Individuals have different beliefs about how long they will live, how healthy they will be in old age, when they will retire, and other aspects of the future. The underlying goal of this project is to explore how these subjective beliefs influence saving decisions. The project uses two new panel data sets, the Health and Retirement Survey (HRS) and the Survey of Asset and Health Dynamics Among the Oldest Old (AHEAD), each of which measures subjective probabilities about mortality and other future events. Because there is virtually no past research using subjective probability measures to explain saving decisions, the project includes a number of exploratory components about the subjective data. One component is to evaluate whether people's subjective expectations are realistic. For example, as future waves of the surveys become available, the project will determine whether differences in subjective expectations across individuals are realized. Specifically, how accurately do people assess differences in their mortality and other future events? Another component of the project investigates the determinants of the subjective probability distributions, and estimates specific models of the formation and evolution of subjective beliefs. The project considers several aspects of how subjective expectations influence saving decisions. First, the subjective probability distributions will be used to estimate models of saving behavior by the elderly. In particular, subjective information about mortality risk and the risk of medical expenditures will be used to estimate the role of expectations in influencing saving decisions; the importance of risk aversion; and the extent to which saving results from differences in expectations, as compared with differences in tastes. Second, the project will assess whether saving decisions are """"""""rational,"""""""" given particular expectations about the future. Third, the project will identify groups with economic hardship, and assess the extent to which their economic hardship results from poor planning, unrealistic beliefs, or bad luck. And fourth, the estimates of saving behavior will be used to forecast saving and wealth of the elderly.
