[The Sunny Dog]


The Theory of Realigning Issues: Why the Issue Structure of Partisan Conflict Changes Over Time


A "brown state-green state clash is likely to encumber any effort to set a mandatory ceiling on the carbon dioxide emissions blamed as the biggest contributor to global warming," reports The New York Times. But the report is puzzling. Congress regularly legislates in the face of much deeper divisions between Red and Blue state delegations on such matters as taxes, health reform, and regulation of the economy. Why do Brown-Green divisions hold up legislation on the environment and the Red-Blue divisions do not hold up legislation on taxes and spending? The answer, most political scientists would agree, is that Congress is organized around issues of economic conflict. Its choices of leaders, procedures, and party reputations are all based on putting together legislation on economic issues. Environmental legislation cross-cuts the usual battle lines. Some Democratic stalwarts, like organized labor, are liberal on most issues but oppose green legislation, whereas some Republican constituencies, like suburban voters, are generally conservative but favor certain green measures. The result is that Congress lacks the means to forge and enforce deals on green issues. This problem appears to apply to all issues that cross-cut the main dimension of congressional conflict. In fact, as I show in the dissertation, if legislator preferences on an issue are orthogonal to the ideological dimension, significant legislative proposals addressing that issue stand only a 3 percent probability of enactment. In comparison, if legislator preferences on the issue are perfectly aligned with the ideological dimension, the probability of enactment is 63 percent. This finding motivates the question: Why are some issues organized into the dominant congressional cleavage and some are not?

Dissertation Summary

United States Congressional history is filled with small changes in the polarization of the parties on different issues and sometimes outright flip-flops. Why, then, do the issues that structure party conflict change over time? My dissertation offers a theory that explains those changes. This theory of realigning issues holds that changes in party conflict are a product of the changing structure of interests in society. When the interests of two different groups become more compatible, new coalitional opportunities emerge. Large changes or "realignments," therefore, are not driven by changing interests themselves but by changes in the relative compatibility of societal groups with the major party coalitions. I argue that episodes of realignment are set off by the activation of realigning issues. Realigning issues are issues that become central to organizing partisan conflict when new coalitional partnerships either commit the coalitions to polar positions on a salient issue or release one of the coalitions from an electorally disadvantageous constraint on a salient issue. For example, the coalitions polarized on slavery when westward expansion of the United States enabled the development of an anti-slavery coalition; and the Democratic Party was able to moderate on race in the 1960s when the northern migration of African Americans provided a coalitional alternative that relaxed the constraint imposed by the Party's dependence on the South. By my account, then, changes in the interest composition of the coalitions caused by realigning issues should lead to predictable changes in how polarized the parties are on all issue domains. To test this theory, I construct an original data set of House roll call votes (1st to 109th Congresses) coded according to the Katznelson-Lapinski policy typology and measure MC issue preferences with DW-NOMINATE scores that scale the subset of roll call votes for each of the major policy categories. I find strong support for my hypothesis by analyzing changes in the levels of inter-party polarization on national defense, social policy, and immigration over the course of American history.

Example Finding

A central hypothesis of my dissertation is that changes in the polarization of the parties on different issues can be explained by how the interest composition of the coalitions is transformed by the realigning issues that have caused transitions between party systems. An issue will be increasingly organized into partisan conflict if preferences on that issue correspond to preferences on the realigning issue and will be displaced from partisan conflict if little correspondence exists. A basic test of this hypothesis is to use MC issue preferences during the mature years of the old party system to predict how much partisan polarization on an issue will change over the course of the new party system. Here is an example where I use MC preferences on the Katznelson-Lapinski issue categories in 1947-62 to predict how much change in partisan polarization on an each issue would develop between the end of the Fifth Party System and today as a result the civil rights realigning issue:
Examples of Variation in Issue Polarization

Here are some examples of changes in the extent to which an issue is organized into congressional conflict over time. My goal is to explain this variation. These plots are for (1) defense, (2) agriculture, and (3) planning and resources. I use first dimension DW-NOMINATE (DW1) as the measure of coalitional alignment, and I use the aggregate proportional reduction in error (APRE) statistic as the measure of association between DW1 and each issue area. These plots present the APRE statistic for all House roll call votes on each respective issue.