Publication Type: Conference Paper/Unpublished Manuscript
Abstract: Although economic integration has progressed steadily during the postwar period, Congressional votes related to trade, economic dislocation, and associated social policies are consistently controversial. Some scholars have explained the endurance of trade barriers by arguing that protectionist interests have organizational advantages (Schattschneider 1935; Destler 1986; Cassing, McKeown and Ochs 1986). More recently, however, Bailey (2003) has claimed that diffuse interests, particularly skilled labor, exert an important and consistent influence on congressional votes by virtue of anticipated reaction (Kingdon 1973; Denzau and Munger 1986; Arnold 1990; Verdier 1994). The logic is that representatives serve diffuse groups of voters because failing to do so will lead rival politicians, interest groups, the media and the president to activate interests that benefit from liberalization.

These competing claims speak to broader debates about the extent to which constituent interests and preferences motivate legislators’ votes. There are, however, three limitations to extant work. First, in assessing the relative influence of constituency, almost all of the studies employ aggregate measures of constituency interest, rather than actual opinion in the member’s state or district. Second, previous work generally only specifies free-trade versus protectionist interests and does not consider whether representatives’ votes reflect the mean interests of the whole state or of the member’s partisans (e.g. partisan delegates). Third, few examinations of trade votes analyze related social policies, which are more likely to be partisan than trade votes. In this way, previous studies may underreport party influence in matters related to economic integration. In this paper, we test the linkages between constituency opinion and key trade and social welfare votes in the Senate from 1993 to 1994 using data from the ANES Senate Election Study. Employing micro-level evidence, we find support for the impact of constituent policy preferences, ideological beliefs, and partisan identification on Senate votes.