When the government denies disability benefits to applicants who are married, spouses step up and earn more. But single individuals saw their incomes decline, managing to earn back only a small share of the denied benefit.

This is the finding of a new working paper by Magne Mogstad, professor in economics, who studied the appeals of people denied government disability insurance (DI) benefits in Norway. Decisions are based on whether or not applicants can be reasonably expected to return to work. For each dollar not granted to an unmarried individual, his or her annual household income was reduced by as much as 90 cents.

Mogstad also identified the degree to which single individuals are at higher economic risk after DI denial because they don’t have a spouse who can contribute to household earnings.

This research provides some of the first clues about how individual DI benefits function in the economic lives of recipients. The study suggests that DI provides the intended insurance against excessive financial hardship for single recipients. But, married recipients may be less in need of the benefits because of their spouse’s ability to offset the loss. The paper was co-authored by David Autor of Massachusetts Institute of Technology and Andreas Ravndal Kostol of the University of Bergen.

These findings may help policymakers better assess the balance between the public financing of costs associated with providing DI insurance and the more general benefits to society and the recipients. As Mogstad documents, payment of disability benefits has become an increasing component of the safety net provided by industrialized nations, but costs have risen with it. In the U.S., for instance, the percentage of nonelderly adults who receive disability payments jumped from 1 percent to 5 percent since the 1960s. In the UK and Norway, the covered population grew even more, by 6 percentage points and 8 percentage points, respectively.

“These increases are dramatic,” remarks Mogstad. “So, we are trying to learn more about what caused this large increase in the disability insurance rolls and what the consequences are.”

Mogstad also notes that very few DI recipients leave the program except to go into retirement pension programs instead, so the numbers are unlikely to fall any time soon.

Some countries have tried to limit the growth of their disability programs by tightening their screening criteria. Mogstad believes that these “enhanced gate-keeping policies” can reduce DI costs to taxpayers by lowering the total number of DI recipients and by increasing tax revenues if rejected applicants return to work. But, he says, stricter screening of DI applicants may also result in net welfare losses for individuals and families who value DI at more than its fiscal cost. Furthermore, some appellants who are denied DI end up receiving financial aid through other publicly funded programs, so measuring precise impact of denial to the program is still a challenge.

“We are starting to think about what the insurance value of the program is,” said Mogstad. “It is called disability insurance, so presumably governments started offering it because we think it might be important to people and families who have a working condition…We find there is sensible insurance value to disability insurance but most of this is concentrated among single people.”

Mogstad reached these conclusions by examining a unique aspect of Norway’s disability insurance approval system: applicants who were denied DI benefits are randomly assigned to judges for their appeal. This random assignment allowed Mogstad to correct for bias and consider the disability applicants somewhat uniformly. Many suffer from difficult-to-verify disorders, such as back pain, and all were under the age of 62. This means Mogstad was learning about those applicants on the margins of qualifying for DI, which might be a starting point for considering how to shrink or better manage the program.

Mogstad notes that simply changing DI qualification rules to benefit single disability recipients over married ones would need to take into consideration how this might create an incentive for marriages to form or dissolve.

This research builds on Mogstad’s earlier work on disability insurance. He previously examined how DI recipients change their work decisions due, in part, to their enrollment in the program and why participation in DI programs sometimes occurs within families from generation-to-generation. Mogstad says now trying to understand the costs and benefits of the program is the natural next step.

—Jennifer Roche