Americans are more sorted into distinct partisan camps than ever before. Affective polarization is growing, with Democrats and Republican becoming increasingly disdainful of one another. Democrats and Republicans appear to be moving toward opposing ideological poles as well—Pew data, for example, demonstrate that the gap between the median self-reported ideology of Democrats and the median self-reported ideology among Republicans is growing larger over time.
What can this mean for Philip Converse’s infamous argument that Americans are largely innocent of ideology (“The Nature of Belief Systems in Mass Publics,” Critical Review, 18(1–3), 1964)? Surely partisan-ideological sorting among the American public runs contrary to Converse’s claim that constraint in the belief systems of the citizenry does not mirror constraint among idea-elements visible at an elite level. If Americans are increasingly taking cues from an ever-polarizing party system, are they at least forming coherent ideological belief systems?
Donald Kinder and Nathan Kalmoe begin Neither Liberal nor Conservative with an exceptionally thorough review of Converse’s classic work; indeed it is one that should be assigned alongside the original work itself in any course on American politics. It is, though, much more than merely a review of Converse’s seminal work; it is an important intervention in the ongoing debate regarding the extent to which polarization pervades American public opinion.
The greatest feat the authors achieve in this book is the vigor with which they tackle the paradox of ideological innocence in an age of polarization. Americans might appear to be divided along ideological lines, but they are by no means ideological. With respect to partisanship, the authors argue, we are polarizing indeed; but ideologically we remain largely innocent, as Converse claimed.
First of all, the majority of us (literally over half) identify as moderate, and the authors take the relatively controversial position (one that I will return to) that “moderate” should not be considered an ideological category at all. In an Appendix devoted to this very issue, Kinder and Kalmoe calculate the ideological consistency of policy preferences among voters, and they find that moderates indeed commit to an incoherent mélange of viewpoints. The correlation between policy references among moderates is a starkly low 0.11.
If Americans are not particularly ideological now, however, the authors consider whether they are perhaps nevertheless becoming more ideological over time. Yet going back through 40 years of survey data, they find just a gentle but steady decrease in the proportion of Americans who call themselves moderate: 55% in 1972 versus 47% in 2012. Ideological extremists, on the other hand, have crept up from 3% in 1972 to 6% in 2012. This movement, the authors contend, is glacially slow and by no means indicative of a populace polarizing toward the extreme ends of the ideological spectrum.
The authors then demonstrate that self-identified liberals and conservatives do not disagree with each other when it comes to policy matters any more now than they did while Converse was writing his original work. If issue disagreement is illustrative of polarization, there is again nothing to see with respect to liberals and conservatives in America. Nor do liberals and conservatives appear to hold in-group/out-group biases against one another, as we do see with Democrats and Republicans. Again, going back 40 years, Kinder and Kalmoe find no consistent evidence that the two ideological camps are becoming less enamored with one another over time.
So why then have we sorted? Why are liberals more likely to identify as Democrats and conservatives more likely to be Republicans? The authors argue that this is, in fact, only true among the well informed. Just as Converse argued decades ago, there is indeed a small segment of the electorate who are knowledgeable and engaged and, it seems, they are largely responsible for the partisan–ideological sorting that is evidently occurring among the American people. Thanks to the polarization of Washington elites, informed Americans now choose the ideological label that best matches their preexisting partisan identity, but most Americans remain innocently “moderate” despite forming strong in-group biases with respect to their partisanship.
Kinder and Kalmoe anticipate some pushback, notably from scholars like Paul Sniderman, John Jost, and James Stimson who argue, respectively, that ideology and partisanship are locked together, that ideology exists and matters, and that Americans may hold views that align with one end of the ideological spectrum while personally identifying with the other. The authors defend their own views against these alternative arguments largely by pointing to the majority of Americans who identify as partisan yet claim to be ideologically moderate, which—in their view—signifies that they are not ideological at all.
Given the massive literature that underscores what we know and think about ideology, there are other views that I would be interested in hearing Kinder and Kalmoe address—particularly when it comes to moderates. For example, Shawn Treier and Sunshine Hillygus (“The Nature of Political Ideology in the Contemporary Electorate,” Public Opinion Quarterly, 73[4], 2009) demonstrate the tendency of cross-pressured individuals (those who hold both liberal and conservative views) to identify as moderate on a single ideological dimension. In my own work I too have found that individuals who hold both conservative and liberal views are drawn toward the midpoint on the ideological scale (Samara Klar, “A Multidimensional Study of Ideological Preferences and Priorities among the American Public,” Public Opinion Quarterly, 78[S1], 2014). Kinder and Kalmoe’s evidence that moderates show weak interitem correlations with respect to policy views might, in fact, complement this multidimensional argument. Going forward, this is an issue I hope these scholars will tackle head-on, as there are perhaps none better to address these questions.
Kinder and Kalmoe conclude their work by urging scholars and spectators of American politics to turn away from ideology for understanding mass attitudes and behaviors and instead to turn toward social groups. This point cannot be made enough, in my view, as group identities are a fundamental informational source in the course of preference formation. But must ideology be cast aside? Perhaps we can instead consider how ideology is intertwined in our identity politics. Liliana Mason (“‘I Disrespectfully Agree’: The Differential Effects of Partisan Sorting on Social and Issue Polarization,” American Journal of Political Science, 59[1], 2015), for example, shows that ideological strength is a significant determinant of affective polarization—even more so among those who have aligned their ideology with their partisan identity. As ideology and partisanship become increasingly aligned, out-group biases increase. It seems that ideology may, in fact, play an important role even as social identities increasingly dictate our political views. In what I hope will be a series of additional studies by these two scholars, this will be another area where their careful analyses will be of great service.
Scholars of political behavior, public opinion, and political psychology are all well served to read this thought-provoking book. Kinder and Kalmoe are tackling among the most difficult questions regarding preference formation in the mass public today. The degree to which voters are ideological with respect to issues, identities, or perhaps neither at al frames our entire understanding of public opinion in a polarizing time.
