Retirement, Death, and the Distribution of Retirement Leisure

Abstract

Using longitudinal data from the Health and Retirement Study and anordered SUR Tobit model, we jointly analyze retirement age and death age to see whoconsumes retirement leisure. We find that retirement age and death age are indeedcorrelated. Hispanic and non-white individuals, and those who suffer from poorer selfassessed health, heart or mobility problems all successfully compensate for differences inmortality risk: they have a statistically significantly different retirement or death age, butstatistically indistinguishable total retirement leisure. Attempts to fix Social Security orpension funding by increasing pension eligibility ages would disproportionately affectsome of these groups (C34, J11, J26).