Files
Abstract
This dissertation takes up the question of how interest group ratings contribute to our understanding of legislator ideology. By judiciously picking which roll calls are included in our estimation techniques, we can gain a more accurate perspective of how legislators line up on the liberal to conservative or other issue-specific scale. This behavior becomes problematic when the standard for distinguishing the “important” roll calls from the multitude of available votes rests on biased, partisan, or otherwise unsound grounds. After fleshing out the consequences of the biased and/or non-standardized subsetting behaviors of interest groups (both general and special-issue) during the creation of interest group ratings, I consider a mathematical standard for creating subsets of ideologically relevant roll calls by which we can efficiently order legislators along a liberal-conservative line. Through an examination of the ADA, ACU, and NEA over the course of more than thirty years, I find that the selection habits of interest groups for which roll calls to include in their legislator ratings actually serve as a mirror into their own political motivations. Additionally, I analyze the consequences of subsetting the roll call record into smaller sizes for the purposes of ideal point estimation in general. I find that a small subset does not necessarily produce inconsistent estimations, but certain kinds of votes can substantially alter the resulting ideal points relative to NOMINATE scores.