Disability Insurance and Retirement Around the World
In the U.S. and many developed countries, older men's labor force participation declined steeply after WWII, during the same period that social security and disability insurance programs were growing rapidly. Since the 1990s, labor force participation has rebounded in many of these countries. Many countries have also made reforms to their pension programs over the past two decades, such as increasing the retirement age or reducing the generosity of the benefit formula.
What are the effects of pension program provisions on the labor force participation of older persons? For over a decade, the International Social Security (ISS) project has used the vast differences in program provisions across a dozen developed countries as a natural laboratory to study this question. A central finding of the ISS project is that the provisions of social security and related government programs provide strong incentives for workers to leave the labor force at relatively young ages, and that reducing the inducement to leave the labor force can lead workers to delay retirement and yield large improvements in the financial position of government budgets.
In Social Security Programs and Retirement Around the World: Disability Insurance and Retirement - Introduction and Summary (NBER Working Paper No. 20120), Courtney Coile, Kevin Milligan, and David Wise summarize the findings of the most recent phase of the ongoing ISS project, which reflects work by three dozen researchers in the represented countries. This phase focuses on the effect of disability insurance (DI) programs, asking how changing the provisions of country DI programs would affect retirement behavior.
The authors begin by pointing out that the share of men age 60 to 64 on DI varies substantially across countries, from 16 percent in the U.K. to 14 percent in the U.S. to 6 percent in France and 2 percent in Japan in 2009.
Trends in DI participation over time are also informative. In most of the countries, DI participation rose until the late-1980s to mid-1990s and fell dramatically thereafter. The DI participation rate for men age 60 to 64 peaked at or above 20 percent in six of these countries, reaching 36 percent in Sweden in 1993. Since then, participation rates have declined by 30 to 50 percent. The U.S. is a distinct outlier from this trend, as participation dipped briefly after 1980 but has climbed steadily since then.
Are these differences in DI participation across countries and within countries over time driven solely by differences in health? The sheer magnitude of cross-country differences casts doubt on this theory, as does the fact that mortality improved steadily in all countries during the period that participation rates first rose then fell.
Education may be another factor affecting DI participation. In all countries, men in the lowest education group are 3 to 7 times more likely to be on DI than men in the highest education group. In part, this reflects the fact that the less educated tend to be in poorer health. However, even among men of similar health status, education matters - U.S. men in the lowest quintile of health who are high school dropouts are 50 percent more likely to be on DI than are college graduates in similar health.
Finally, the time trends suggest that there may be a relationship between DI participation and employment. In Canada, Denmark, Italy, the Netherlands, Sweden, and the U.K., the employment rate fell as DI participation increased and then reversed course and rose as DI participation decreased.
To explore this relationship further, members of the ISS project team in each country conduct empirical analyses. The first step is constructing a retirement incentive measure that reflects how the provisions of the country's social security, DI, and other programs provide a greater or lesser return to continued work at a given age for each worker. This process incorporates the potential value of retirement via different pathways, for example social security and DI, into a single measure, the "inclusive option value."
Next, the researchers construct a comprehensive health index that aggregates information from over 20 health indicators. This index can be constructed in a comparable way across countries and has been validated by its strong correlation to future health events and mortality. Finally, the researchers estimate models that relate a worker's labor force participation to the inclusive option value and health index measures.
The researchers find that the inclusive option value measure has a strong and highly significant relationship with retirement in nearly all of the countries. The authors then use these results to simulate the effect of increasing the stringency of admission to the DI program. In the U.S., for example, eliminating eligibility for DI for one-third of DI participants is estimated to increase years of work between ages 50 and 69 by 5 percent among current DI participants. Comparable estimates in other countries range from 2 to 13 percent. Thus, the results of the empirical analysis support the theory that making it easier or harder for workers to access the DI program can affect labor supply at older ages.
The authors conclude, "with large increases in life expectancy in all participating countries, there is considerable interest in prolonging work lives. Indeed there has been a large increase in the employment of men in most of the participating countries since the late 1980s and mid-1990s - the same period over which DI participation has been declining in most countries." In future work, the ISS team plans to explore how the capacity to work at older ages varies across countries and over time.
Funding for this project was provided by the National Institute on Aging grant numbers P01-AG005842 and P30_AG012810 to the NBER.
