UC San Diego

In Chapter 1, I investigate a novel channel of polarization: divergent interpretations of information. I conduct an online experiment with Democrats and Republicans in the US to study beliefs about racial discrimination in the labor market, a topic on which Democrats and Republicans are polarized. I find that Democrats' beliefs about racial labor market discrimination are responsive to information on racial wage disparities, while Republicans' beliefs are not. As a result, wage gap information fails to reduce (and even increases) the partisan difference in discrimination beliefs. Moreover, even after both groups agree about the extent of racial hiring discrimination, participants change their opinions about whether it is a problem depending on their political affiliation, enabling disagreement in policy demand. Together, these findings highlight key challenges in using information to reduce polarization.

In Chapter 2, we leverage a randomized evaluation of an early childhood program to study the impact of early life investments on resilience to negative shocks. When the children in our study were 3-5 years old, they were randomized to a preschool program, a parenting program or to a control group. Ten years later, the children were exposed to school shut-downs during the Covid-19 pandemic. With nearly 900 observations, we show that the parenting program had a protective causal impact on the decrease in academic test scores during the year that schools were closed. While the control group saw a 0.31 SD decline in standardized test scores after Covid-19, the parenting group saw only a 0.12 SD decline. We provide a conceptual framework and evidence on potential mechanisms driving this effect.

In Chapter 3, we explore the robustness of rank independence of equalizing reductions with respect to experimental procedures. Bernheim and Sprenger (2020) devise and implement a novel test of rank-dependent probability weighting both in general and as formulated in cumulative prospect theory (CPT). They reject both hypotheses decisively. CPT cannot simultaneously account for the rank independence of "equalizing reductions" for three-outcome lotteries, which it construes as indicating linear probability weighting, and the relationship between equalizing reductions and probabilities, which it interprets as indicating highly nonlinear probability weighting.