The National Bureau of Economic Research declared that the U.S. recession ended in June 2009, yet 2012 didn't look like much of a resurgence. To me, the number that sums up the year's doldrums is the 1.27 million increase in the number of disabled Americans without jobs. There are now 8.8 million workers receiving disability payments from Social Security. I find this number haunting.

The disabled are part of the far larger number of Americans who have left the labor force altogether since the recession, and who don't seem to be coming back. About 88.9 million people in the United States are now out of the labor force, 2.4 million more than a year ago and 11.4 million more than in 2006. Thirty years ago, there was a 40-to-1 ratio between the total labor force and those workers receiving Social Security disability payments. Today that ratio is less than 18-to-1.

In November 1982, unemployment hit its postwar high of 10.8 percent, far higher than the current rate of 7.7 percent. But the total share of workers who are either unemployed or receiving disability payments from the government totals 12.6 percent today.

The steady rise in disability claims presents something of a puzzle. Medicine has improved substantially. Far fewer of us labor in dangerous industrial jobs like the ones that originally motivated disability insurance. The rate of deaths due to injuries has plummeted. Behavior that can cause disability, such as alcohol use and smoking, has declined.

The aging of the baby-boom generation is often cited as one explanation for the rise in disability insurance rolls. Yet economists Mark Duggan and Scott Imberman estimate that "this factor can explain just 15.5 percent of the growth in the likelihood that a nonelderly adult male receives DI benefits."

The two main alternative hypotheses are that work has become less attractive or that disability insurance has become more attractive and available. The disability-claims approval process and society itself have become more accepting of people receiving the benefits even if they have no visible ailment.

Duggan and Imberman argue that changes in the award formulas for recipients have made disability substantially more generous for poorer workers. For example, a male worker who is 30 to 39 and in the bottom 25th percentile of earnings distribution could expect disability insurance to pay 41 percent of his previous earnings in 1984 and 49 percent of his previous earnings in 2002. An older study examined a Canadian experiment and found that as disability became more generous, the overall labor supply declined. Duggan and Imberman say that changes in availability can plausibly explain about one-quarter of the increase in the insurance rolls from 1984 to 2002.

The economists say that the most important cause of the increasing number of recipients is the loosening of eligibility criteria. In 1984, Congress "shifted the criteria for DI eligibility from a list of specific impairments to a more general consideration of a person's medical condition and ability to work." As a result, the typical disability recipient today is far less likely to have an easily verifiable ailment.

Seventy percent of workers receiving disability assistance fall into three large diagnostic groups: mental disorders (about 32 percent), musculoskeletal system and connective tissue (29 percent), and nervous system and sense organs (about 9 percent). The growth in these conditions explains more than 80 percent of the rise in disability payments since 1996. These medical gray areas can involve a judgment call.

I believe that recipients of the aid are typically in pain, but many have a choice between suffering at work and going on disability. In boom times, the work option may seem more attractive. But as labor options contract, a steady check from the federal government can seem the better choice. The recession surely explains much of the 24 percent increase in the number of people receiving Social Security disability insurance since 2007.

The big fear is that those who went on disability during a downturn may stay out of work forever, even as the economy recovers. The combination of a generous welfare system and a sharp recession is what led to long-term unemployment in Western Europe starting in the early 1970s.

Ten years ago, Alberto Alesina, Bruce Sacerdote and I wrote a paper trying to understand why Americans worked so much harder than Europeans. Those differences are starting to look less stark.

The political response to the low labor-force participation is to offer policies aimed at job creation. The right favors tax cuts. The left favors subsidized help through building roads and other infrastructure, and favoring particular sectors, such as clean energy.

I'm skeptical that either approach can succeed, especially in helping the lower-skilled. Most sensible government investments require well-trained employees, such as engineers, not high-school dropouts. Even if tax cuts help private-sector activity, there is little guarantee that businesses will start hiring from the bottom of the labor market.

We can do better by replacing the archaic system we have now with ones that offer better incentives to work. We have too many distinct programs (unemployment insurance, food stamps, housing vouchers) that penalize earnings, collectively creating a huge implicit tax that discourages work among the poor.

Duggan and David Autor, a Massachusetts Institute of Technology economist, have proposed reforming disability insurance in ways that encourage companies to keep employing disabled workers and impose higher costs on those that produce more disabled workers.

Ultimately, the best recipe for fighting poverty is investment in human capital. This starts with improving our education system, so it provides marketable skills to less-advantaged Americans. But human capital is also acquired at work, and that's why unemployment is so dangerous. When workers leave the workforce, their skills depreciate, and they end up locked out of prosperity.

As we reform the tax code, we must focus on providing stronger incentives to work, through the earned-income tax credit and reductions in the payroll tax for poorer Americans. The future of America depends on preventing a temporary economic crisis from becoming a permanent labor market catastrophe.

------

Edward Glaeser is an economics professor at Harvard.