This article examines whether organized interests alter the strong opinion-policy linkage observed by Erikson, Wright, and McIver (1993). We first replicate the EWM model circa 1980 with the addition of measures of interest organization density and diversity under a variety of specifications. We then extend the analysis to the contemporary period, using EWMs' new public opinion data through 1999. To execute this second stage of the analysis, we construct a new index of policy liberalism circa 2000 comparable to EWMs' 1980 measure. Our results suggest that interest community bias in favor of traditional economic interests modestly weakens policy liberalism responsiveness, but that simply having many organized interests registered to lobby may promote more liberal policy.
The classic question of whether a democracy's policies are responsive to its citizens' wants has been investigated in many different contexts. In the American states, the question is usually posed as one of policy congruence-the extent to which a state's policies are congruent with citizens' desires as expressed through public opinion polls. Early studies of policy congruence or responsiveness were hampered by lack of public opinion data comparable across the fifty states. Scholars (e.g., Weber et al. 1972) instead simulated citizen opinion from state demographic data, using these proxy measures to test for policy responsiveness. Erikson, Wright, and McIver 's (EWM) (1993) State-house Democracy reinvigorated and revolutionized the study of opinion-policy congruence using aggregated CBS/NYT polls over several years to develop a reliable and valid measure of opinion ideology. They found that state policies were highly correlated (0.82) with opinion ideology and that the impact of traditional demographic factors paled in comparison. EWM found that public opinion exerted both a sizeable direct effect on policy and an indirect effect through both party elites and party control of state legislatures. State policies generally reflect the ideological sentiment of a state's citizenry.
Although EWMs' model is a classic in the state politics subfield, it also provoked others to suggest improvements in their model. Generally, these center on additional variables that might plausibly condition the relationship between public opinion and public policy over time or space.1 This research focuses on the role of organized interests in the opinion-policy linkage process: does the structure of communities of organized interests alter or perturb the translation of public opinion into public policy? If so, how? Our investigation will proceed in three steps. First, we discuss several-and sometimes compeling-hypotheses about how organized interests might perturb opinion-policy congruence. We then replicate the EWM model circa 1980 with the addition of measures of interest system density and diversity in 1980 under a variety of specifications. And last, we extend both the EWM analysis and our additions to it to the contemporary period, using EWMs' new public opinion data through 1999. To execute this third stage of the analysis, we construct a new index of policy liberalism circa 2000.
ORGANIZED INTERESTS AND OPTNION-POLICY CONGRUENCE
Despite the arguments of David Truman (1951) and other pluralists, organized interests are often assumed to thwart public opinion. Jeffrey Berry ('1997: 19) sums up a lengthy catalog of their sins: "the popular perception is that interest groups are a cancer spreading unchecked throughout the body politic, making it gradually weaker, until they eventually kill it." Though many political scientists are more positive about the role of organized interests in democracies, a significant tradition of scholarship argues that the interest system is biased and, therefore, unrepresentative of the popular will. Schattschneider's assertion (1960: 34-35) that "The flaw in the pluralist heaven is that the heavenly chorus sings with a strong upper-class accent" expresses one type of bias, a class bias. …