An odd thing has been happening since the middle of 2014, according to a recent analysis of Census Bureau data from the economist Ernie Tedeschi. The number of prime-age (25–54) Americans who are out of the work force, and say they do not work because they are disabled, has declined about 7 percent, reversing a trend that had been in place for decades. The disabled have become more likely to rejoin the work force and less likely to leave it to begin with.
The Social Security Disability Insurance (SSDI) rolls show a similar pattern, peaking in May 2014 — the one month in history they topped 11 million, having almost tripled since the mid 1980s. As of last month, they stood at about 10.3 million, and the shift in the trend line has been so dramatic and unexpected that it’s caused confusion about when, exactly, the beleaguered program will run out of money. A 2016 report from the program’s trustees predicted that 2023 would be the year of doom; last year’s report put it at 2028; this year’s, 2032.
Either working-age Americans got a lot healthier all of a sudden, or in reality there are jobs that can accommodate a lot of these individuals, especially when the economy is booming, unemployment is about as low as it’s gotten in half a century, and companies are scrambling to find help.
These new developments, aside from being great news in themselves, are just the latest pieces of evidence for a proposition that ought to be common sense: “Disability” is a category with fuzzy boundaries. Many people have ailments that pose an obstacle to work but need not prevent them from working entirely. And this is a reality that our disability programs largely fail to recognize.
At minimum, this failure reduces work for a significant share of enrollees who are capable of it; at most, it may have substantially contributed to a shocking, decades-long decline in labor-force participation among prime-age men (and, since the turn of the century, prime-age women too). Either way, it needs to be fixed — even as, for now, a strong economy is pulling many off the disability rolls all by itself.
In theory, SSDI — the nation’s biggest disability program but not the only one, focusing on those below retirement age and with a significant work history — isn’t supposed to cover borderline cases at all. To qualify for benefits, people must be unable to perform any “substantial gainful activity” that would pay even $1,180 a month ($14,160 per year, roughly the income from a full-time, minimum-wage job), whether in their previous field of employment or in another field they could plausibly enter. In addition, their condition must be expected to last at least a year or until death.
It is inevitable that such decisions will be subjective to an extent, though — and since the program’s standards were loosened in the 1980s, ailments that are difficult to verify have accounted for a rising share of awards (today more than half of SSDI recipients have mental-health or musculoskeletal conditions such as joint pain), and the judgments of applicants’ personal physicians have taken on a greater role in the process. The share of working-age adults who were on the program grew from 2.2 percent in the late 1970s to 4.4 percent in 2013. Population aging does account for much of this growth, but far from all of it.
Experts broadly agree that disability programs can discourage work. In a review of the evidence earlier this year, the University of Maryland economists Katharine G. Abraham and Melissa S. Kearney drew attention to several rigorous studies of the question — in particular, those looking at situations where applicants were assigned to disability examiners or appeals judges who varied in their leniency, allowing researchers to see what happened when similar applicants were denied or given benefits.
One study suggested that, for the 23 percent of applicants with marginal health conditions — meaning they might get benefits or not, depending on how lenient their examiner was — receiving benefits reduced employment by 28 percentage points. This implies that for every 100 people in this group granted benefits, 28 go on not to work who otherwise would have.
There is strenuous disagreement, though, about exactly how much disability programs are to blame for the overall drop in labor-force participation among prime-age men. In 1967, just 3 percent of these men were neither working nor looking for work: not “unemployed,” but out of the labor force entirely. In 2016, that number was 12 percent. And while women flooded into the work force in the second half of the 20th century, since then they too have been slowly leaving.
In Men without Work, his 2016 book on the topic, Nicholas Eberstadt of the conservative American Enterprise Institute noted survey data from 2013 indicating that 57 percent of prime-age men out of the work force received disability benefits, whether from SSDI or other programs — a figure that had risen from 38 percent in 1985. (People tend to underreport their government benefits in surveys, so both of these numbers are likely low.) As correlation does not prove causation, however, Eberstadt hedged his argument, claiming that whether or not rising disability receipt caused men’s flight from work, it has financed it.
In a recent report for the libertarian Mercatus Center, Scott Winship took a different approach to the issue, looking at a Census Bureau survey question asking why prime-age men hadn’t worked during the previous year and checking to see which explanations had risen the most over time. More than a third of the decline of work since 1968 was nothing to worry about, resulting from men becoming more likely to be in school at later ages, to stay home to care for family, or to retire early. But 47 percent of the decline came from an increase in men saying they didn’t work because they were disabled or sick.
Winship pushes aside several explanations for the trends. Contrary to some claims from the left, wages for low-skilled men are not declining once inflation is properly accounted for — and just 9 percent of the drop comes from men who say they can’t find a job — so it’s not that there just isn’t enough demand for these men’s skills. It’s also unlikely that men increasingly face health problems that make them unable to perform the available jobs, as workplace injuries are down, exercise has increased, rising obesity doesn’t cause much labor-force dropout, and employed men are taking less sick leave than they used to.
One plausible explanation, Winship suggests in an accompanying essay, could be that “the payoff for not working has increased,” including payoff through disability programs. This finds some support in a 2013 analysis from the Congressional Research Service, which noted that SSDI awards have grown faster than wages for low earners.
Abraham and Kearney, though, present a very different analysis, roughly echoing earlier findings by the Obama administration’s Council of Economic Advisers. Extrapolating from the research on SSDI and a disability program for veterans — including the above study estimating a 28 percent employment drop for 23 percent of the SSDI-applicant population when granted benefits — they estimate that the expansion of these benefits caused a 0.2-percentage-point decline in the employment-to-population ratio after 1999. (This ratio, as the name implies, measures whether people are employed, regardless of whether they’re looking for work. It includes both sexes and everyone 16 or older.) The overall ratio declined 4.5 points, or 3.6 points after changes in the age and sex composition of the work force have been accounted for, so the benefits were not a major contribution to labor-force dropout. Abraham and Kearney charitably call it “not inconsequential.”
Far more important were the kinds of demand-side factors that Winship pooh-poohs: expanded trade with China, which Abraham and Kearney credit with a full point of employment decline, and increased automation, which gets more than half a point.
Whether disability reform can meaningfully combat the decline of work or would just help the share of applicants who can work but struggle to, it is imperative. The current system is structured in a dysfunctional way — in particular owing to the “all or nothing” nature of benefits — and struggling financially as well. In 2015 Congress was forced to transfer money from Social Security’s retirement funds to SSDI, and as mentioned above, it will run out of money again in 2032.
Last year, the Washington Post profiled an Alabama man struggling with the question of whether to apply for disability. He in fact had health problems but was still able to work and was trying to find a job. And in cases such as this one, the decision to apply for disability is highly consequential. As the Post noted,
most people aren’t employed when they apply for disability — one reason applicant rates skyrocketed during the recession. Full-time employment would, in fact, disqualify most applicants. And once on it, few ever get off, their ranks uncounted in the national unemployment rate, which doesn’t include people on disability.
The decision to apply, in many cases, is a decision to effectively abandon working altogether. For the severely disabled, this choice is, in essence, made for them. But for others, it’s murkier. Aches accumulate. Years pile up. Job prospects diminish.
In other words, after the difficult decision to apply, the next choice is simple: You can’t work, because by doing so you’d prove you are capable of working. This begins during the lengthy approval process; even rejected applications typically take one to three years with appeals. And this period, often right after someone becomes disabled, happens to be the most important time. It’s an opportunity to avoid a spell of joblessness on one’s résumé and work out accommodations, if possible, with an employer in one’s field. Then, once an individual is accepted to the disability rolls, earning beyond the modest “substantial gainful activity” threshold can trigger a review to ensure he still qualifies for the program.
To be fair, it’s far from true that SSDI does nothing to encourage work. Beneficiaries are periodically reevaluated to ensure they are still disabled, with more frequent reviews for those whose conditions are expected to improve. Those who are self-motivated to get back into the work force have opportunities as well, including a “trial work period” in which they can earn money without losing their benefits, an “extended period of eligibility” in which their benefits can be restored after leaving the program for a job that didn’t work out, and voluntary rehabilitation. Light work, providing income that falls below the “substantial gainful activity” threshold, is also allowed.
But there are plenty of aspects to the program that should operate differently, to ensure that those able to work have the incentives and assistance they need to do so. They range from obvious fixes, such as tightening up the definitions of mental-health ailments and speeding up the application process, to far more technical tweaks. Eventually the system needs an almost complete overhaul, though changes should likely begin with an extensive series of pilot programs to test the reforms, as the Trump administration has proposed in its 2018 budget.
One of the most common suggestions is to award temporary or partial benefits to those who face genuine obstacles to work and yet are not completely unemployable. In a report late last year for the American Enterprise Institute, Tejesh Pradhan and James C. Capretta proposed such a benefit that would be smaller than traditional SSDI payments (which vary case by case but average about $1,000 a month) but would not disappear as quickly when beneficiaries earned more at their jobs.
Pradhan and Capretta would also mandate physical therapy or treatment for applicants who might benefit from it — before their applications are even considered. By the same token, the disability pilot programs envisioned in Trump’s budget proposal focus “on early intervention return-to-work initiatives that would help the individual worker maintain attachment to the labor force.”
A more controversial idea, though one with some bipartisan support in the think-tank world, is to “experience-rate” employer taxes: Whenever a worker went on disability, his former employer’s taxes would go up. (A similar rule is already in place for unemployment and worker’s-compensation benefits.) This could be handled through the existing payroll-tax system or, as proposed in a 2010 report from the liberal Center for American Progress and the Brookings Institution’s Hamilton Project, by bringing in private disability insurers, who would have an obvious incentive to get people back to work.
The idea is to give employers an extra reason to keep the workplace safe and to accommodate injured workers so they don’t go on disability. A risk, however, is that employers would refuse to hire workers who are more likely to go on disability in the future. And there’s a question of fairness as well, because SSDI covers all disabilities that keep people from working, not just those stemming from injuries in the workplace itself. The business lobby would likely push back, but it is an idea worth exploring, considering that the Netherlands has seen enormous success with reforms based on the concept of making employers more responsible for their workers’ disability claims.
The Committee for a Responsible Federal Budget, a bipartisan think tank focused on reducing the deficit and debt, has also dived into the debate, publishing an ongoing series of detailed proposals to shore up the program. Beyond ideas broadly similar to some of those outlined above, suggestions thus far have included subsidizing temporary employment to ease those with disabilities back into the work force, as well as limiting the “double payments” that occur when people are eligible for unemployment in addition to disability insurance.
We are, obviously, far from figuring out exactly what the ideal disability system would look like. We can be confident, though, that it bears little resemblance to our own — and that dramatic improvement should be a priority.