Replication data for: Representation in Congress: Constituents and Roll Calls in the 106th House

Aggregating 100,814 randomly selected respondents to measure sub-constituency preferences provides an unprecedented ability to measure sub-constituency preferences in the House. Looking at the relationship over all votes, "key votes," and on individual votes confirms that representatives are not completely responsive to the district mean voter, that only majority party Republicans are especially responsive to the preferences of same-party constituents, and that same-party constituency preferences cannot entirely account for systematic differences in Republican and Democratic voting behavior.