Electoral Consequences of Network Sophistication and Uninformed Abstention

• In Event: Poster Session: Religion and Politics & Political Networks
  In Poster Session: Poster Session: Political Networks

Abstract

Reduced electoral participation is one of the most established empirical consequences of the lack of voter political information. On the other hand, less is known if this relationship is conditional on social context. Here, I argue that uninformed voters abstain from the election especially when the political sophistication level of their social network is high. In addition to the theoretical background and empirical evidence to this claim, the current paper constructs a simple agent-based simulation to explore aggregated consequences of conditional uninformed abstention. The results contribute to the literature revealing why and how the combination of social context, voter information, and conditional incentives for abstention changes the aggregated electoral outcome.

Previous studies make two conflicting claims regarding the logical origin of uninformed abstention. The first suggests that uninformed voters, those who have uncertain preferences, cannot motivate themselves enough to overcome the cost of voting participation (Matsusaka 1995). Uninformed abstention under this logic is unconditional on social context: it occurs regardless of the surrounding social environments. For the second, the theoretical development in strategic voting indicates that uninformed voters have a reason to actively abstain from the election (Feddersen and Pesendorfer 1996). They argue that if there is a shared true preference, less informed voters may be better off abstaining and delegate the electoral decision to more informed voters. The theoretical extension of this argument suggests that the optimal strategy of uninformed abstention is based on an individual’s relative, rather than absolute, lack of information (McMurray 2010). Voters with the low absolute level of information may not abstain if they are relatively well informed in the society (i.e., the society is uninformed on average).

This paper assesses the latter argument by examining if relative, rather than absolute, level of information in political network determines the participatory action of voters. The original theoretical argument assumes that all voters know the information distribution of the society, but such knowledge is rarely available in the real world. Instead, I argue that voters speculate the societal information distribution based on their network. The empirical analysis utilizing network question batteries offered in national election surveys including American National Election Study (ANES) provides the confirming evidence to this claim. I find that the highly sophisticated network strengthens the connection between the lack of information and abstention.

The additional analysis explores the consequences of the empirical findings. Even when voters potentially treat network sophistication as the proxy for the societal level of sophistication, the personal network does not necessarily reflect the society. Given this discrepancy, I construct the simple agent-based model with different spatial allocations of preferences and information to examine the consequences of uninformed abstention conditioned by network contexts. The results suggest that the special composition of the society has significant implication regarding how the logic of uninformed abstention consequently influences the electoral outcome.

The inquiry in the current paper sheds new lights on the studies of information and voting behavior by focusing on the role of the social network. It also gives new insights regarding how individual-level social network can influence the aggregated electoral outcome. By revealing the important connection between social network, uninformed abstention, and electoral outcome, this study bridges important studies that are previously discussed in separate contexts and methods. In future studies, I plan to conduct an experiment that directly examines the micro-level relationship between the personal network and uninformed voting. It will help to identify the exact mechanism that potentially explains when and how uninformed voters abstain.

Feddersen, T. J. & Pesendorfer, W. (1996), 'The Swing Voter's Curse,' The American Economic Review 86(3), 408-424.
Matsusaka, J. G. (1995), 'Explaining Voter Turnout Patterns: An Information Theory,' Public Choice 84(1), 91-117.
McMurray, J. C. (2010), 'Strategic Incentives in Elections with Heterogeneous Information,' Ph.D. thesis, Department of Economics, University of Rochester.

Author

• Gento Kato, University of California, Davis