Economists have long struggled to explain why a growing proportion of men in the prime of their lives are not employed or looking for work. A new study has found that nearly half of these men are on painkillers and many are disabled.

The working paper by Alan Krueger, a Princeton economist, casts light on this population, which grew during the recession that started in 2007. As of last month, 11.4 percent of men between the ages of 25 and 54 — or about seven million — were not in the labor force, which means that they were not employed and were not seeking a job. This percentage has been rising for decades (it was less than 4 percent in the 1950s), but the trend accelerated in the last 20 years.

Surveys taken between 2010 and this year show that 40 percent of prime working-age men who are not in the labor force report having pain that prevents them from taking jobs for which they are qualified. More than a third of the men not in the labor force said they had difficulty walking or climbing stairs or had another disability. Forty-four percent said they took painkillers daily and two-thirds of that subset were on prescription medicines. By contrast, just 20 percent of employed men and 19 percent of unemployed men (those looking for work) in the same age group reported taking any painkillers.

The connection between chronic joblessness and painkiller dependency is hard to quantify. Krueger and other experts cannot say which came first: the men's health problems or their absence from the labor force. Some experts suspect that frequent use of painkillers is a result of being out of work, because people who have no job prospects are more likely to be depressed, become addicted to drugs and alcohol and have other mental-health problems.

While it's hard to generalize across a large group of people, it's clear that job market changes can have significant health effects on the labor force. Increased automation and the offshoring of jobs have hit men with less than a college education particularly hard. Add to that soaring levels of prescription opioid addiction in the general population, and the result of the Krueger study becomes less surprising.

More research is clearly needed. In the meantime, some things could be done to help workers who've given up. Congress could appropriate money for the opioid addiction treatment and prevention programs they authorized in July. And federal and state governments could focus economic initiatives where long-term joblessness is highest, especially in the South, Southwest and Midwest.

FROM AN EDITORIAL IN THE NEW YORK TIMES