A Coordination Game of Electoral Representation and the Quality of Democracy

• In Event: Democracy at Risk

Abstract

The advantages of electoral accountability depend to a considerable degree, on citizens sharing a common informational and deliberative environment among themselves and with public officials; yet, that condition has been significantly strained by changes in the media environment, together with ideological polarization. Our aims in this paper are to describe this challenge to representative democracy in a way that lays bare mechanisms by which polarization and contemporary technologies of communication contribute to the loss of effective accountability and then to address this challenge, from the standpoint of normative institutional design.
We proceed by developing a coordination game-based model of the relationship between representatives and voters. In this game, the voters choose the amount of attention and “benefit of a doubt” to grant to their representatives, while representatives choose how much to depart in their policy decisions from their constituents’ existing preferences. When representatives expect that they will have a meaningful opportunity to explain their votes or when the policy outcomes are likely to materialize relatively quickly, they will be more willing to depart from the voters’ prior beliefs -- in effect, to act as voters’ trustees. What emerges is a situation in which voters, in effect, give officials a benefit of the doubt without simply deferring to them. In contrast, when officials expect that voters will not give them the opportunity to meaningfully explain their choices in office or when the policy outcomes are not expected to become apparent sufficiently quickly to affect their re-election chances, representatives are more likely to pander to voters’ prior preferences -- in effect, acting as voters’ delegates.
As a coordination game, this interaction has multiple equilibria, including what we may call the “trustee equilibrium” with less pandering and a meaningful practice of “explaining the vote,” and the “delegate equilibrium” where those expectations are reversed. The equilibrium status of these outcomes comes from mutual reinforcement of voters’ responses and representatives’ choices -- if representatives act as “delegates,” there is no reason for voters to pay attention to anything other than whether representatives’ choices match their own prior beliefs; if, by contrast, representatives act as “trustees,” the value of attending to the explanations that they give for their votes is substantially greater.
Our interpretation of the challenge described above is that the technological changes affecting the ease and speed of communication in social networks as well as ideological polarization, which has been linked to these changes, make the selection of the “delegate” equilibrium more likely. The basic intuition is as follows. One immediate consequence of the near-instantaneous communication within social networks is that a politician’s action is communicated, commented upon, and ultimately “binned” within one’s social network long before she has an opportunity to engage with the constituents in any kind of sustained attempt at explanation or persuasion. Voters’ cognitive biases, such as correlation neglect (Levy and Razin 2015; Enke and Zimmermann 2019) and failures of negative introspection (Dickson, Hafer, and Landa 2008), lead to exaggeratedly strong and biased beliefs following interactions in self-selected social networks. The increase in the strength of voters’ prior beliefs about the appropriateness of policy choices decreases the attention that they may otherwise have given to the representative’s attempt at an explanation, which, in turn, leads the representative to prefer choices that are consistent with voters’ priors – that is, in effect, acting as a delegate.
The discussion has theoretical import in the context of the analysis of representative democracy’s justifiability as a regime type, as well as a practical dimension, in providing a way of thinking about how best to respond to the dysfunction that we are currently observing in many developed representative democracies. We argue that there are available institutional choices that can help nudge us toward an environment that is more conducive to realizing the potential of representative democracy. We explore two such choices in some detail. These choices are (1) particular forms of lottocratic institutions (such as deliberative polls), in which such institutions play a circumscribed informational role in improving citizen deliberation, and (2) forms of media regulation that may help curb the worst excesses of echo chambers. These choices do not exhaust the set of possible institutional elements within representative democracy that could help improve the informational environment of electoral accountability. However, they are of particular importance because they seek to cultivate a shared informational environment, which is important to representative democracy’s performance.

Authors

• Ryan Pevnick, New York University

• Dimitri Landa, NYU