This chapter explores the policy consequences of polarization – asking whether we see different policy choices when parties are more (versus less) polarized. This analysis is motivated by recent research suggesting a constraining effect of partisan polarization on redistributive policy – and contradicting earlier expectations that greater benefits will accrue to the poor when parties are competitive and offer clear policy choices. To better understand these consequences, we capitalized on variation in both policy choices and party polarization over time and across the 50 American states.

1. • Across eight distinct forms of policy redistribution, we found that party polarization tended to result in lower levels of redistributive policy.

2. • This was most pronounced for redistributive policies that are not indexed to inflation (so without new legislation, the real value of the policy decreases over time). These policies, such as the minimum wage, are particularly impacted by polarization-induced gridlock.

3. • For redistributive policies that do increase in magnitude as inflation, incomes, or earnings rise, such as the corporate tax rate, polarization had more of an indirect effect – shifting the impact of party control of the state government.

4. • These distinct paths of influence for different types of policies illustrate both the importance of polarization for policymaking and the complexity of these relationships.

In recent decades, political parties in the United States have become more ideologically distinct (Rohde 1991; Aldrich 1995; Fiorina 1996; Poole and Rosenthal 1997). During the same time period, we have seen a parallel increase in income inequality (McCarty, Poole, and Rosenthal 2006), as well as the transformation of social welfare policy away from entitlements and cash assistance toward a work- and market-focused system of neoliberal reform (Pierson 2001; Korpi and Palme 2003; Hacker 2004; Schram, Fording, and Soss 2008; Allard 2009). Accounts of these political developments often draw linkages between these trends – describing how party polarization contributes to gridlock and policy drift that tends to disadvantage those with less power and organization (Hacker and Pierson 2010).

The most explicit examination of a relationship between party polarization and policy outcomes is found in McCarty (2007) and McCarty, Poole, and Rosenthal (2006), who describe a “dance of inequality” in which the upward trend in party polarization tracks alongside the declining real value of the minimum wage, as well as regressive shifts in estate and income tax policies.

Political parties hold a special place in modern theories of democracy. They were not popular when anticipated by the founders of the U.S. Constitution (see Madison's Federalist Paper no. 10), and they are not generally thought well of by the general U.S. public today (Dennis, 1978; Wattenberg, 1986). Nevertheless, it is difficult to envision representative democracy in complex societies without competitive political parties.

Two different models of party competition are commonly presented, usually as if the two models are in direct competition with one another. One is the “responsible parties model, ” which as a prescriptive doctrine promotes ideologically distinct political parties. The other, sometimes called the “public opinion” model (Page, 1978), is more commonly called the “Downs” model after the theorist Anthony Downs (1957). While it is common to juxtapose these two models as competing with one another, they are in fact complementary, each drawing on different aspects of party motivation. The responsible parties model emphasizes the policy or ideological motivations of party elites, which push party positions away from center. The Downs model emphasizes the electoral incentives pushing politicians toward the center in order to get elected. As we will see, the positions of state political parties represent both sets of forces.

The responsible parties doctrine was a product of political science thinking in the 1940s and 1950s. Inspired by British electoral experience, its advocates (Schattschneider, 1942; Committee on Political Parties, 1950) favored strongly disciplined and cohesive political parties that would offer distinct policy programs. According to the design, responsible parties were to be internally democratic, so that party activists would control the content of party programs.

One central question about modern democracies is, How much do partisan elections contribute to the effective representation of public opinion? In theory, programmatic parties provide the electorate with a clear choice of policy options, with election results followed by predictable policy consequences. But empirically, does it matter which party (or coalition) governs? Most recent cross-national comparisons of Western democracies answer affirmatively. Socialist or leftist party participation in national governments is associated with a high inflation to unemployment tradeoff (Hibbs, 1977), growth of the governmental sector (Cameron, 1978; Lange and Garrett, 1985), high public expenditure (Tufte, 1978; Castles, 1982), welfare spending (Castles, 1982; Hicks and Swank, 1984), quality-of-life (Moon and Dixon 1985), and income equality/ redistribution (Tufte, 1978; van Arnhem, Corina, and Schotsman, 1982; but see also, Jackman, 1975, 1980). In sum, nations where leftist parties are most influential also tend to have the kinds of policies and policy consequences with the greatest appeal to leftist constituencies. Time-series analyses for individual nations also show policy consequences resulting from party control (Hibbs, 1977, 1987; Alt, 1985). Party control does matter.

It may be the case, however, that party control matters only in terms of the sharp divisions between European socialist parties of the left and parties of the right. In the U.S. context, it is by no means clear that major policy differences flow from variation in Democratic versus Republican control. At the national level, while some time series find policy consequences from party control of the presidency and/or Congress (Hibbs, 1977, 1987; Kiewiet and McCubbins, 1985), the role of parties remains in some dispute (Beck, 1982; Browning, 1985; Lowery, 1985).

In recent years, citizens of many nations have struggled to establish democratic institutions where none had existed before. As political scientists, we can ask what justifies this passion for “democracy?” Part of the answer is easy: political freedom. Empirically as well as in theory, where democratic institutions prosper, people are freer to speak and act without fear of arbitrary intrusion from government authority. But freedom from government intrusion is not the sole justification for democratic government. In theory, the democratic ideal of popular sovereignty means that collectively, citizens can actively shape what their governments do. The relevant empirical question becomes whether, in practice, democratic institutions allow public opinion to influence government policies very much. Modern political science is still working on the answer to this important question.

This book addresses the question of democratic representation for one set of democratic governments: those of the 50 separate states of the United States. The Constitution of the United States reserves many government powers to the states. The policies that states enact are often described in terms of their ideological content as relatively liberal or relatively conservative. We use this ideological dimension to assess the correspondence between public opinion and policies across the states.

We began this project several years ago, from a sense that public opinion had been seriously neglected in the political science literature on state policymaking. As we assembled our statistical evidence, we found a pattern that was even stronger than our initial suspicions: State ideological preferences appeared to dominate all other variables as a cause of the ideological tilt of a state's policies.

When analysts classify the citizens of the various American states in terms of their political differences and similarities, the standards most often imposed are in terms of partisanship and ideology. One can generally classify a state's partisan tendencies with some assurance, since we can make inferences from state voting patterns and, in some instances, party registration. Classifying electorates on the basis of ideology is riskier for the reason that available indicators of ideology are at best indirect. Politically knowledgeable observers do commonly attribute liberal or conservative tendencies to state electorates. Among the factors that enter into these impressionistic judgments are the states’ electoral affinities for liberal and conservative candidates, the ideological proclivities of their congressional delegations, and the general imprints of their unique political histories. Still, it is far from certain that state electorates’ ideological reputations are deserved. It may be an unwarranted leap of democratic faith to attribute ideological motive to ideological consequences.

For our investigation of the public opinion-policy connection in the U.S. states, the most crucial challenge is the measurement of state-level public opinion. This chapter presents our measures of state-level ideological identification and partisan identification, from CBS/NYT polls. Our overall strategy is straightforward. We simply aggregate by state the responses to the 122 national CBS/NYT telephone polls for the period 1976-88. These polls are conducted on a continuous basis, maintain the same questions for party identification and ideology throughout the time period, and use a sampling design that serves our purposes very well.

Popular control of public policy is a central tenet of democratic theory. Indeed, we often gauge the quality of democratic government by the responsiveness of public policymakers to the preferences of the mass public as well as by formal opportunities for, and the practice of, mass participation in political life. The potential mechanisms of democratic popular control can be stated briefly. In elections, citizens have the opportunity to choose from leaders who offer differing futures for government action. Once elected, political leaders have incentives to be responsive to public preferences. Elected politicians who offer policies that prove unpopular or unpleasant in their consequences can be replaced at the next election by other politicians who offer something different.

Of course, this picture describes only the democratic ideal. A cynic would describe the electoral process quite differently: Election campaigns sell candidates in a manner that allows little intrusion by serious issues. Once in office, winning candidates often ignore whatever issue positions they had espoused. Voters, who seem to expect little from their politicians, pay little attention anyway.

The actual performance of any electoral democracy probably falls between these extremes. Acknowledging the factors that impede effective democratic representation, we can ask to what degree does public opinion manage to influence government decisions? This is an empirical question, often noted as the central question of public opinion research (Key, 1961; Converse, 1975; Burstein, 1981; Kinder, 1983; Erikson, Luttbeg, and Tedin, 1991).