Abstract
There is an interesting institutional difference between political science and sociology as it is organized in most American universities. Political science—like, in many ways, anthropology and psychology—is divided relatively rigidly into subfields, each of which functions relatively autonomously. Like disciplines in general, these subfields—American politics, comparative politics, international relations, political methodology, and political theory—tend to be simultaneously very productive internally and somewhat insular with respect to cross-subfield work.
Sociology’s subdivisions are substantially less rigid, which results in an interesting trade-off between subdisciplinary progress and cross-fertilization. In this somewhat haphazard review of several recent books from political science—recommended by University of North Carolina political science colleagues—I will highlight strengths from which sociologists might borrow as well as intellectual weaknesses that might benefit from sociologists’ insights.
An important strain of current political science, particularly in American politics, is boldly reductionist, seeking to isolate root causes in the characteristics of individual voters or single pathways. This paradigm is very rich and productive, and it has produced some of the field’s most enduring findings (think, for example, of The American Voter, which has essentially defined the study of voting for over 50 years). But it is also quite constraining, as scholars in this paradigm seek to demonstrate that one specific set of predictors is truly foundational. These quests for foundational causes often end up ignoring the mediating and moderating processes that may well have more important effects than the foundations. In this review I begin with detailed discussions of two books in this tradition, then consider several others that, in one way or another, seek to transcend it.
The poster child for the reductionist approach is Predisposed, by John R. Hibbing, Kevin B. Smith, and John R. Alford. It builds on years of work by this team on genetic causes for political viewpoints. Much of this work uses twin studies to partition the variance in political preferences among genetic, shared environmental, and unshared environmental causes. The heritability of a political outlook is the extent to which that outlook is more similar between monozygotic (“identical”) twin pairs than it is between heterozygotic (“fraternal”) pairs. There are other methods, too, including focusing on specific candidate genes and using Genome-Wide Association Studies (GWAS), both of which aim to identify specific genetic codes for political traits.
The book’s central claim is summed up well on page 91: “Just as flies’ taste for beer is biologically based and relates to their behavior, humans’ tastes are often biologically based and relate to their behavior, right down to political orientations.” Why? Well, political orientations are ways of comprehending the world. Different people react to the same stimuli differently. “Each person experiences the world differently because the biological machinery responsible for that experience . . . differs from one person to the next” (p. 114; emphasis added). Much of the actual argument proceeds by analogy. Brain injury, for example, can cause personality change (pp. 147–148), so, by analogy, brain structures can cause political orientations. Polygraphs use physiological indicators to assess psychological states, so, by analogy, physiological characteristics might cause beliefs and behaviors, including political ones. In each case, though, the extension to political orientations originates with a much simpler example; the reader is asked to ignore the fact that political orientations are more complicated than whether or not one is knowingly lying.
This ontological complexity re-emerges as a methodological point in Chapter Seven (“Politics Right Down to Your DNA”). It turns out that attempts to locate specific genes for political orientation have been generally disappointing, though they have turned up some weak associations. GWAS approaches suffer from many false positives, and “findings from molecular genetic studies have not replicated well” (p. 192). Hibbing et al.’s view is that “fairly small genetic differences get magnified by environmental forces to create distinct political predispositions” (p. 227). Political conversions are identified as “going home” (p. 227)—as realigning political position with natural inclination. But this view of environmental forces enormously underestimates the complexity of environmental influences—yes, they may magnify genetic predispositions, but they may also distort or diminish any such predispositions. And, most interestingly for sociologists, they may have contradictory effects at different analytical levels and in different historical contexts.
The most basic claim in Predisposed is likely both true and not particularly important: the claim that there is some non-negligible heritability involved in people’s political development. That is because broad-brush personality traits are modestly heritable, and personality presumably figures in individuals’ political orientations. But in order to become political outlooks, these traits have to be refracted through a thick haze of mediators: political institutions, party coalitions, epistemic cultures, media regimes, historical contingencies, and emotional states, to name a few. And by structuring, teaching, and distorting those traits, that haze is likely a far more important cause of political outlook than the presocial dispositions examined here.
Some of this complexity is on display in Milton Lodge and Charles S. Taber’s The Rationalizing Voter. Like Predisposed, this book is the capstone on a long scholarly literature: the motivated reasoning literature. In a nutshell, this literature—based largely on experimental designs—holds that most, if not all, political reasoning is essentially retrospective. Rather than assessing the evidence and arguments to determine a position, people begin with the position they prefer and rationalize that position by selecting and reinterpreting evidence to fit. This rests on current thinking in psychology and neighboring disciplines about “dual-process models” of cognition: cool, deliberative, conscious cognition is one process, but it tends to be dwarfed by “hot,” automatic, subconscious cognition, because the latter can be accomplished faster and with less investment than the former.
On the very first page of the preface, the book lays out its object: “unconscious political thinking and the subterranean forces that determine how citizens evaluate political leaders, groups, and issues” (p. xiii). As in Predisposed, the goal is to come up with an abstract, general model of the political subject based on psychological findings. Much of this research is very strong, and sociologists ought to pay close attention to the overwhelming evidence that much human action is habitual or automatic, and furthermore that even when actors seem to be acting in conscious, deliberative ways, they may well be engaging in rationalization instead of the calm consideration of evidence and argument. (It’s a bit ironic, though, that a book whose point is that people are rarely able to assess evidence and argument even when they want to is written as a dispassionate presentation of evidence and argument!)
The authors spend a great deal of the book building toward a general model of the political actor, a model they summarize at the end of the book in what they call their JQP (John Q. Public) model. The model incorporates affective, experiential, and learning components and is designed to be implemented in a computer simulation that can then be used to examine how hypothetical JQPs will react to differences in the environment such as varying information, preferences, and emotional responses. JQP is a truly remarkable achievement. It reflects a triumph of this style of research: a fully functioning model of the theorized actor that can be assessed against competing visions of political actors. And in the final chapter, the book pits that actor against other models promoted in political science and shows a positive result.
But the triumph of JQP also shows the weakness of its paradigm. Like other literature in the dual-process and motivated-learning paradigms, virtually no attention is paid to the key question of where “automatic” reasoning comes from. At times, the book veers toward an untenable claim that automatic reasoning is more authentic than deliberative, though it never makes that claim explicit. How, for example, does it come to pass that liberal-leaning subjects “automatically” react relatively positively to pro-labor and pro-environment cues, even though the coalition that brings these two interests together is neither ideologically necessary nor historically persistent? This is particularly concerning when so many of the studies are made up of college students or convenience samples.
The JQP model contains no opportunities for subjects to be challenged, either intellectually or emotionally, on their beliefs; no institutional context making certain actions easier or harder; no cultural repertoire of belief or action; no dynamics in the environment for action facilitating or foreclosing particular directions. That means it is ultimately limited in similar ways to the model from Predisposed: it identifies individual-level tendencies when so much of the action in political outcomes happens at the supra-individual level.
The same can certainly not be said of Andrew Gelman’s well-known Red State, Blue State, Rich State, Poor State. Gelman uses representative data about voting behavior to model state-by-state predictors of elections. Thus while the data themselves are individual-level, the argument is in the aggregate and the actors are set within appropriate contexts. Gelman uses David Brooks’s silly “One Nation, Slightly Divisible” (Brooks 2001) as a foil; Brooks claimed that Red State-Blue State differences were essentially about taste and lifestyle, not income or social class. Gelman shows convincingly that old-fashioned class, not consumption-based culture, is the primary predictor of individual and state vote choices. More specifically, the interaction between place and class forms the backbone of Gelman’s argument. Across many different states and regions within states, individual-level income continues to be among the biggest predictors of vote choice.
The bottom line for Gelman, then, is that voting remains mostly polarized by class, with lower-income voters breaking more for Democrats and higher-income ones for Republicans. And this pattern holds true within “red” states like Kansas and “blue” states like Connecticut, as well as swing states like Ohio. Responding to Thomas Frank’s popular book What’s the Matter with Kansas, Gelman asked in a blog post, “what’s the matter with Connecticut?” In fact, it is in the wealthy state of Connecticut that the relationship between income and vote choice is muted relative to national trends.
Gelman’s book is less purely reductionist than are the prior two; the historical and contextual narratives in the book retain more of the complexity and even contingency that produced the observed outcomes. But the intellectual style is similar, in that the book identifies a single dimension of political behavior and isolates that dimension from its roots and its institutional setting.
The next two books mitigate that problem by paying specific attention to the institutional processes that aggregate individual decisions into collective outcomes. In Changing Minds or Changing Channels?, Kevin Arcenaux and Martin Johnson take up the question of the influence partisan media outlets have on individuals’ opinions. The research on media influence has taken place largely in laboratories, demonstrating that exposure to particular forms of news—such as partisan news like Fox News or MSNBC—can affect individuals’ beliefs and opinions, albeit more weakly than one might think. Other research has shown that the changing media regime affects people’s political knowledge. What this book does is demonstrate that the media regime—which channels and other media streams are available and how—moderates the effects of partisan news on opinion. Essentially, the ability for people to select into media exposure blunts the effects of those media. That means things seem more polarized than they are because selection effects add to exposure effects.
Thus, an institutional configuration (the structure of media) changes the individual-level effects of partisan media. That configuration is part of what I referred to as the “thick haze of mediators” that make individual-level differences relevant. The book presents an “active audience theory” to account for these aggregate effects. The theory is effective for that use, but it doesn’t account for the fact that interpretation of information and ideas is, itself, an active process. The settings where people consume and discuss media, and the backgrounds they bring to media, affect what those media’s messages mean to them.
A similar approach animates The Logic of Connective Action, a careful study of the role of social media in facilitating and changing (though, interestingly, not hampering) contentious politics. W. Lance Bennett and Alexandra Segerberg present a typology of the kinds of movements that can be facilitated through social media. It comes as no real surprise that social-media-based movements tend to prefer less organization, less hierarchy, and more self-expression. Different online-enabled movements are different from one another, though; some, like Occupy, tend to be extremely decentralized, while others are more hybrids and maintain a level of organizational control. That’s important because, while the former facilitate more self-expression (at least for relatively privileged movement members), the latter seem more effective for movement outcomes and longevity.
If anything, Bennet and Segerberg underplay individual-level effects such as selection into, and social capital within, online networks. If people choose which online movements to be part of based in part on individual styles and preferences, that cannot be attributed solely to the character of the network itself. But nevertheless, this book, like Changing Minds or Changing Channels?, pays real and welcome attention to the institutional climates that structure individual political action.
But institutional context alone is not sufficient to capture that thick haze. Each of the final three books I’ll consider uses a different approach to develop a more thorough assessment of the haze’s moderating process. In Changing Minds, If Not Hearts, James M. Glaser and Timothy J. Ryan begin with an individual-level model not so different from JQP. They accept the idea that much political reasoning is automatic and that the unconscious attitudes implicated in that reasoning may be impervious to change (although this is an important question that deserves further investigation). But they also demonstrate that, given specific cues, people’s automatic reasoning on questions of racial fairness can be tempered by information and deliberation. In other words: even if the political subject is essentially JQP, she may not act in that capacity if the context favors a more reasoned choice.
If Glaser and Ryan use psychological insights to transcend the simplified models, Claudia Strauss uses a model of culture that will be familiar to sociologists to conceptualize the nature and dynamics of public opinion. It may not really be fair to include Making Sense of Public Opinion here, as the author is an anthropologist. But the book’s logic speaks directly to the concerns of political science and does so in a way that both sociologists and political scientists ought to pay attention to. Strauss provides a detailed model for the way people form opinions out of the “opinion communities” they are part of. But her model is far from reductive. Rather, it assumes both access to numerous such communities and active work on the part of people sorting out the discourse in those communities to arrive at an opinion—formed only “when someone is supposed to state an opinion” (p. 111). It is an elegant, complex theory of individual-level opinion formation in the context of cultural and institutional moderators and should be required reading as an antidote to the individual acting subject models in the books at the beginning of this review.
Finally, where Strauss brings in culture, Ira Katznelson’s magisterial Fear Itself deploys history to account for the vast complexity of the origins of contemporary American politics. The first chapter alone makes this book worth reading. That chapter lays out the historical role of the New Deal, not just as an economic intervention in the face of the Depression but also as a fundamental shift in what people expect of politics and policy. The New Deal ushered in nothing less than a new vision of the state and a new framework for doing politics. Katznelson’s careful analysis of the historical dynamics that made that so shows why reductionism is unnecessary for making a causal claim.
Beyond the first chapter, the book details the ways Jim Crow, World War II, and labor politics coincided to produce strange coalitions as well as ideological constraints and opportunities. The power held by southern Democratic lawmakers meant that civil rights was off the agenda. But the international environment favored a demonstration that American democracy was effective and representative. Roosevelt capitalized on that dynamic, relying on southern Democrats to support the New Deal’s class-based mitigation in order to avoid having to confront racial justice. Both cause and effect in this narrative are complex. A combination of local and national culture, international political environment, and inside-the-beltway strategizing produces the New Deal’s successful implementation. Meanwhile, the cultural and institutional effects of the New Deal set up the ways people—JQP included—imagine the possibilities of, and their preferences for, state action.
Political science has made great headway with relatively abstract models of human cognition and behavior that reduce political processes to aggregations of individual thought and action. Political sociology can learn from that exercise and from its critics. When political science ignores the social, institutional, cultural, and historical mechanisms that moderate between such thought and action and their collective outcomes, it generates models that are likely wrong but still instructive. Sociologists ought to resist thinking of these models as accurate, but should consider them and their extensions as ways of simplifying and abstracting ideas of social and political behavior.