Best Dissertation Defended in 2019


Winner: Kyle Peyton, Yale University

Title of the winning dissertation: “Experiments on Legitimacy and Intergroup Relations: Policing, Trust, and Prejudice in the United States” 

Committee Members: Kristin Michelitch (Chair), Vanderbilt University; Alexander Coppock, Yale University; Saad Gulzar, Stanford University.

Kyle Peyton’s dissertation is an experimental tour-de-force that demonstrates mastery of three common modes of experimentation. Particularly strong across all three papers is rigorous analysis, which employs a plethora of careful robustness checks. Paper 1 is a field experiment that finds support for contact theory in the context of police-citizen interrelations. Specifically,  positive door-to-door conversations with police increase citizen perceptions of police legitimacy and performance as well subjects’ stated willingness to cooperate with police. This experiment not only is one of the few that adequately test Allport’s contact theory by carefully controlling police officers’ interactions with citizens, but it is also very policy relevant, since many police departments have touted some form of community policing to improve policy-community relations. This paper, coauthored with Michael Sierra-Arevalo and David Rand, has been published in the Proceedings of the National Academy of Sciences. 

The second paper uses five survey experiments to reassess the validity of a causal claim lifted from prominent observational research. In American politics, one empirical regularity (e.g., using ANES data) is that trust in government is positively correlated with support for redistribution, even when controlling for many other factors.  Peyton devises a survey experimental manipulation that manipulates trust using exemplar OpEd news articles about the presence or absence of legislator corruption and crimes, but the manipulation does not affect support for welfare programs, as previous authors have suggested. This paper is noteworthy for being both a well-powered null finding and a solo-authored (forthcoming) publication in the American Political Science Review.

The third paper uses a series of laboratory experiments to obtain a behavioral measure of racial discrimination. Peyton uses this behavioral measure to compare two survey based measures of anti-Black prejudice. The racial resentment scale correlates poorly with the behavioral measure whereas the explicit prejudice measure correlates strongly. This paper helps to move forward the measurement conversation in the ethnic identity subfield of American politics. It is coauthored with Greg Huber and has been accepted for publication at the Journal of Politics.

In closing, the committee would like to recognize Peyton for his first-rate application of a wide range of experimental tools. The three papers of Peyton’s dissertation have already been recognized as important contributions in their respective literatures and we have no doubt that they will continue to be influential in the years to come.