The metaphor of the public behaving as if it were a thermostat—that is, responding to increasingly liberal policies by becoming more conservative, and vice versa—was first proposed by Christopher Wlezien in his important 1995 article in the American Journal of Political Science. That model, which turns the more traditional notion of representation (opinion causes policy) on its head, reaches its fullest explication in this excellent new book by Stuart N. Soroka and Wlezien.
To be sure, the metaphor, and the entire conception of the connection between the mass public and elected officials, is inextricably tied to time; there are no representational dyads to be found in these pages. Soroka and Wlezien examine the connections between the over-time movement of public opinion and public policy across several policy domains in three countries (the United States, Canada, and the United Kingdom). After an introductory chapter, they lay out their theory of the relationship between public opinion and public policy in Chapter 2. There, they clarify that the theory is not exceedingly demanding of what decades of research has shown to be a public with limited appetites and abilities to process political information. In Chapter 3, they add important comparative wrinkles to their theory, specifying two dimensions that should moderate the opinion–policy connection. The first is issue salience; the authors expect both more representation and more public responsiveness to policy on issues that are of high salience, and less of both on issues that are of low salience. (Importantly, salience can vary over time.) The second moderator is the nature of institutional arrangements in a political system. They argue that both representation and responsiveness are diminished in systems with federal (as opposed to unitary) organization and also in parliamentary (as opposed to presidential) systems.
Soroka and Wlezien lay out the specifics of their conceptualizations of both public opinion and public policy in Chapter 4. Public opinion, in their examination, is a rather blunt instrument, characterized by a Goldilocks-like sense of preferring either too much, too little, or about the right amount in varying policy areas, such as health, defense, and education; the domains vary by country. Although the time series in each country covary substantially, the authors resist the urge to go to the macro-level extreme and aggregate across all issues into a single “mood” reminiscent of James Stimson's influential book Public Opinion in America (1991). There is, they argue, enough unique variation within each policy domain to warrant separate analysis. Having measured public preferences for spending, the authors likewise conceptualize and measure public policy in a similarly broad manner, examining over-time levels of spending on various policy domains. In Chapters 5 and 6, the thermostatic model is tested and largely confirmed. The public, in all three countries, on almost all policy domains, reacts thermostatically to government spending. When the government spends more, the public comes to prefer less; when the government spends less, the public comes to prefer more. The predicted moderators, federalism and issue salience, turn out to have powerful effects.
In Chapter 7, Soroka and Wlezien document the more familiar representational connections, showing that when the public prefers more liberal policy, it gets it, and when it prefers less, it gets that, too. Strikingly, this is true across both issue and country contexts. At least for students of macro politics, however, this portion of the book will be familiar turf. The controversial question beneath the fully aggregated analysis is explored a bit in Chapter 8, where the authors examine whose opinions are represented. Consistent with other recent over-time analyses (but not with some cross-sectional works, like: Martin Gilens, “Inequality and Democratic Responsiveness,” Public Opinion Quarterly 65: 778–96; and Larry Bartels, Unequal Democracy: The Political Economy of the New Guilded Age, 2008), Soroka and Wlezien find that across many salient subdivisions of the public (like education), opinions move in parallel over time. So who is represented? In these formulations, seemingly to the authors' surprise the answer is more or less “everyone.”
The book's primary (and quite substantial) theoretical value lies in a unification of the literatures on representation and public responsiveness. The scholarly literature on the connections between public preferences and public policy has seen the causal arrow go in both directions. And yet these works, though cognizant of one another, have not been linked theoretically until this book—and this represents its most significant and novel theoretical achievement. Indeed, in the current volume, Soroka and Wlezien make it clear that the opinion–policy representational connection only makes theoretical sense if the policy–opinion feedback connection also exists. They write that “without such [public] responsiveness [to policy], policymakers would have little incentive to represent what the public wants in policy—without public responsiveness, expressed public preferences would contain little meaningful information. There not only would be a limited basis for holding politicians accountable; registered preferences would be of little use even to those politicians motivated to represent the public for other reasons” (p. 22).
This is an important synthesis. It serves as a reminder that scholars of representation must necessarily assume that feedback will take place. And it serves as a reminder to scholars of public opinion that they must expect and look for the effects of policy feedback. Both of these postures will require adjustments from scholars in the field.
The empirical scope of the book is impressive. Soroka and Wlezien commandeer all of the aforementioned data on opinion and spending dynamics, showing both their similarities and their contextual uniqueness in the best comparative tradition. The lack of data from other regions of the world produces intriguing possibilities for future scholars: Chapter 2 (the main theoretical chapter) contains interesting theoretical insights—about majoritarian systems, for example—that remain untested in the current book. As more data on democratic systems becomes available, there will be opportunities to subject the thermostatic model to increasing amounts of scrutiny. As the time series in the Latinobarometer and other sources grow, for example, scholars should focus on the challenges of collecting policy data of similar quality in the hopes of testing the mechanics of democracy further.
All metaphors, of course, break down at some point. The most useful metaphors only succumb after extensive probing; the less useful ones crumble after only the most trivial questioning. Each reader will come to his or her own conclusion about the usefulness of the metaphor of the thermostat. For an actual thermostat, it is easy to see why, during the winter, the thermostat demands more heat: It is cold outside. Then, yes, the furnace kicks on, provides some heat, and is followed by a reduced demand for more heat. But when the heat goes off, it will again get cold because the heat dissipates, and the cycle starts all over again. That is not exactly what happens in Soroka and Wlezien's metaphor. What, after all, is their parallel for the winter cold (or, equivalently, summer heat)?
Overall, this is an important scholarly work that will be essential reading for scholars of representation, of public opinion, and of empirical democratic theory. It is well written, methodologically quite accessible (the technical material is relegated to an appendix), and appropriate for a broad variety of graduate-level courses in the subjects just mentioned, and for specialized undergraduate courses in comparative politics.
