Nick’s story. The research shows that it is almost always better to work than not to work, that marginal incentives point to the benefits of working more rather than fewer hours, and most importantly, that the safety net as a whole has little effect on the work effort of low-income workers.²

While Nick is on more solid ground regarding disability, both the CEA and recent work by disability expert Kathy Ruffing argue against disability insurance (DI) being a major factor in declining men’s employment rates. According to the CEA, the increase in prime-age men on disability insurance is too small to explain the lion’s share of the decline in work. The CEAs simulations end up assigning less than half a percentage point (out of the 7.5 points just noted) to DI, suggesting it accounts for less than 10 percent of the decline in work by these metrics.

Moreover, suppose, contrary to the CEA’s findings, there were a large rise of prime-age men on disability as employment rates fall. It could easily be that jobs disappeared and these men are using disability insurance as a type of longterm unemployment insurance (a misuse, I would agree, of disability insurance). Weak labor demand could be a reason for higher disability insurance rolls rather than the disability insurance rolls being an explanation for weak labor supply.

Though Ruffing does not look exclusively at prime-age workers, she shows that population growth, aging, and the increase in women eligible for Social Security Disability Insurance explain much of the increase in the disability rolls. She also makes an interesting demand-side observation: while disability insurance applications spiked in the Great Recession, awards did not. This suggests that higher disability insurance rolls (applications) may be a function of weak demand for prime-age men.

What about Nick’s evidence in table 8.1? Here again, there’s a disconnect between his evidence and conclusion. As I read the table, it shows that prime-age men not in the labor force were more likely to receive disability and Medicaid benefits in 2013 than in 1985. But nothing in Nick’s argument shows this results from prime-age men rejecting available employment and taking benefits instead. Absent that, this becomes an argument as to whether NILF men are getting too much in the way of benefits (or the wrong type of benefits).

I do agree with Nick that the large number of prime-age (and younger) men in the criminal justice system is a serious factor in the “men without work” phenomenon, especially among minority men. Incarceration rates have risen partly as a function of lengthy sentences for nonviolent crimes, nudged along with a strong dose of discrimination against nonwhites. Even in periods of strong labor demand, these men would (and do) face high labor market barriers.

WHAT’S TO BE DONE?

Nick’s analysis of the problem is one of the most exhaustive I have seen. His solutions, however, fall short. After several pages of diagnosis, we get minimal prescription. That’s unfortunate, given the welcome urgency Nick brings to the subject.

Nick’s first solution is more entrepreneurialism, though he does not suggest how to achieve this. More business creation would surely help job creation. But economists have few explanations as to why business startup rates are low.

I clearly disagree with Nick’s “work first” ideas. I see no good coming from making these programs less generous or further conditioning them on work. Essentially, I think Nick makes the same mistake that other conservatives make in assuming that, if poor people wanted a job, they could get one. Not every prime-age, nonworking male is poor, of course, but they are disproportionately low-income and less educated, and often face steep barriers to labor market entry, including criminal records and racial discrimination. Researchers at the Georgetown Center on Poverty and Inequality recently published a comprehensive examination of subsidized employment programs that have consistently seen high rates of voluntary participation over the past forty years. They report that the “number of disadvantaged people willing to work consistently exceeds the number in competitive employment.”

Policy can achieve stronger labor demand conditions. The CEA offers several recommendations (expanding the Earned Income Tax Credit for childless workers, higher minimum wages, paid leave, and child care assistance). The agenda I lay out in my recent book, The Reconnection Agenda, emphasizes pro-work/full employment fiscal and monetary policies, lowering the trade deficit and, perhaps most important here, direct job creation. As long as there is a surfeit of takers for such programs and they have a good track record, I see no downside at all to significantly expanding their footprint with resource streams. Perhaps all of us who are concerned about men without work can agree that this is a useful way forward, even as we debate the relative roles of supply-side and demand-side factors.

Finally, those on all sides of the political aisle agree that criminal justice reform is integral to solving the men without work problem. Such reforms must focus both on reducing individuals’ exposure to the criminal justice system and mitigating the effects of that exposure, through policing and sentencing reforms, expanding alternatives to incarceration, and “fair chance” hiring policies that protect these men from labor market discrimination.