Estimating Life Cycle Effects of Subjective Survival Probabilites in the Health and Retirement Study.

Abstract

This paper attempts to confirm the life-cycle relationship that lower subjective survival probabilities should lead to less positively sloped consumption trajectories. I use the results of six waves of subjective survival probability questions in the HRS to construct an index of survival belief that exploits the panel nature of the data by summarizing all of a respondent’s answers to such questions. In conjunction with constructed consumption values from the financial section of the HRS, I test the life-cycle relationship using OLS and Least-Absolute Deviation regression. I find weak evidence that the life-cycle effect of subjective survival probability is significant in a high-cognitive-ability sub-sample of the HRS. Measurement error in the constructed consumption data is problematic.