Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Three Essays on Reducing Outcome Bias and Racial Disparities in Decision Making

Wang, Xiaohong (2023) Three Essays on Reducing Outcome Bias and Racial Disparities in Decision Making. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

Download (2MB) | Preview


In this dissertation, I investigate the presence of outcome bias and racial disparities in various societal institutions such as the labor market, the criminal justice system, and electoral politics. The primary focus of this study is to understand the underlying mechanisms that cause these biases and to propose solutions to mitigate their impact on decision-making.
The first paper is a co-authored paper with Dr. Robizon Khubulashvili and Dr. Sera Linardi. We ask if visually observing effort can reduce outcome bias. Outcome bias is pervasive and persistent across different environments. In our noisy gift-exchange game, where agents can perform a real effort task to improve principals’ lottery win probability, we replicate outcome bias in effort rewarding when the effort is only numerically observable. To investigate the role of principals’ beliefs about effort cost, we employed a visual treatment in which principals watch a 30-second video of the agents performing the task. We show that visually observing agents’ work corrects asymmetry in rewarding effort. The post-experiment survey suggests that the mechanism through which visually observing effort reduces the outcome bias in reciprocating effort is informing principals about the cost of effort.
In the second paper, I study the effect of Black Lives Matter Protests on racial disparities in nonfatal police-civilian interactions. Protests against police brutality and systemic racism have been prevalent in the United States, and most recently hastened by the killing of George Floyd. This paper evaluates how George Floyd protests affect racial disparities in nonfatal police-civilian interactions using police practice data across 17 cities in 12 states and a combination method of regression discontinuity (RD) and difference-in-differences (DiD). The results show that the protests have not impacted the proportion of African Americans in stops, but have reduced the proportion of African Americans in arrests from 30% to 26%. When dividing all interactions into cases in daylight and in darkness, the decreased effect of the protests only holds during the daytime, instead of nighttime when public supervision is absent. It suggests that Black Lives Matter protests did affect nonfatal police-civilian interactions when it comes to race. However, the day-night differences imply that the decrease in police interactions with African Americans may not be due to the change in police attitudes/beliefs. It is possible that it is a temporary change yielding to strong public attention at that time.
My third paper explores the racial differences in politicians’ persistence in elections. Empirical data from California city council elections and a close election regression discontinuity design (CERDD) suggest that losing an election causes 70% attrition in rerunning for office. After a loss, however, nonwhite candidates are 59% more likely to run for office again compared to white candidates. The possibility of winning the subsequent election remains the same for different racial groups conditional on rerunning. As such, the persistence of losing nonwhite candidates contributes to closing the racial representation gap.


Social Networking:
Share |


Item Type: University of Pittsburgh ETD
Status: Unpublished
CreatorsEmailPitt UsernameORCID
Wang, Xiaohongxiw131@pitt.eduxiw1310000-0002-8993-1295
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairLinardi,
Committee CoChairJones,
Committee MemberCondra,
Committee MemberCastillo,
Date: 20 June 2023
Date Type: Publication
Defense Date: 19 April 2023
Approval Date: 20 June 2023
Submission Date: 14 May 2023
Access Restriction: No restriction; Release the ETD for access worldwide immediately.
Number of Pages: 70
Institution: University of Pittsburgh
Schools and Programs: Graduate School of Public and International Affairs > Public and International Affairs
Degree: PhD - Doctor of Philosophy
Thesis Type: Doctoral Dissertation
Refereed: Yes
Uncontrolled Keywords: Racial Disparities, Policing, BLM Protests, Minority Representation, Outcome Bias, Visibility
Date Deposited: 20 Jun 2023 16:12
Last Modified: 20 Jun 2023 16:12


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item