Chris Achen and Larry Bartels have written a brilliant book. It is also a distinctly pessimistic assessment of the quality of American democracy. They are quite successful in showing a number of major deficiencies our common sense idea of democracy, what they call “the Folk Theory.” The Folk Theory is not an empirical model, it is a normative ideal. It may be on the naive side, but is it nevertheless deeply embedded in both popular conceptions of democracy and in the fabric of a great deal of political science research. In many parts of their book, however, it appears that Achen and Bartels attack the Folk Theory as though it were being offered as an empirical model of how the system actually works. But we have known for generations that the American voter falls short of the Folk Theory’s idealized democratic citizen. While Achen and Bartels provide interesting new data, their basic message is not news.
I doubt that Achen and Bartels will be successful in getting political scientists to abandon the Folk Theory. The reason is not the weakness of their data analyses, but rather that this normative model drives much of our research, and it is the model, or how far we are from it, that makes much of our research interesting. The important question is whether there is enough evidence in the collected works on U.S. politics to still have some faith in the underlying ideas of popular rule. I think the answer is a tentative yes.
A sizable fraction of American politics research is motivated in some fashion by our efforts to assess the relationships between citizens and their government. We can cite dozens of studies that find evidence of responsiveness between elected official and voters, or between public preferences and governmental policies, or between changes in things like the economy or war deaths and public assessment of those in power (Canes-Wrone Reference Canes-Wrone2015). None of these gives full-blown support to the Folk Theory, but many provide evidence that government is frequently responsive to its citizens. This is the case even though it is also true that sometimes government pays more attention to interest groups, or to the wealthy, or to ideologues or even campaign contributors. The democratic glass probably is half-empty, but there is still something there that makes it normatively worthwhile.
One of the saving gimmicks for people who harken back to the Folk Theory is the idea that together the voters are a lot smarter than they are as individuals. Achen and Bartels discount quite a bit the idea of “the miracle of aggregation” (Converse Reference Converse, Ferejohn and Kuklinski1990). A half-full approach, however, does not demand that aggregated opinion be statistically on the mark, but only that in response to government and elected officials, aggregate opinion behaves in reasonable ways. And aggregated opinion certainly shows a great deal more order and rationality than we would expect from what we learn from studies of individual citizens (Erikson Reference Erikson2007).
This brings me to V. O. Key, Jr.’s definition of public opinion, which I think is useful for reflecting back on the types of evidence that Achen and Bartels offer in Democracy for Realists. Key saw it as “opinions held by private persons which governments find it prudent to heed” (Key Reference Key1961). He realized that only some of the electorate is well-informed and active, those he termed “the attentive electorate” (1961). This clearly presents an upper class bias to what officials “find it prudent to heed.” But officials nevertheless are smart enough to know that elections are not random walks. In the states and districts, we see a striking consistency between the liberalism-conservatism of public preferences and both the voting records of elected official and the policies they produce. There is a lot to support one of the important corollaries of the Folk Theory, which is that elected officials pay attention to public opinion.
Bartels (Reference Bartels2008, Reference Bartels2016) has elaborated interesting counter examples such as the minimum wage or the estate tax where the combination of institutional status-quo bias and strong partisan preferences in Congress hold sway against public opinion. But it is also the case that these issues are not of burning concern to majorities of voters. When these, or other issues, become noticeably salient to large sectors of the electorate, elected official pay attention (Bartels Reference Bartels1991).
Reading Democracy for Realists I frequently thought there are interpretations to the authors’ findings that cast the mass public in a less damning light. I will mention just two examples, both dealing with Achen and Bartels’ arguments against the retrospective voting version of the Folk Theory. One is their focus on “myopia,” whereby the evidence shows that voters rely on economic change in the last year running up to the election rather than averaging growth over the term of the incumbent administration. The other is their demonstration that there was no policy mandate in the formation of the New Deal coalition which governed from the Great Depression well into the post-World War II era.
Averaging economic performance over the full course of the administration’s record is one way to evaluate and decide whether to reward or punish the administration. However, one can also argue that considering trends in economic change is rational. Thus, an improving economy would be judged better than one that is stagnant or even declining, even from a positive performance earlier. That is, what the authors see as irrational myopia might also be seen as a sensible way to achieve a better economic future. Similarly, Achen and Bartels show that across countries that incumbent administrations, whether left or right, were punished for the Depression, and that the out-of-power parties who happened to be on the other side of the ideological spectrum were elected and (as the economies improved) reelected. This probably does show that mass publics do not have “ideal points” of the levels of government control of the economy, but the pattern is consistent with parties being punished when things go bad and being rewarded when things are improving. People want a healthy growing economy, and whether this is accomplished by the left or the right is not of much concern to the average voter. Even this policy-blind but performance-sensitive simple rule is enough to make elected officials heed public concerns about economic performance.
Taking up Achen and Bartels’ call to drop the Folk Theory and explore the promise of group-based identities may yield significant empirical understandings of how people behave in politics, including important things not predicted by the Folk Theory. That research should be done. But without an underlying concern for the quality of democracy—the Folk Theory in some guise—I fear we would become social-psychologists who happen to look at political behavior. That is, if the normative stake in what our research says about the performance of democratic governance is removed, the enterprise political research just becomes a whole lot less interesting.
