District Competitiveness and Legislative Roll-Call Behavior: A Reassessment of the Marginality Hypothesis

Noting the conceptual and methodological confusion which has surrounded previous examinations of the marginality hypothesis, this paper directly tests its two major predictions: (1) The relationship between district opinion and roll-call behavior is greater in competitive than in noncompetitive districts; (2) The relationship between party position and roll-call behavior is less in competitive than in noncompetitive districts. Neither hypothesis is completely supported when examined across three policy dimensions. The findings raise serious questions about previous studies which have relied upon party loyalty measures.