Risk perceptions and lifestyle choices Empirical and theoretical findings for smoking and obesity

Risk is everywhere, but because individual risk perceptions are inherently difficult to measure researchers often ignore them and resort to deterministic models. There is emerging evidence however that risk perception plays a fundamental role in influencing (perhaps subconsciously) lifestyle choices. In this dissertation I show that financial risk influences both smoking and obesity, two lifestyle choices that have previously been treated almost exclusively as deterministic phenomena. In my first paper I analyze the relationship between perceptions of economic insecurity and smoking behavior. An empirical model, designed to test theory motivated by findings in economics, psychology, and neuroscience, examines the effects of economic insecurity, defined as the probability of catastrophic income loss, on smoking. It is proposed that smoking is used as a form of "self-medication" in times of insecurity, in effect, decreasing an individual's "perceived" risk of income loss. Proxies for various measures of economic insecurity are derived from longitudinal data on income and employment history. Findings suggest that economic insecurity causes individuals to be more likely to smoke. Perceptions of economic insecurity and weight are also positively correlated. In my second paper I develop a two-period, two-state model, in which I endogenize preferences for body fat and observe optimal fattening under economic uncertainty. As motivated by behavioral ecology, weight gain is seen as a type of precautionary savings, in which humans gain weight when they perceive food insecurity. Comparative static results suggests that fattening is a form of self-insurance--individual's gain weight with increases in perceived economic uncertainty and lose weight with increases in perceived economic security. My final paper is a theoretical and empirical investigation of the effect of household composition (i.e., number of workers and non-workers in the home) on weight. It is hypothesized that household composition affects weight through three mechanisms already established in the economics literature as determinants of obesity: risk, income, and time costs. Evidence in this paper suggests that increasing the number of workers and non-workers in the home decreases weight through the proposed mechanisms of reduced risk and the decreased time costs of eating healthy