Skip to Navigation

Polarization in America: An Interview with Delia Baldassarri

Rohan Mascarenhas, Russell Sage Foundation
March 20, 2012

election-2012As Republican voters in Illinois head to the polls, we're pleased to announce a new feature on RSF Review: the Election 2012 series. Through the campaign season, we hope to add insights, data and commentary from our scholars and research programs to the political conversation. In our first installment, RSF Visiting Scholar Delia Baldassarri discusses the conventional wisdom about political polarization in the United States—and why it may be wrong.

Q: Although it is widely assumed that public opinion in America has sharply polarized over the past 40 years, your analysis of the evidence and scholarly literature suggests a more nuanced conclusion. What does the research show?

A: Over the last four decades, and especially starting in the 1990s, public opinion increasingly divided on a few specific issues, such as abortion, gay rights and the war in Iraq, while we observe relative stability and even depolarization on all remaining economic, civil rights, social, and foreign policy issues. Indeed, Americans are actually less divided in their opinion on the role of women in society, racial integration, and criminality than they were forty years ago. As new generations replace old ones, and women achieve higher positions in society, the collective mood shifts accordingly.

However, there is clear evidence of polarization among political partisans: those who are politically active, who identify with a party or strongly identify as liberal or conservative, tend to have more extreme positions than the rest of the population. In sum, while the distribution of opinions across the population has not changed a lot, with the exception of a few ‘hot-button’ issues, Republicans and Democrats are significantly more divided on a wide range of issues. This is a process of partisan alignment, which is partly consequence of the increased polarization of the parties, Congress, and political activists: since parties are more polarized, they are now better at sorting individuals along ideological lines.

Q: Do you have any theories for why pundits believe attitudes are increasingly polarized?

A: There are a few reasons that might lead lay observers to believe that political polarization is much more widespread than it actually is. First of all, they might attend to the wrong population. Namely, they might focus on political activists and leaders, and assume that their ideological positions reflect those of the population at large. Second, they devote most of their attention to issues that are focus of political debate, such as abortion, or gay rights, and these issues, often for limited periods of time, tend to be more polarized.

Finally, there is a third, and more pervasive reason that may induce all of us, and especially those who are interested in politics, to perceive greater division than there actually is. I’ll give you an example. I lived in New York City for several years, but I have met only 3 Republicans. Being an academic, and a sociologist, makes me more likely to meet Democrats, but this cannot account for the very small number of Republicans I know. The truth is that I have probably met many more Republicans, but we never talked politics. People will talk about trivial matters—the weather or what to have for dinner—with most anyone, but for topics that are important to them, they reveal their views more selectively, disproportionately choosing people who they believe share views broadly similar to their own. Because people tend to talk to others who share their beliefs, the discussion of important issues will always induce, at the microlevel, the experience of a very homogeneous environment, due to the fact that people reveal their view selectively in order to avoid conflict. Also, since I tend not to talk politics with Republicans, it is easier to ‘demonize’ them, to think that they are Limbaugh-like, thus very different from me. Overall, this leads people to believe that they live in a politically divided country.

Q: There is a worry that increased polarization poses a danger to democracy, as it will hurt dialogue and cooperation. You suggest that evaluating this risk partly depends on the level of "issue constraint" – could you define the term? How has the level of issue constraint changed over the years?

A: Usually, people think of polarization as a process of radicalization and increased extremisms along a single dimension. In this perspective, people would say that public opinion is polarized if there are very few individuals holding moderate views on, for example, abortion, and most people are either strongly pro-life or pro-choice. However, an integrated society is not a society in which conflict is absent: there will always be pro-choice and pro-life supporters, or people in favor of the welfare state, and people that oppose government spending for the poor. Rather, an integrated society is one in which conflict expresses itself through crosscutting interests and identities. I conceive of polarization as a process of alignment along multiple lines of potential division, which organizes individuals and groups around exclusive identities. If people align along multiple dimensions (for instance, if most pro-choice citizens were also against government spending while the pro-choice people are also in favor of welfare state intervention), even if they do not take extreme positions on single issues, the end result would be a polarized society.

The empirical question therefore is whether, along with the partisan alignment described before, there has also been a process of issue alignment, namely whether individual’s opinions on economic, civil rights and moral issues has become increasingly more correlated. Here, the finding is quite intriguing. There is a consistent subgroup of people, around 30 to 40% percent of the population, for which we observe increased "issue constraint." Simultaneously, there is a second group of people, as large as the first one, that have experienced a process of ‘de-alignment’ between economic and civil rights issues, on the one hand, and moral issues, on the other. This latter group has increasingly rejected the notion that economic and moral conservatism (or liberalism) necessarily entail one another. Who are they? These are 'rich but secular' or 'poor but religious' citizens whose combinations of economic interests and religious identities make it particularly difficult to be consistently conservative (or liberal) on both moral and economic issues. Indeed, they deviate from the orthodox understanding of politics, adopting a political logic in which conservatism and liberalism are not entirely at odds. Interestingly, these individuals holding alternative preferences tend to be influenced by their conservative tendencies (irrespective of whether conservatism is on moral or economic issues), and vote in favor of the Republican Party.

Q: How does the polarization of the political elite – activists, leaders, etc. – affect public opinion in general?

A: The current trend in political polarization is mainly a top-down phenomenon, mostly driven by parties, interest groups, and political activists. The increased partisanship and polarization in Congress is the by-product of a process of selection and reinforcement according to which moderate representatives have been replaced by more ideological ones, and those who maintained their seats had embraced a more ideological agenda to be re-elected. In addition to the re-alignment of the Southern states, this process has been triggered by the growing cost of the campaigns and parties increased dependence on interest groups, political activists, single-issues advocates, PACs, and now even superPACs.

The partisan polarization of public opinion should be interpreted mostly as a consequence of parties’ increased polarization, rather than a factor that has caused it. Looking at the glass half-full, some scholars have actually argued that increased partisan divisions might increase citizens’ interest in politics. Moreover, a widely accepted argument in the literature on partisan alignment is that party polarization has made it easier for voters to identify with a party because parties have become more distinguishable on a broad set of issues. However, there are reasons to believe that parties' growing division in their stances on moral, economic, and civil rights issues have also made it more difficult for certain Americans to identify wholeheartedly with either the Republican or Democratic party. The reason for this is quite intuitive. At least to some extent, people's political preferences map into the their socio-demographic characteristics: for instance, poor people and blacks favor welfare state interventions and tend to support the Democratic party, and regular churchgoers, at least in recent years, have turned to the Republican party. However, how can a low-income, highly religious African-American voter reconcile liberal tendencies on economic redistribution and civil rights with moral conservatism? Similarly, will a wealthy, non-religious voter identify with the Republican party for its economic policies, or with the Democratic party's moral progressivism? In other words, in a context in which parties have clearly defined, alternative positions on multiple issue dimensions, voters who do not fully subscribe to party's positions on all dimensions may find it harder to define their political allegiance.

Q: Which population subgroups tend to have the most coherent political beliefs? What are the implications for the equal representation of opinions?

A: As it is usually the case, those who are more interested in politics are more coherent in their political preferences than the remainder of the population, and this gap has substantially increased in the last few decades. A similar and more striking pattern is observable among the richest third of the population, who have become more coherent in their political preferences, and in the relation between these preferences and partisanship, while the poorest have remained essentially inconsistent. The wealthier part of the population knows well what it wants, and it is likely, now more than in the past, to affect the political process, not only through campaign financing and lobbying activity, but also in the ballot.

A comprehensive discussion of inequality in political representation should definitely start from the role of money in politics, interest groups' capacity to influence the political agenda, and a series of other aspects that affect the electoral and legislative processes. Here I limit myself to a few considerations concerning the relationship between parties and their constituencies. As a consequence of the process of partisan polarization, extreme positions have gained prominence within the two parties. Given the partisan realignment, the average opinion within partisan subgroups is now more extreme. Party voters, having become more consistent in their political preferences, are likely to convey more extreme preferences to their party leaders. Finally, since party voters are now more divided, they constitute an easily identifiable target for a party elite concerned with preserving its constituency. To the extent that parties pay some attention to voters in defining their strategies and political agenda, nonvoters, by not showing up at the polls, are undermining their representation capacity both because they do not get to choose their representatives and because parties’ strategies are less likely to consider their preferences.