UC Berkeley

Economic theory hinges on the fact that humans are rational. However, in the wild, research demonstrates human behavior often deviates from rationality. This deviation may result in suboptimal behavior. Researchers in behavioral economics and psychology have tried understand these irrational behaviors and clarified many of the ways humans are likely to be biased. Yet, we are still exploring ways to help people overcome their behavioral biases. This dissertation explores behavioral biases in three different contexts: technology, human cooperation, and banking.

This dissertation demonstrates a behavioral bias in A/B testing in technology and quantifies the amount to which this bias is a problem. Second, this dissertation proposes a light institutional intervention of giving more information to study the impacts on trust. Third, this dissertation explores the effects of offering a new financial product to overcome behavioral biases around opening bank accounts and savings. Overall, these papers demonstrate behavioral biases can lead to suboptimal outcomes such as making the wrong business decision or missing out on the benefits of cooperation, or failure to open a bank account and save. Luckily, there are some ways we can overcome biases (Chapter 3), but not all interventions work in the ways we would expect (Chapter 2).

The first chapter introduces the behavior of ''p-hacking", where decision makers stop experiments earlier or later than proper statistical validity requires, possibly because they are overly eager to obtain significant results. Such behavior may result in invalid test conclusions and financial losses. We investigate whether online A/B experimenters ``p-hack'' by stopping their experiment based on the p-value of the effect. Our data comes from a leading platform and contains 2,101 A/B tests that track the magnitude and significance level of the effect on every day of the experiment. We estimate the causal effect of reaching a particular p-value on stopping behavior by applying a regression discontinuity design to hazard modeling. Experimenters indeed p-hack, especially for positive lift values. Moreover, experimenters p-hack more if the lift is mildly positive rather than strongly positive. A latent class analysis shows that approximately 57% of experimenters p-hack at the 90% confidence threshold. A false discovery rate (FDR) analysis estimates that p-hacking increases false discoveries by 27.5%, while the overall rate of false-discoveries is 38%. This chapter is coauthored with Ron Berman, Leo Pekelis, and Christophe Van den Bulte.

In the second chapter, I introduce an information signal and role organization that may engender more trusting behavior. Trust is an essential ingredient for unlocking economic surplus. However, consider the prisoner's dilemma---all parties gain from cooperation, yet each party has an incentive to deviate. How can we organize society to unlock the possible gains from trust in such situations? We've all had experiences that indicate it is possible. Studies have shown prosocial individuals are more trustworthy. We can take advantage of this fact and suggest pairing prosocial individuals with less prosocial individuals who will trust them if their type is known. In this case, it takes information, timing, and only one pro-social individual to unlock the trust surplus. I find information actually decreases overall trust and does not impact . Consequently, too much information might negatively influence cooperation and trust by changing our biases.

In the final chapter coauthored with Paul Gertler, Sean Higgins, and Enrique Siera, I explore whether a financial incentive can nudge people into opening a bank account and saving. Despite the benefits of saving in formal financial institutions, take-up of no-fee formal savings accounts is low among the poor. Surprisingly, even after opening a savings account, use of the account is often low. In a large randomized experiment across 110 bank branches throughout Mexico, we provide a temporary incentive to both open and use a savings account: we offer prize-linked savings accounts with cash-prize lotteries, where lottery tickets are awarded as a function of savings balances. We find that 41% more accounts are opened in treatment branches than in control branches on average, and the number of accounts opened in treatment branches increases steadily over time while the lotteries were being offered. Although the incentive to save is temporary as lotteries are only offered for two months, the new accounts continue to be used over time. After five years, clients who opened accounts in response to the lottery continue saving and making transactions at the same rates as those who opened accounts in control branches during the same months.