Mortality, Health, and Disability Insurance Around the World

High and rising expenditures on disability insurance (DI) programs are a major concern in the U.S. and many other developed countries. The proportion of men collecting disability benefits at older ages varies greatly across countries - for example, more than 35 percent of 64-year-old men in Sweden and more than 25 percent of those in the Netherlands are on DI, versus 10 percent or less in Belgium, Italy, and Spain. Does this reflect differences in the underlying health status of older individuals in these countries? Or do differences in the provisions of the DI systems explain this variation in DI take-up rates?

These and related questions are explored in Social Security and Retirement Around the World: Mortality and Health, Employment, and Disability Insurance Participation and Reforms - Introduction and Summary (NBER Working Paper 16719). As editors Kevin Milligan and David Wise explain, this is the introduction to the fifth volume in the NBER's ongoing international social security project. This project involves a team of analysts in twelve developed countries conducting comparable analyses of their own country's social security program, effectively treating the vast differences in program provisions across countries as a natural laboratory to study the effect of these provisions on labor force participation.

The authors begin by considering changes in mortality over time and the relationship between mortality and labor force participation. Mortality is of particular interest as a health measure because it is comparable across countries and over time within countries. The authors find that mortality at age 65 was fairly constant in the 1950s and 1960s but began to decline after 1970, with those countries that initially had the highest mortality rates experiencing the most rapid improvements. The variation in mortality across countries is small relative to the variation in labor force participation, suggesting that differences in health may explain relatively little of the cross-country differences in labor force participation. The authors also show that cross-country mortality rates diverge as mortality increases.

Mindful that mortality is not the only health outcome measure of interest, the authors next explore how trends in mortality compare to trends in self-assessed health within a country. In general, there is a strong relationship between the two trends, though there are large differences in the level of self-reported health across countries. The individual country analyses broaden this to examine other indicators of health status, the availability of which varies greatly from country to country.

Next, the authors consider whether trends in DI participation are related to changes in mortality and other health measures within a country over time. They find a very weak relationship between the two, leading them to conclude "DI insurance reforms are largely a train on their own track and not endogenously determined with respect to health."

Finally, the authors consider "natural experiments" in several of the countries, in which DI reforms were enacted for reasons other than changes in the health or employment behavior of older individuals (e.g., by a court decision). In Canada, for example, concern about the cost of the DI program led to the removal of socio-economic considerations in eligibility determinations and a new emphasis on self-sufficiency and returning to work, among other changes. After the enactment of the reform, the share of men age 60 to 64 on DI fell from 14 percent to under 8 percent, a decline of nearly half. A key lesson from many of the reforms is the importance of substitution between programs for older workers (for example, between DI, unemployment benefits, and early retirement benefits) and the importance of non-health factors in DI take-up decisions.

This volume lays the groundwork for the next stage of the project, which will address the following question: "given health status, to what extent are the differences in labor force participation across countries determined by the provisions of DI programs?" The current volume has explored many issues that are inputs into this later study, such as the comparability of health outcome measures across countries and the extent to which DI provisions are prompted by a country's health status or employment circumstances.

Funding for this project was provided by grants from the National Institute on Aging (grant numbers P01-AG005842 and P30-AG012810) to the NBER.


National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us