Adjusting Social Security for Increasing Life Expectancy: Effects on Progressivity
Publication Date: August 2010
Publisher(s): Center for Retirement Research at Boston College
Keywords: Social Security
Coverage: United States
Achieving long-run Social Security solvency requires addressing rising life expectancy. Increasing the Full Retirement Age (FRA), while holding the Early Entitlement Age (EEA) fixed, could be effective but eventually will result in replacement rates that are viewed by many as too low. A possible policy to prop up replacement rates is to raise the EEA, which has been age 62 for more than 40 years. However, an increase in the EEA introduces unfairness because the variation in life expectancy across socioeconomic groups is positively correlated with lifetime income. Using data from the Health and Retirement Study to investigate how earnings relate to mortality risk and health limitations, this project explores the possibility of constructing a flexible FRA that could preserve or even enhance the progressivity of Social Security benefits. If life expectancy were correlated with lifetime income, Social Security policy could use the AIME (Average Indexed Monthly Earnings) to target policies that are more equitable for people with both lower lifetime income and lower life expectancy. Unfortunately, we find that while life expectancy is strongly correlated with AIME for men, it is only weakly correlated for women, and when pooling the genders the correlation disappears. We then investigate whether targeting could be done by the max AIME, which is the AIME for single persons and the maximum of the husband‘s or wife‘s AIME for married couples. We find that the max AIME, which is a household measure of lifetime income, could be used for constructing a flexible FRA because it is negatively correlated with mortality
risk and also negatively correlated with other measures of economic vulnerability or inability to work at older ages. With a flexible FRA, individuals in households with a
low max AIME would have a lower FRA than other individuals.