Understanding the Increase in Disability Insurance Spending

Jeffrey Liebman

NBER Disability Research Center Paper No. NB 13-01
Issued in January 2013

The share of working age Americans receiving disability insurance benefits has increased substantially in recent years. Some experts attribute this increase primarily to more lenient eligibility criteria and to labor market changes that have increased the incentive for low-wage workers to apply for benefits. Other ex-perts suggest that the aging of baby boomers into prime disability-benefit-receiving ages and the increased labor force participation of women can explain most of the increase. In order to distinguish among these perspectives, this paper decomposes the increase in spending on Disability Insurance (DI) into the share attributable to different factors and conducts simulations of the DI caseload under counterfactual scenari-os in which various combinations of these factors are held constant. There are five main findings. First, while the share of the working age population receiving benefits has risen substantially in recent decades, average benefits relative to GDP have fallen. The net effect is that spending on DI and SSI benefits in-creased by only 0.20% of GDP between the late 1970s and the years immediately preceding the 2007-2009 recession. Second, for men DI spending as a share of GDP was at approximately the same level in 2007 as it was in the late 1970s. Incidence levels fell in the late 1970s and early 1980s and partially re-bounded in the late 1980s, but male incidence has been flat since 1991. Measured from 1985, the rise in the beneficiary-to-population ratio among men is attributable to a combination of population aging, a de-cline in mortality, and an increase in incidence. Measured from 1991, nearly all of the rise is attributable to population aging and a decline in mortality. Third, spending on DI benefits for women has increased by almost 0.15% of GDP over the past 30 years. Almost all of the increase is the result of population aging, an increase in fraction of women with earnings sufficient for them to be covered by DI, and a significant increase in the incidence of DI receipt, with female incidence now approximately equal to male incidence. Fourth, there has been a significant decline in incidence related to impairments from circulatory and neo-plasm conditions. This has been offset by increased incidence related to musculoskeletal and mental con-ditions. Absent the rise in the incidence of musculoskeletal and mental conditions, the number of benefi-ciaries today would be 21 percent lower. It is unclear to what extent the rising incidence of musculoskele-tal and mental conditions reflects more lenient eligibility criteria and increased incentives for low-wage workers to claim benefits versus a reclassification of circulatory cases as musculoskeletal/mental cases. Fifth, incidence rates adjusted for age and the national unemployment rate have been flat for men since 1991 and have exhibited only a slight upward trend for women. Because this constancy has resulted from large underlying changes that have happened to balance each other out, it would not be surprising to see incidence either rise or fall going forward.

PDF (739 K)