My research contributes both to quantitative methodology and American politics. In research methods, I develop tools for improving the connection between statistical methods and social science theories, and making reliable inferences based on non-experimental data. My goal is the development of methods and models that reflect with fidelity the constraints implied by substantive theory, while also minimizing the use of ancillary modeling assumptions unmotivated by theory. Substantively, I focus on evaluating the structure of political conflict and the quality of political representation in the United States.