When making intertemporal decisions like saving for retirement individuals need to form expectations over various uncertain future events, such as the chances of high medical expenditures or chances of survival into old age. Research on individuals' expectations has made considerable progress on how to elicit individuals' expectations about the future, leading to various subjective probabilities measures being added to a number of general-purpose surveys. While there has been considerable research into the predictive power of these variables of subsequent outcomes and into measurement properties of subjective probabilities, very little is known about how individuals form and update the expectations they report. In this project we will use up to 60 waves of unique high-frequency longitudinal data on individuals' expectations that we have been collecting in the RAND American Life Panel (ALP). In the context of the ALP Financial Crisis Surveys we have queried respondents monthly about their expectations of losing their job, future stock market returns and gasoline prices, and we have asked them every quarter about their expectations about house prices, retirement, survival and out-of-pocket medical expenses. The overall objective of this proposed research is to establish the empirical foundations of how individuals form and update their expectations, recognizing that the expectation formation process is likely to vary by domain of outcome and by individuals. We will start out by studying the empirical properties of the high-frequency expectations data in cross-section and panel and what they imply for the basic features of individuals' expectation formation process in each domain. We willuse the insights to formulate models of expectation updating. We will estimate and test Bayesian updating models, as well as alternative models depending on the findings. Having a large number of observations for each individual we will estimate the expectation formation models at the individual-level and relate the resulting parameters to individual characteristics. We will further investigate the predictive power of individuals'subjective probabilities for economic decisions in panel data.
Studies have shown a strong link between socioeconomic status and health: How individuals make inter- temporal: decisions relating to their health and economic status under uncertainty will depend on their expectations about the future. Thus expectations will indirectly influence the evolution of health and economic positions. In this project we study individuals' expectation formation in several domains and how these influence economic decision making.
Showing the most recent 10 out of 309 publications
