Replication Data for: Candidate Positions, Responsiveness, and Returns to Extremism

The concept of candidate positioning is central to the study of U.S. elections, representation, and political behavior. Existing work, however, overwhelmingly relies on indirect measures which may not reflect candidates’ stated positions. I analyze foundational relationships between candidate positions and district partisanship, primary electoral success, and primary fundraising performance with existing approaches versus text scaling estimates based on an original collection of campaign platforms from House primary candidates' websites in 2016, 2018, 2020, 2022, and 2024. Directly measuring candidates' positions using campaign platforms leads to conclusions vastly different than those reached with existing measures. While platform-based measures suggest candidates are responsive to their districts, existing measures do not. Within district, however, existing measures show financial and electoral penalties to extremism in primaries, but platform-based measures show no such penalty. These findings have wide-ranging implications for a number of ongoing scholarly debates which involve congressional candidates' positions.