A Method of Estimating the Personal Ideology of Political Representatives

Abstract
We suggest a method for estimating a political representative's personal ideology and its effect on his or her voting decisions. The current practice of using the ratings of a pressure group such as Americans for Democratic Action (ADA) as a proxy for personal ideology is shown to have a number of theoretical and interpretive flaws. Our technique uses the residuals from two regression equations to provide answers to two questions: Is there a systematic ideological component to the voting behavior of political representatives after taking account of other political determinants, and if a systematic ideological component exists, is it possible to determine its role in voting on particular issues? The technique developed and the currently accepted practice are compared using votes on labor issues as an empirical example.