This award funds a doctoral dissertation that will use laboratory experiments to study the consequences of compulsory voting mechanisms on voting behavior. Voting may be voluntary or compulsory. For instance, voting may be compulsory (no abstention) or voluntary (abstention allowed) in small committees or in jury deliberations. In U.S. federal district courts, juror abstention from voting on a verdict in a criminal matter is not allowed while juror abstention is allowed in certain U.S. state courts; for example in civil cases where unanimity is not required. There are also differences in voting rules for larger-scale, national elections. Argentina, Australia and Belgium are among several nations where voting (more accurately, showing up to vote) in national elections is compelled by law.

The coPI plans a laboratory experiment to study the consequences of compulsory or voluntary voting mechanisms on voting behavior using the set-up of the Condorcet jury model. In that setting, individual voters all share a common interest, for instance, to acquit the innocent or to convict the guilty, but prior to any voting, each individual receives noisy private information regarding the true binary state of nature, e.g., whether the accused is innocent or guilty. The jury's (electorate's) choice is determined by majority rule. He studies two different voting mechanisms: (1) compulsory voting, where all voters are required to vote, and (2) voluntary voting, where each voter may first independently choose whether to abstain or to vote. In the latter case, we also consider whether voting is costly or not.

Game theory predicts that under either voting mechanism rational (i.e. strategic) voters will employ mixed strategies in equilibrium (under certain conditions). In the mixed strategy equilibrium of the compulsory voting mechanism, rational voters should sometimes vote strategically, i.e., against their private signal regarding the true state of nature, and otherwise vote sincerely, i.e., in accordance with their private signal. Such behavior can be rationalized by the recognition that other voters may have different signals and in small groups, there is some chance that an individual voter?s choice is pivotal to the outcome; these pivotality concerns can outweigh the value of the private signal, creating incentives to vote strategically. By contrast, in the mixed strategy equilibrium of the voluntary voting mechanism, rational voters resolve pivotality issues by playing a mixed strategy not with respect to the sincerity of their vote but with respect to their decision to vote or to abstain from voting; under the voluntary mechanism, all those who choose to vote are predicted to vote sincerely, according to their private signal. Voter participation rates under the voluntary mechanism are all endogenously determined. Participation decisions are further affected if there are costs to voting.

The laboratory study funded by this award investigates all of these behavioral predictions. The study uses a 2x2 design where the treatment variables are whether 1) voting is compulsory or voluntary and 2) (in the voluntary treatment) whether voting is costly or not. Experimental methods are well-suited to testing these theoretical predictions as the laboratory environment enables precise control over the voting mechanisms, the states of nature, private signals, group sizes and voting costs; environmental features that are difficult (if not impossible) to control in the field. The experimental findings from this first-ever study comparing compulsory and voluntary voting mechanisms will yield us a better understanding of voting behavior under these two mechanisms and may also provide some rationale as to why both mechanisms are used.

Project Report

for: Doctoral Dissertation Research: Compulsory versus Voluntary Voting Mechanisms: An Experimental Study #SES-1123914. The aim of this project was to use experiments to assess the extent of strategic voting and information aggregation under two different voting mechanisms that have been studied theoretically and are also found in nature: compulsory voting, where all voters are compelled to cast ballots and voluntary voting where abstention is allowed. We designed a laboratory experiment to study these different voting rules on voting behavior using the set-up of the Condorcet jury model. In that setting, individual voters all share a common interest, for instance, to acquit the innocent or to convict the guilty, but prior to any voting, each individual receives a noisy, private but informative signal regarding the true binary state of nature, e.g., whether the accused is innocent or guilty. The jury's (electorate’s) choice is determined by majority rule. In this setting we studied (1) compulsory voting, where all voters are required to vote, (2) voluntary voting, where each voter may first independently choose whether to abstain or to vote and (3) voluntary but costly voting, where voting is voluntary but voters have to pay a small cost to vote. The theoretical literature predicts that, under either voting mechanism, rational (i.e. strategic) voters will employ mixed strategies in equilibrium (under certain conditions). In the mixed strategy equilibrium of the compulsory voting mechanism, rational voters should sometimes vote strategically, i.e., against their private signal regarding the true state of nature, and otherwise vote sincerely, i.e., in accordance with their private signal. Such behavior can be rationalized by the recognition that other voters may have different signals and in small groups, there is some chance that an individual voter’s choice is pivotal to the outcome; these pivotality concerns can outweigh the value of the private signal, creating incentives to vote strategically. By contrast, in the mixed strategy equilibrium of the voluntary voting mechanisms, rational voters resolve pivotality issues by playing a mixed strategy not with respect to the sincerity of their vote but with respect to their decision to vote or to abstain from voting; under the voluntary mechanism, all those who choose to vote are predicted to vote sincerely, according to their private signal. Voter participation rates under the voluntary mechanism are all endogenously determined. Participation rates are further reduced if there are costs to voting as in our voluntary but costly voting treatments. The results of this NSF funded project appear in the peer-reviewed paper "Compulsory versus Voluntary Voting: An Experimental Study" that was published in Games and Economic Behavior, 84 (March 2014), pp. 111-131. NSF funding is acknowledged. The main findings are as follows. Consistent with rational choice predictions, subjects in our compulsory voting treatment engaged in some strategic voting. In particular, the group whose signal was more precise strategically voted against their signal a significant percentage of the time. Such behavior is consistent with the Nash equilibrium prediction for this model. By contrast, under the voluntary voting treatments, there is no evidence of strategic (or insincere) voting; instead, subjects adjusted their participation in voting according to the precision of the signal received and, in the case of costly voting, such adjustments also depended on the cost of voting, all in line with theoretical predictions. Finally, regarding the efficiency of these voting mechanisms in terms of achieving the correct outcome (e.g., convicting the guilty and acquitting the innocent), because subjects adapted their behavior to the institutional rules – voting insincerely when voting was compulsory and participating less frequently when their signal was more precise and/or the costs of voting were not too high, we found no significant differences in efficiency across the three treatments. We concluded that our experimental findings help us to account for the co-existence of both voluntary and compulsory voting mechanisms in nature.

Agency
National Science Foundation (NSF)
Institute
Division of Social and Economic Sciences (SES)
Type
Standard Grant (Standard)
Application #
1123914
Program Officer
Nancy A. Lutz
Project Start
Project End
Budget Start
2011-08-15
Budget End
2014-07-31
Support Year
Fiscal Year
2011
Total Cost
$11,300
Indirect Cost
Name
University of Pittsburgh
Department
Type
DUNS #
City
Pittsburgh
State
PA
Country
United States
Zip Code
15213