Essay on household financial and career decisions

This dissertation advances the understanding of household economic behavior. The first essay addresses the investment decisions of households, in particular the often sizable allocation to individual stocks. This essay estimates a structural model of the beliefs required to rationalize this direct stock ownership. In the model, households believe they can learn about individual stock returns through costly research. This essay shows that the model produces choices consistent with many of the observed features of household portfolios. Further, the model identifies both the distribution of research costs and the distribution of household beliefs about the predictability of individual stock returns. Estimation results indicate that most households have only modestly optimistic beliefs about individual stock return predictability, although a minority must expect research to yield exceptionally high returns.;The second essay explores the role of idiosyncratic taste and skill in career choice. We use expected utility to cardinalize a logit model of career choice in a setting where we observe the income risk of chosen careers and the risk-aversion of the people who choose them. The key parameter of interest - the importance of idiosyncratic taste and skill in career choice - is identified from the shift in the distribution of income risk with risk aversion. We estimate the model using proxies for income risk and risk preference, both from the PSID. We separate idiosyncratic career taste from skill using the pay gap between high- and low-income risk people with high and low risk-aversion.;The third essay is an empirical investigation of the presumed effort shirking induced by long-term guaranteed labor contracts. Previous studies have explored effort shirking among professional athletes. In cases where a negative correlation between player performance and the time until contract expiration is found, the correlation is viewed as evidence of effort shirking. An alternative explanation is that contract lengths are chosen endogenously. Using panel data from the NBA, this paper tests for the presence of effort shirking while explicitly controlling for endogenously chosen contract lengths. Estimation results indicate that no effort shirking is present in the data once endogenous contract lengths are appropriately accounted for.