This dissertation consists of two chapters on behavioral economics.

Chapter one analyzes mutual fund flows from 1999 to 2012 to provide evidence that financial advisers exploit their clients' extrapolative beliefs to sell funds that pay high commissions. Utilizing comparisons between twin retail mutual fund share classes that differ only in commissions, I find a strong positive correlation between commission levels and return chasing. Fund flows from high commission shares show increased purchases of past winners, but no increased redemptions of past losers. This pattern is consistent with the incentives from 12b-1 commissions, which pay advisers to keep clients invested in high commission funds, but is hard to reconcile with alternative explanations such as advisers reducing search costs for quality funds, or return chasers self-selecting into high commission advisers. My findings support the SEC's recent efforts at restructuring 12b-1 commissions, as well as the White House's recent push to require fiduciary duty from all retirement advisers.

Chapter two attempts to isolate the effects of violent video game sales on crime using a novel data set of state-level game sales from 2006 to 2010. The panel database, constructed using NPD group publications and state-level sales weights computed from Google Trends search data, overcomes the time-series data restrictions faced by previous works. However, the state-level variation in this panel data was not enough to disentangle the impact of game sales from time trends because video game sales are highly correlated across states. I document the methodology used to reconstruct the panel data using Google Trends, as this could prove useful in other situations where panel data is difficult or expensive to obtain.

This dissertation uses tools from behavioral and labor economics to study longstanding questions about people and policies.

The first chapter, co-authored with Evan K. Rose, provides a comprehensive look at what happens to criminal offending around childbirth, marriage, and divorce for women and men. Our event study analysis suggests that pregnancy is a strong inducement for fathers and especially mothers to reduce criminal behavior. For mothers, criminal offending drops precipitously in the first few months of pregnancy, stabilizing at half of pre-pregnancy levels three years after the birth. Men show a 25 percent decline beginning at the onset pregnancy; however, domestic violence arrests spike for fathers immediately after the birth. A design using stillbirths as counterfactuals suggests a causal role for children. In contrast, marriage marks the completion of a 50 percent decline in offending for both men and women. Finally, people headed for divorce show relative increases in crime following childbirth and marriage. The patterns in drug offenses for new mothers are consistent with a Beckerian model of habit formation.

In the second chapter, I use a new dataset on unemployment insurance recipients to study the effects of benefits on search behavior. In a regression kink design, I find that monetary weekly benefits have a positive effect on unemployment duration but no impact on search behaviors. As evidence against misreporting, I show that reservation wages predict reemployment wages and decrease with unemployment duration and the local unemployment rate. These results suggest that explicitly measured search behaviors may not explain the duration response to benefits.

This dissertation studies modeling and identifying cognitive mistakes in decision making in the context of education system in China. Chapter 1 studies the cognitive distortion in complex school choice problems in Chinese centralized admission system. Chapter 2 studies the impact of education on students’ ability of decision making. Chapter 3 studies the psychological impact of religious obligation on education attainment.

In Chapter 1 (coauthored with Shaoda Wang and Xiaoyang Ye), we empirically study an admission system that employs a constrained Deferred Acceptance Algorithm to understand how students construct their lists. Students appear overly cautious with their top choices and most of them do not always put safer choices at a lower-ranked spot on the list. We propose that the Model of Directed Cognition could explain such choices. Applicants using the model myopically focus on the spot they are contemplating and neglect its impact on the rest of the list. To differentiate from alternative hypotheses, we deploy an in-field experiment that pinpoints a core prediction of our model concerning framing effects and find clear evidence of it. Structural estimation suggests that 45%∼55% of the sample are better described by our model and that this boundedly rational decision rule explains 83% of outcome inequality across socioeconomic groups.

Chapter 2 (coauthored with Binkai Chen and Wei Lin) intends to investigate the causal impact of collegiate economics courses on individual learning and decision-making under a development context. By exploiting a Chinese college-admission system that quasi-randomly assigns students to economics/business majors given students’ preferences and the College Entrance Exam’s cutoff scores for economics/business majors, we are able to isolate the treatment effects of an economics education on students’ responses to a decision-making survey. Specifically, we compare the survey responses of students who narrowly meet the cutoffs for the economics/business majors to those who do not and find that students educated in economics/business courses are more likely to be risk neutral and less prone to common biases in probabilistic beliefs. While students in economics/business majors do not show significant changes in social preferences, they appear more inclined to believe that others behave selfishly.

Chapter 3 (coauthored with Shaoda Wang and Xiaoyang Ye). We reports a field experiment that tests the effect of motivated cognition on information acquisition. When the high-stakes College Entrance Exam is held in the month of Ramadan, Chinese Muslim students not only underestimate the cost of fasting when uninformed, but further, misread clear empirical evidence of the cost, which we obtain by analyzing administrative data on past students’ exam performance. Inspired by the theory of motivated cognition, we tackle this learning failure by randomly offering a subset of the students reading materials in which well-respected Muslim clerics explain that it is permissible to postpone the fast until after the exam. Students who receive the material are substantially less likely to misread our empirical analysis and more willing to postpone the fast.

The findings in this dissertation can deepen our understanding of the impact of psychological factor and cognitive limitation on decision making, particularly in the context of education.

This dissertation establishes that loss aversion fundamentally influences the tax avoidance behavior of property taxpayers. Upon receiving notice of a new assessed value, homeowners have the option to appeal, which, if successful, could lower their tax base. I conjecture that a lagged, salient value---a property's assessed value in the previous year---serves as a natural and prominent reference point to property owners. Guided by a reference-dependent model of assessment protests, I demonstrate various predictions using a sample of 8.2 million administrative property assessment records associated with 1.6 million appeals. Foremost, loss aversion introduces an extensive margin effect around the reference point that induces property owners to disproportionately appeal assessments that have increased relative to the prior tax year. In aggregate, this leads to a sharp kink in the probability of protesting as a function of percent change in assessed value exactly at zero percent change. Additionally, homeowners not only achieve but also seek out value adjustments that result in a final assessed value precisely at the property's previous assessed value. Evidence is strongest for owner-protesters, for whom the reference point is presumably most relevant. Finally, I employ a simple counterfactual estimation strategy. It suggests that loss aversion has a sizable impact on annual household property taxes, most notably for properties constituting the top quartile of value, while also highlighting the importance of one's position in the loss domain in understanding average effect sizes.

This dissertation consists of three chapters which all study parental decision-making for their children in the domain of competition. Willingness to compete is an important determinant of education and labor market outcomes. A growing literature has documented a robust gender differences in competitiveness which may explain the observed differences between men and women in these domains. Parents play an important role in shaping children's preferences and long-term outcomes. The key motivation for the dissertation is to understand the influence of parents on gender differences in competitiveness and later life outcomes.

In the first chapter, I study how parents make competitiveness choices for their adolescent children in Norway. In an experiment with 1480 parents and children, parents choose if their child will do a task for a competitive or non-competitive pay scheme. The paper establishes a number of novel facts on parents' choices for children. First, parents choose more competition for boys than for girls. The gender gap in parents' choices is smaller than that in children's own choices. Second, two main mechanisms explain the gender gap in parents' choices: their beliefs about children's preferences and paternalistic behavior. Third, parents' choices are more responsive to the ability of boys than girls, which results in many high-ability girls not entering into competition. Fourth, parent gender matters: fathers are more likely than mothers to enter their child into competition. Finally, children are unaware of the gender difference in parents' choices and believe that parents will make the same choices for boys and girls.

The second chapter uses data from the same experiment as in chapter one, to study the transmission of competitiveness preferences within social networks. I document a positive correlation between preferences both within family and within peer groups. I study parents' beliefs about the correlation of preferences, and find that parents overstate how close their own preferences are to their children's preferences.

In the third chapter, which is joint work with Edward Miguel (UC Berkeley), we study how parents make competitiveness choices for their young children in Kenya. We present preliminary data from two lab-in-the-field experiments, and highlight three findings. First, in the sample of parents, men are more likely than women to compete. Second, in the sample of children, there is no evidence that boys are more competitive than girls. In fact, girls are more likely to compete than boys, but this difference is not significant. Third, when parents are asked to choose if their child should compete or not, there is a significant gender difference in choices; parents choose more competition for boys than for girls. We study mechanisms for parents' competition choices. Importantly, the difference in choices for boys and girls, is not explained by parents' beliefs about their children's preferences.

This dissertation is comprised of three essays in empirical economics. These essays are united by three clear intellectual and methodological themes. First, each essay attempts to bring theories and insights from psychology to bear on open questions in economics, with a focus on topics of importance to developing countries. Second, they use experiments - both randomized field experiments and natural experiments - to test economic theory. Finally, each paper attempts to measure economically important but difficult-to-observe behaviors and preferences - self-control problems in Chapter 1, social norms in Chapter 2, and discrimination and fairness preferences in Chapter 3.

In chapter one, coauthors Liang Bai, Edward Miguel, Ben Handel and I construct a simple model of preventive health behavior under present biased time preferences, and show how beliefs about future time preferences (sophistication, partial naivete, and perfect naivete) affect how agents are predicted to use, under-use or misuse different types of commitment contracts. We propose a type of commitment contract that has the potential to benefit not just sophisticated present biased agents, but also naifs. We conduct a field experiment focused on increasing the share of patients who actively manage their hypertension by visiting a doctor periodically. The experiment is closely tied to the theory, allowing us to estimate the key parameters of the model.

In chapter two, coauthors Stefano DellaVigna, John List, Ulrike Malmendier and I ask the question: Why do people vote? We argue that social image plays a significant role in explaining turnout. People vote because others will ask. The expectation of being asked motivates turnout if individuals derive pride from telling others that they voted, or feel shame from admitting that they did not vote, provided that lying is costly. We design a field experiment to estimate the effect of social image concerns on voting. In a door-to-door survey about election turnout, we experimentally vary (i) the informational content and use of a flyer pre-announcing the survey, (ii) the duration and payment for the survey, and (iii) the incentives to lie about past voting. Our estimates suggest significant social image concerns. For a plausible range of lying costs, we estimate the monetary value of voting because others will ask to be in the range of $5-$15 for the 2010 Congressional election. In a complementary get-out-the-vote experiment, we inform potential voters before the election that we will ask them later whether they voted. We find suggestive evidence that the treatment increases turnout.

In chapter three, I exploit a natural experiment in India to identify how mixing rich and poor students in schools affects social preferences and behaviors. A policy change in 2007 forced many private schools in Delhi to meet a quota of poor children in admissions. This led to a sharp increase in the presence of poor children in new cohorts in those schools, but not in older cohorts or in other schools. Exploiting this variation, I study impacts on three classes of outcomes: (i) prosocial behavior, (ii) discrimination, and (iii) academic outcomes. First, I find that having poor classmates makes wealthy students more prosocial. In particular, they become more likely to volunteer for a charity at school. Second, having poor classmates makes wealthy students discriminate less against poor children, measured by their teammate choice in an sports contest. Third, I find marginally significant negative effects on test scores in English, but no effect on Hindi or Math. Overall, I conclude that mixing in schools had substantial positive effects on the social behaviors of wealthy students, at the cost of negative but arguably modest impacts on academic achievement. To shed light on mechanisms, I exploit administrative records on the idiosyncratic assignment of students to study groups and find that the effects on social behaviors are largely driven by personal interactions between wealthy and poor students, rather than by changes in teacher behavior or curriculum.

This dissertation studies the determinants and consequences of standard, and in particular, non-standard economic preferences. Preferences are the primitives of choice theories in economics. An active body of research investigates the heterogeneity and determinants of economic preferences and their consequences for decision-making. In three chapters, this dissertation studies different aspects of this topic both theoretically and empirically. Chapter 1 Preferences over Relative Income within the Household investigates the existence of non-standard, immaterial preferences over relative income within the household. In intuitive terms, preferences over relative income capture the notion that individual partners have a preference over their share of total household income independently of the level of total household income. For instance, such preferences include inequality aversion or a preference for being the primary earner. The chapter documents the existence of preferences over relative income, quantifies their extent and documents heterogeneity by gender. The chapter also investigates the consequences of these preferences for household decision making and socio-economic gender inequality. Chapter 2 Relationship of Gender Differences in Preferences to Economic Development and Gender Equality empirically investigates cross-cultural variation of gender differences in several standard and non-standard economic preferences. The chapter investigates the role of economic development and gender equality in shaping gender differences in preferences across countries. Chapter 3 Longevity and Patience investigates the role of longevity in driving patience across countries and individuals. The chapter documents empirical evidence for the long-standing hypothesis that higher longevity fosters patience. In addition, the chapter explores theoretically the consequences of a positive link between longevity and patience for the emergence of development traps.

The first chapter Preferences over Relative Income within the Household tests for the existence and quantitative extent of non-standard, immaterial preferences over relative income within the household and analyzes their consequences for the matching of couples and family outcomes. This chapter presents a theoretical model and empirical evidence from administrative German tax data and furthermore documents results from an online experiment. The model and evidence from tax data have been developed in joint work with Nikolaus Hildebrand. First, to guide the empirical analysis, the chapter outlines a marriage market matching model in which prospective partners can hold non-standard preferences over relative income, such as inequality aversion or a preference or aversion for being the primary earner. The model yields testable predictions for the presence and structural form of partners' preferences over relative income either as a kink or notch at the point of equality in the relative income distribution of households, i.e. the distribution of the female share in total household income. The predictions are tested for the case of Germany using administrative tax data. The data indicates a kink at the 50%-threshold in the relative income distribution which indicates the presence of kinked preferences over relative income. To disentangle the preferences of women and men, a survey experiment is conducted and analyzed. The results indicate that women exhibit inequality aversion while men show a preference for being the primary earner. Finally, the chapter explores the consequences of non-standard preferences over relative income on further family outcomes such as divorce and household public good provision.

The second chapter Relationship of Gender Differences in Preferences to Economic Development and Gender Equality investigates the role of economic development and gender equality in driving gender differences in standard and non-standard economic preferences. The chapter includes material from a study joint with Armin Falk and published in Science. The chapter first lays out two contrasting hypotheses that make opposing predictions about the relationship of gender differences in preferences to economic development and gender equality. According to the Social Role Hypothesis, higher economic development and gender equality promote a dissolution of traditional gender roles, leading to a narrowing of gender differences in preferences. In contrast, according to the Resource Hypothesis, higher economic development and gender equality yields greater availability of material and social resources to both women and men and facilitates the independent development and expression of gender-specific preferences, ultimately leading to an expansion of gender differences. These hypotheses are tested using 76 representative country samples of validated measures of six preferences, including willingness to take risks, patience, positive reciprocity, negative reciprocity, altruism, and trust. The analysis shows that gender differences in preferences are positively associated with both economic development and gender equality, confirming the predictions of the Resource Hypothesis. The findings point toward the critical role of availability of and equal access to material and social resources for both women and men in facilitating the independent formation and expression of gender-specific preferences across countries.

The third chapter Longevity and Patience provides an empirical test of the long-standing and influential hypothesis stating that greater longevity fosters higher patience.The chapter includes material that has been developed in joint work with Armin Falk and Uwe Sunde. The chapter tests this hypothesis using data on an experimentally validated survey measure of patience in 76 representative country samples in conjunction with life table measures of longevity that vary across country-age-gender cells. The empirical framework analyzes the longevity-patience link in an econometric design that is akin to a difference-in-difference specification isolating the effect of longevity from age-, country-, and gender-specific confounders. The empirical results document that higher longevity is associated with higher patience: a ten-year increase in life expectancy is associated with a 5-percentage point increase in the discount factor. This relationship remains robust for different proxies of life expectancy, for various sub-samples, when applying instrumental variable estimations, and when conditioning on lifetime experiences related to economic development, institutional quality, or violence. Finally, the chapter analyzes the longevity-patience link in a theoretical framework that demonstrates the consequences for the emergence of development traps due to a vicious cycle of high mortality, low patience, and low human capital investments.

How do investors respond to predictable shifts in profitability? We consider how demographic shifts affect profits and returns across industries. Cohort size fluctuations produce forecastable demand changes for age-sensitive sectors, such as toys, bicycles, beer, life insurance, and nursing homes. These demand changes are predictable once a specific cohort is born. We use lagged consumption and demographic data to forecast future consumption demand growth induced by changes in age structure. We find that demand forecasts predict profitability by industry. Moreover, forecast demand changes five to ten years in the future predict annual industry stock returns. One additional percentage point of annualized demand growth due to demographics predicts a 5 to 10 percentage point increase in annual abnormal industry stock returns. However, forecasted demand changes over shorter horizons do not predict stock returns. A trading strategy exploiting demographic information earns an annualized risk-adjusted return of approximately 6 percent. We present a model of inattention to information about the distant future that is consistent with the findings. We also discuss alternative explanations, including omitted risk-based factors.

This dissertation incorporates two major themes from behavioral economics and political economy. The first theme focuses on how social forces—such as culture, norms, and identity—shape economically meaningful behavior. This work builds on key theories from psychology and economics that embody intuitive ideas: we compare what we have relative to our peers or expectations, and are constrained by social norms yet impose them on others as well. The first two chapters explore these ideas in historically or politically consequential contexts. They evaluate the historic role of commercial television in setting consumer expectations and study the motives behind pressuring others to follow norms on political civic duty.

The second theme in this dissertation moves onto governments as decision-makers and studies the determinants of policy adoption across the United States. The first chapter within this theme describes the general dynamic patterns of policy diffusion for a broad set of U.S. state-level policies. The second then focuses on municipal governments in their adoption of non-partisan, low-cost “nudge” interventions.

This dissertation applies behavioral economics concepts to understand key issues in firms’ and individuals’ decision making. In particular, this dissertation investigates the presence of behavioral biases on firms’ managerial decisions and on individuals’ annuitization decisions through the use of field and incentivized experiments, respectively.

In the first chapter, we run a field experiment with micro-entrepreneurs in Brazil (N = 742) to shed light on the constraints that lead to the under-adoption of improved business practices. We randomly offer entrepreneurs micro-incentives, which include reminders, deadlines and small monetary payments, to implement record keeping or marketing for three consecutive months, following a business training program. Our intervention is designed to have a significant impact on firms’ decisions only in the presence of behavioral biases. Compared to traditional business training, micro-incentives significantly increase adoption of marketing (13.2 p.p.) and record keeping (19.2 p.p.), with positive effects on firm survivaland investment over four months. Additional survey evidence is consistent with biases, such as inattention, time inconsistency and information avoidance, inhibiting the adoption of improved practices. Taken together, our results show that behavioral biases have a significant impact on firms’ managerial decisions.

In the second chapter, we study psychological biases in take-up of annuities, using an incentivized experiment with a probability-based sample (N = 3,038). Choosing an annuity was payoff-maximizing in the experiment at all prices, but take-up was incomplete and price elastic. Reformulating decisions as insurance against a “bad” outcome rather than insuranceagainst “longevity risk” did not increase take-up. Instead, we find substantial failures of contingent reasoning: participants underappreciated how annuitization mitigated the need for less-efficient means of saving for retirement. Increasing the salience of the interaction with savings decisions, or eliminating the need to think through this interaction altogether, substantially increased annuity take-up.