In 2016, American citizens voted against the wisdom of policy experts and elites. A historic election campaign came to a close with a result that was feared by many and predicted by few. The economic and political consequences of the future policies enacted by the Trump administration still remain uncertain—much depends on the degree to which campaign promises (and threats) will be enacted. However, President-elect Trump’s campaign proposals for trade, foreign policy, climate change, and energy policy (to name a few) alarmed experts and politicians alike, and a vocal elite from both sides of the aisle warned repeatedly of the potentially long-lasting deleterious effects of such policies on the lives of ordinary citizens. These concerns were exacerbated by perceived shortcomings on requisite leadership qualifications normally associated with a president. Indeed, Trump’s election seems a mismatch with what should have been the policy preferences of a sizeable part of the American electorate who voted for him.
What can explain this apparent disconnect between the policy preferences and voting decisions of ordinary citizens? In Democracy for Realists: Why Elections Do Not Produce Responsive Government, Christopher Achen and Larry Bartels offer a new way of thinking about democratic processes by refuting a variety of tired assumptions, including the notion that outcomes of democratic elections reflect the policy preferences of a well-informed electorate. More generally, Achen and Bartels illustrate that the ideal of popular sovereignty, embodied in the “folk theory of democracy” (Dahl Reference Dahl1956, Downs Reference Downs1957), falsely relies on unrealistic expectations of the ordinary citizen. These expectations include the assumption that voters have distinct policy preferences, are aware of these preferences, are well-informed about a range of policy alternatives they face in elections, are able to choose the candidate that most closely resembles their own preferences, and then cast their ballot for their preferred candidate. Further, the authors challenge the folk democratic expectation that political parties respond to these preferences, and cater to the demands of the electorate by tailoring policy positions accordingly.
An alternative explanation of the democratic process, “retrospective voting” (Key Reference Key1966, Fiorina Reference Fiorina1981, Kramer Reference Kramer1971), does not, in Achen and Bartel’s view, fare any better in terms of suggesting democratic accountability. In Chapter 4–7, the authors analyze several historical cases of restrospective voting and conclude that voters do not carefully assess the performance of an incumbent during the entire time in office and then base their voting decision in the next election on whether a party succeeded or failed in improving the voter’s personal well-being. Instead, the electorate often votes according to “blind” restrospection or a short-sighted evaluation of the economy. Therefore the “most important single factor in determining who wins—myopic retrospection—is, from the standpoint of democratic accountability, essentially arbitrary” (176).
What then, is a more appropriate way to think about the realities that determine party identification and voting behavior? In Chapters 8 and 9, the authors develop the group theory of democracy, where policy preferences do not play a significant role in determining party identification and voting decisions. Instead, Achen and Bartels emphasize social identity and group attachment as core motivations that lead a voter to identify herself with a major party. Voting behavior, according to Achen and Bartels, is thus not primarily determined by ideology, as the “folk theory of democracy” would suggest, but instead determined by social identity, which becomes more salient when certain political events heighten the individual’s attachment to a group. Policy preferences, if they exist at all, are merely a byproduct of partisanship, developed by the most-informed voters as a post hoc justification for their party identification (Chapter 10, 310).
But in Democracy for Realists Achen and Bartels do not only take on the voter side of the democratic process, they also theorize the implications of group-based electoral behavior for party responsiveness (Chapter 11). Since identity—not ideology—drives vote choice, parties do not have a clear political mandate and “issue congruence [between voters and parties], insofar as it exists, is mostly a byproduct of other connections, most of them lacking policy content” (301). In the same way, “the relationship between constituents’ opinions and their representative’s policy choices is correlation, not causation” (313). Thus, the heart of Achen and Bartels’ realist view of democratic representation is not issue congruence, but identity congruence between voter groups and the candidates they elect (313).
So how, according to Achen and Bartels, do elections actually work? In political campaigns, parties mostly remind voters of their partisan identities and mobilize them to go to the polls, aided by “formal communications by the groups and informal communication networks among group members” to help citizens understand how their identity group connects to the candidates and parties (311). Since in most elections, the balance between Democratic and Republican identifiers is close, the real difference-maker are “pure” Independent voters, who may base their voting decision on “the familiarity of an incumbent, the charisma of a fresh challenger, or a sense that it is ‘time for a change’” (312). These volatile assessments by independent voters swing elections one way or another, turning election outcomes into nothing more than “random choices among the available parties—musical chairs” (312).
The strength of this book is its development of the group theory of politics and the wealth of empirical evidence that Achen and Bartels marshal in support of this fresh alternative to the folk theory of democracy. In Chapter 9, they show convincingly how several social identity groups (including Catholics, white Southerners, and women) have strengthened or changed their party identification as a result of political events or the emergence of relationship-changing political candidates. For example, in the 1960 Presidential election, 83 percent of Catholic voters reported voting for John F. Kennedy (the first Catholic Presidential candidate since Al Smith in 1928)—a huge increase from the previous support of Catholics for the Democratic party (243). In Chapter 10, Bartels and Achen use the case of public perception of U.S. budget deficit changes during the Clinton administration to show that voters do not only use their own partisanship to construct what the party stands for, but also to construct “objective facts” tainted by partisanship bias (267).
The empirical analysis of the party responsiveness aspect of their realist view is only taken up in the concluding chapter of the book and is thus less well developed than the assessment of the group theory. And the hypothesis that “pure” Independents are the force that tips the scale one way or another in an election, based on random myopic retrospection or candidate assessment by these Independents, is simply suggested, not empirically analyzed. This is surprising, since the assertion that election outcomes merely reflect “musical chairs” is a strong statement that will cause considerable discomfort even for supporters of a group theory of democracy.
So how can a realist theory of democracy contribute to an explanation of the surprising voter behavior in the 2016 US presidential elections? It is at the very least an excellent starting point to assess the (sometimes surprising) voter groups that formed—often across party lines—in response to the nationalist appeals by the Republican candidate Donald Trump in the US election. Achen and Bartels repeatedly point to James Madison’s description of factions in Federalist Number 10 as a psychological foundation for the group theory of politics. The authors describe convincingly how Madison foresaw the tendency of people to form groups, to attach to groups easily and strongly, to construct “us” versus “them” identities, and the powerful role of emotions rather than reasoning in group activity (215). The emotional attachment to a social identity group and the subsequent attachment to a political candidate based on identity congruence could not only explain the electoral support for Donald Trump, but also the fierce opposition against him, as well as the attachment of social identity groups like women, minorities, and African-Americans to the Democratic candidate Hillary Clinton.
A careful assessment of the way social identities translated into political loyalties in the 2016 presidential election, however, requires some refinement of Achen’s and Bartels’ group theory of politics. More specifically, the process of identity construction needs more elaboration by the authors, theoretically and empirically. For example, what constitutes a social identity? If race and gender are bases for identity groups, how large or small can these groups be? The authors point out that social identities can be “ethnic, racial, occupational, religious or other sorts of groups” (299), but fail to define identity groups beyond this general description.
And while Achen and Bartels point to the importance of the strength of identity attachment in determining if a particular social identity becomes an important determinant for partisanship, they do not specify what exactly causes a voter to feel more strongly about one social identity than another. Their analysis of the effect of womens’ identity on partisan preferences with regard to abortion policy (258–264) falls short of their promised disentanglement of the process of choosing between conflicting social identities, because it does not specify general conditions under which one identity supercedes the other. Women’s gender identity in relation to abortion is a highly specific and personal issue that cannot easily be generalized.
An underdeveloped, yet very appealing part of their argument is that the strength of group attachment increases “when political events make a particular identity salient or threatened” (267), which happens in all of their detailed case studies of social identity group attachments in Chapter 9. The authors do not analyze the timing of or the process by which either threat to identity or identity congruence can increase an individual’s attachment to one of their social identity groups. But the 2016 Presidential election is an excellent example of how threatened social identities—for example women, intellectuals, immigrants, and African-Americans—adjusted or reinforced their partisanship based on solidarity or identification with a threatened group. Pride of identity congruence was also not only a factor in the partisanship change of Republican Catholics in the 1960 election for John F. Kennedy, but arguably also an important determinant of group attachment of African-Americans in the 2008 election to Barack Obama and women to Hillary Clinton in 2016.
Lastly, what accounts for the divergence of individuals from their social identity group’s behavior? An interesting example would be an analysis of the decision of a considerable number of Latinos to vote for Donald Trump in the 2016 election.
The importance of a clear definition of a social identity group becomes evident when Achen and Bartels draw their conclusions for possible democracy reform in the last chapter of the book. After dismissing plebiscitary suggestions such as initiatives, direct primaries, and term limits as counterproductive if the folk theory of democracy does not hold, they suggest that in order to create a more effective democracy, the “power discrepancy” between groups needs to be equalized through a greater degree of economic and social equality. This in turn would result in an increase of the real political power of a variety of currently underrepresented groups (326).
However, in their examples of power discrepancies between groups the authors compare well-institutionalized interest groups with social identity groups. Most experts on group politics would agree that interest groups are distinct from many other groups in their degree of organizational structure and focus on particular political issues. This illustrates that we need a clearer empirical understanding of social identity group politics and social identity group organization (as the authors correctly point out) and an analysis of whether social identity groups do or do not, in fact, voice policy preferences that they would like to see represented in government. Are there certain preferences that automatically come with certain social identities, other than the survival of the identity group under threat? If loose group identifications only serve to facilitate party identification and do not need to be represented beyond identity congruence in government (in other words, “I just want a Catholic in office, I don’t care what the f@#% he does when he is there”), then the authors’ suggestion to overturn Citizens United v. Federal Election Commission as part of an effective democratic reform (326–327) will not affect the representation of social identity groups versus other organized interests. As a side note, the 2016 presidential election has also illustrated that seemingly resource-disadvantaged social identity groups have taken great advantage of social media as an organizing tool, a tool that also showed the potential to increase group loyalties through constant exchange between group members. Whether these cheap resources could also serve to influence policy outcomes (if these groups wish to be represented) remains to be seen.
Nonetheless, even if the “realist view of democracy” leaves many open questions, the depth and scope of these questions show what an important contribution the book is to our understanding of the democratic process. The strength of the argument lies in the way the authors refute a fundamental assumption of the democratic process, popular sovereignty, by revealing its “folk” nature as one that would be easily dismissed as naïve by anyone outside of political science. Instead, they replace these assumptions with the notion of group identity that makes intuitive sense to observers of recent political events. Any contribution to the political science literature that accomplishes such a fundamental “reality check” can only be called major and transformative. At the same time, the book’s theory opens many potential avenues of further research. We have much work to do.
