Tests of Vote-Buyer Theories of Coalition Formation in Legislatures

Abstract

Formal theories of “vote-buying” aim to explain legislative coalition building, and lobbying. While anecdotal evidence suggests that something approximating vote-buying occurs, these theories have not been subjected to substantial empirical tests. Using roll-call data from all House bills subject to votes on final passage in the House of Representatives in the 103rd Congress, this study tests the coalition size implications of vote-buying theories. The primary method of analysis is ordinary least squares. Variation in coalition size, as represented by the percentage of the chamber voting yea, is accounted for in a manner consistent with vote buying theories. By making an assumption about what kinds of legislation are likely to attract one, or multiple, vote-buyers, we are able to operationalize the number of vote-buyers lobbying on a given bill. Other theoretical predictors of coalition size are also entered into the analysis, and the results offer support for the family of theories. When controlling for different parliamentary procedures, bills that are likely to attract two vote-buyers generate coalitions that are approximately 6 percent larger than single vote-buyer bills.