Jeffery Jenkins, Bedrosian Center Director, and co-author Thomas Gray will be presenting a paper at the 2018 Congress & History Conference at Prinecton University on June 7-8, 2018.
The paper is titled “A Bridge Too Far?: Examining Bridging Assumptions in Common-Space Estimations.” It will be presented as part of a three paper panel on “Methodological Advance in the Study of Congress,” and discussed by Robert Erickson and David Bateman.
“A Bridge Too Far?” focuses on a subset of House members who move to the Senate to examine the constancy of common-space voting assumptions.
While most common-space estimations rely upon members who served in both the House and Senate as “bridges” to scale the remaining members, this assumes that these “bridge members” do not change their preferences when they change chambers. Such an assumption conflicts with standard notions of representation, that is, that legislators’ votes reflect (at least to some degree) the wishes of their constituents. And a House member’s district constituency is often very different (smaller and more homogenous) than a senator’s statewide constituency. We examine the constancy of this common-space voting assumption, by focusing on a subset of House members who move to the Senate – those who come from statewide House districts. Using these members as the bridge actors (and thus bridging by constituency explicitly) in a one-dimensional IRT model, we find that the standard assumption of chamber-switchers in common-space estimations is technically, but immaterially, false. While there are statistically distinguishable differences in House and Senate voting records for chamber switchers, they are not sufficiently large to meaningfully undermine bridging. These results largely support the work and assumptions of Poole, Rosenthal, and others who have used these bridging assumptions to create