Duke 2008 | ASHE | Equity and Efficiency in Health and Healthcare

The proportion of large employers offering retiree health insurance in the U.S. has declined by half in the past 20 years. This paper examines the potential implications of this change by estimating the effects of a retiree health insurance (RHI) offer on a comprehensive set of labor, health and health care use outcomes in the near-elderly population.

The analysis is based on ten years of longitudinal data from the Health and Retirement Survey (1992-2002). Concerns about endogeneity with respect to individual selection into jobs/employers that offer coverage and with respect to retirement are addressed with several methods. First, I restrict the analysis to individuals who have employer-sponsored health insurance coverage at baseline. Robustness checks include propensity score weighting to adjust for remaining baseline differences in observable characteristics. I also test for differential selection into retirement with respect to health and conduct placebo tests based on the not retired subsample. These checks generally support the baseline results.

I find that an RHI offer increases the probability of early retirement by 35 percent for both men and women. An RHI offer decreases the probability of being uninsured by 55 percent, while increasing the probability of employer-sponsored coverage and decreasing the probability of public coverage. There is suggestive evidence that individuals with an RHI offer are more likely to visit the doctor, use prescription drugs on a regular basis and have outpatient surgery. There is no evidence of differences in inpatient or preventive care utilization. While the results suggest that an RHI offer has little, if any, effect on health, there is strong evidence that RHI provides significant protection from high out-of-pocket medical costs. In the top 40 percent of the out-of-pocket spending distribution, those with an offer of retiree coverage spend 21 percent less on average.

Estimates of the value of retiree health insurance of $3,400 per year for men and $3,100 per year for women suggest that increasing opportunities for the near-elderly to purchase coverage at actuarially-fair prices through the individual market or public programs could significantly reduce the projected increase in uninsurance for this age group. For example, decreasing loading fees by half could bring average individual market premiums to the level of average willingness-to-pay for the near-elderly, which could have a significant impact on take-up rates of individual insurance. Policies that stabilize employer-based group coverage, address imperfections in the individual insurance market and use public health insurance programs to address the insurance needs of the near-elderly will be important as the impacts of the decline in RHI offer rates become manifest in the coming years.