This article examines the influence that state party registration laws have on individual-level party attachments. It tests the hypothesis that individuals living in states with party registration laws are more likely than those living in states without such laws to identify themselves as partisans. This occurs primarily because of self-perception processes by which registrants infer their party attitudes from their own behaviors. Using the state-based Senate Election Study data to test this expectation, we find strong evidence for both statistically and substantively significant effects of party registration on individual partisanship. Registered individuals living in states with party registration are, by about ten percentage points, more likely to identify as partisans than those in other states even when controlling for alternative hypotheses dealing with state culture, attitudes toward the parties, retrospective evaluations, interest in politics, and demographic factors. Importantly, the effect is not observed for individuals who are not registered to vote but is for registered nonvoters. However, registration-induced party identification is shallow, as individuals living in registration states are also more likely to vote for candidates from the other party.
In this article we consider the influence that state party registration laws have on party identification. We argue that the presence of a state party registration law fosters individual-level party attachments. In all conceptions of party identification, whether partisanship is conceived of as exogenous or endogenous, partisan attitudes are thought to drive behavior. In contrast, we demonstrate that in some cases partisan attitudes are actually inferred from behavior-in this case, registering with a party Using self perception theory as a plausible mechanism, we demonstrate the effects that the process of formally registering with a political party has on citizens' party loyalties. According to the theory, individuals infer many of their attitudes from observing their own behaviors. For voters lacking strong predispositions, the act of verbally or physically registering with a party ought to make one more likely to consider one attached to that party based on self-attribution. We test this hypothesis using state-based surveys conducted between 1988 and 1992 to assess the direct impact of registration laws on individual-level party attachments. Even when controlling for potential alternative variables at the individual and state levels, we find consistent and strong support for this expectation, strengthened by the fact that the effect is observed only for individuals who are registered to vote.
The authors of The American Voter believed that party registration requirements could affect voters' political attitudes. In a chapter entitled "Election Laws, Political Systems, and the Voter" Campbell and his Michigan colleagues outlined a fundamental relationship between electoral regulations and partisanship. They believed that "the rules governing the conduct of partisan politics constitute an important aspect of the individual's political environment that bears directly upon our analysis of electoral behavior" (Campbell et al. 1960: 267). Using bivariate tests, they found some evidence that respondents who lived in states with laws facilitating partisanship through party registration and closed primaries had higher percentages of partisans than did other states.
Others have also examined this relationship at the aggregate level. Jewell (1984) found that states with relaxed, open primary systems had more Independents than closed primary states requiring party registration. In a more comprehensive analysis Finkel and Scarrow (1985) demonstrated the strong relationship between state registration laws and the proportions of Independents and partisans. Indeed, they found a .87 correlation between a state's rank in Independent party enrollment and Independent identification. …