My research examines how political institutions shape the behavior of public officials and their accountability to the people they serve. I pursue this question in three areas. My work on representation asks how racial identity shapes legislators' strategies for navigating institutional constraints rooted in marginalization. It also asks how well elected officials represent constituencies where racial identity is obscured by the narratives of nationalist and neo-colonial projects. My work on electoral systems turns to the rules that put legislators in office, examining how reforms such as alternative primary systems and congressional redistricting influence legislative behavior and outcomes. Finally, my public policy research follows this question from legislatures into bureaucracies, studying how policies that integrate artificial intelligence (AI) into policing can improve transparency and governance.

Methodologically, I combine causal inference with computational approaches. In my work on race and representation, I often leverage electoral turnovers as quasi-experiments and estimate their effects with matching, difference-in-differences, synthetic counterfactual methods, and nonlinear models. My work on electoral systems relies on the same matching and synthetic counterfactual methods but adds computational tools such as simulated ensembles of district maps and geospatial analysis. In my policing and AI policy research, the emphasis shifts to measurement, where I use established tools for natural language processing and multimodal analysis. More recently, I have also focused on approaches to integrating large language and multimodal models into social science research—particularly for content analysis and measurement across text, images, audio, and video.