| This dissertation consists of three essays in behavioral economics. The first essay draws out the consequences of selective attention in a learning model. I show that selective attention may lead people to persistently fail to recognize important empirical regularities and make biased forecasts. In addition, I identify factors that make such errors more likely or persistent. The model sheds light on a set of documented biases in inference and is applied to help understand the formation and stability of systematically incorrect stereotypes and beliefs.;The second essay, written jointly with Sendhil Mullainathan and Andrei Shleifer, presents a model of uninformative persuasion in which individuals "think coarsely": they group situations into categories and apply the same model of inference to all situations within a category. We apply the model to study uninformative advertising and product branding, as well as to understand some otherwise anomalous evidence on mutual fund advertising.;The final essay, written jointly with Sendhil Mullainathan, builds a general framework for analyzing health insurance when people may make mistakes in deciding whether to utilize medical services. Our analysis provides qualitative and quantitative guidance for optimal insurance policy. We produce a formula for the optimal co-pay and show how it can be empirically implemented by measuring demand elasticities and demand responses to "nudges" --- small factors that affect the choices of behavioral but not standard agents. |
