Cognitive and emotional factors have played a larger role in economists' understanding of the world in the last decades. While earlier work has focused on experimental and theoretical results, a larger number of recent contributions have tested ideas from the field of Psychology using econometric methods for causal identification on field data. This line of research seeks to analyze market situations in which specific psychological factors can be identified to cause observed economic behavior. My dissertation, at the intersection of Behavioral and Applied Microeconomics, offers examples of behavior in which cognitive aspects are shown to play a central role and is unified across the three chapters by a common methodological approach. The first chapter, based on joint work with Kareem Haggag, reports evidence from tipping behavior of New York city taxicab customers. For credit card payments, the payment screen in the car displays suggested tip amounts. In particular, for one of the main companies, the suggested amounts are $2, $3, $4 for fares below $15, and 20, 25 or 30 percent above $15. Using this variation, the chapter shows that suggestions play an important role in tipping behavior of customers: comparing rides below and above $15 using regression discontinuity methods, it is possible to show a large local causal effect of the suggestions on average tips. Moreover, a backlash effect is observed, as more customers decide not to tip on a credit card at all. These findings contribute to our understanding of default effects beyond the area of tipping, for instance in savings. An even broader lesson is that these findings isolate a case in which cognitive and emotional responses are likely to mediate the relationship between preferences and choice. The second chapter, based on a joint work with Kareem Haggag, presents field evidence on cheating behavior. During the two years 2008-2010, several taxi drivers cheated customers by charging a higher fare amount that is allowed only for rides outside the city even for rides in the city. The choice of whether to cheat a customer on a individual ride is shown to be affected by loss aversion. The estimates can be effectively reconciled by models of reference-dependent preferences that take drivers' expectation as reference points: drivers are more likely to cheat on those rides within a shift in which they are below expectations. The results highlight the role played by a classic decision-making bias in shaping unethical behavior in a market. These findings suggests that cognitive and emotional aspects of the valuation of benefits are relevant to our economic understanding of ethical problems. The third chapter presents regression-discontinuity evidence on an investment-incentive program. The methodology, which compares firms who received the award with those that marginally lost it, allows for a cleaner identification of the effect of the policy. In this last essay, the conceptual tools from Applied Microeconomics used in the first chapter are put to work in the context of firms' behavior. The tool allows one to show in a straightforward manner the main outcomes of the policy.

Before making a choice, we often have the opportunity to learn more about the options that are available. For example, we can check the characteristics of a product before buying it, or read different newspapers before a political election. Understanding what shapes the demand for information, and its role in the decision process, is important to study economic choices. This dissertation contains three essays in behavioral and information economics that utilize experimental data and modeling to analyze how people choose and use information to make decisions. The first chapter, "Coarse and Precise Information in Food Labeling," uses experimental data to determine whether precise food labels can be more effective and informative than coarse ones. In a preregistered online study conducted on a representative US sample, I manipulate front-of-package labels about foods' calorie content. I find that coarse-categorical labels generate a larger reduction in calories per serving compared to detailed-numerical labels despite providing less information. Choices violate the predictions of Bayesian decision theory, suggesting that consumers are less responsive to detailed information. Results also show that participants prefer coarse labels, suggesting a general preference for simple, easy-to-interpret information. The second chapter, "The Status Quo and Belief Polarization of Inattentive Agents," studies how differences across agents can drive information acquisition and generate polarization. In a rational inattention model, optimal information acquisition and subsequent belief formation depend crucially on the agent-specific status quo valuation. Beliefs can systematically update away from the realized truth and even agents with the same initial beliefs might become polarized. A laboratory experiment confirms the model's predictions about the information acquisition and its effect on beliefs. Differently from the model's predictions, participants display preferences for simple messages that can provide certainty. The third chapter, "Dynamic Information Choice with Biased Information Sources," uses experimental data to study how people decide what kind of information to acquire when they have multiple opportunities to learn. Standard theory predicts that decision makers should collect the stream of information that leads to the maximization of the expected reward from the final choice. An online experiment on sequential information acquisition shows that people systematically deviate from the predictions of the standard normative model. Participants display a certainty-seeking information acquisition behavior and under-respond to the new evidence collected, reviewing rarely their own information acquisition strategy.