Article Summary
Introduction
Previous research has found that plurality electoral systems, where the party with the most votes emerges as the winner, are associated with higher levels of affective polarization compared to proportional representation electoral systems, where legislative power is assigned based on the relative number of votes for each party. However, these and other findings establishing a relationship between political institutions and affective polarization are only correlational. Political institutions are likely connected to other factors that can influence both partisan animosities and the institutional designs of political systems. For instance, prior research has shown that more affluent democracies are more likely to adopt proportional representation systems, which can also impact varying levels of affective polarization. To address concerns regarding these potential confounding factors, the authors have implemented a survey experimental study, posing the following question: How do electoral rules and the number of parties influence partisan animosities?
Analytical Approach
The incentivized experiment comprised two main parts. In the first part, respondents encountered a voting scenario where they were matched with other players participating simultaneously. They observed whether another participant belonged to the same team and the number of other teams in a hypothetical election, ranging from three to five. Each team received a randomly assigned unique position on a number line. Based on this randomized configuration, respondents were asked if they wanted to participate in the election and vote for their team. Participation increased the chances of their team winning but required a cost of one token (i.e., $0.25) to cast a vote. While the setup of this voting simulation was consistent for all respondents, the electoral formula determining the outcome of the hypothetical election varied among treatment groups. Respondents in the plurality and proportional representation (PR) treatment conditions could select only one team, whereas participants in the ranked-choice voting (RCV) condition submitted a ranked preference list.
| Treatment group | Payout formula |
|---|---|
| Plurality | The winning team in the election is the one with the most votes. Respondents receive a reward of 11 tokens minus the distance to the winning team. |
| Proportional representation (PR) | Respondents received tokens weighted by their distance from the other teams and the votes cast for each of the teams. |
| Ranked choice voting (RCV) | If no majority of votes emerged for a given team, the votes for the team with the fewest votes were reassigned to other teams based on the next preference on the ranked-choice list until a team reached a majority. Similar to the plurality condition, the team with the most votes wins the election, and respondents receive 11 tokens minus the distance to the winning team. |
After casting their votes or choosing not to participate, respondents learned about the election outcome. Subsequently, they were paired with another player, informed about their position on the number line, and whether their team won. This was followed by two rounds of an adjusted dictator game. In the first round, respondents proposed an allocation of five tokens between themselves and the other player, who could accept or refuse, resulting in no tokens for either player. In the second round, respondents received a proposed allocation and could choose to accept or refuse. The survey concluded by asking respondents about their perceived legitimacy of and satisfaction with the election.
Main Findings
Several of the study’s findings are consistent with previous research. Firstly, winning an election and increasing the number of parties contribute to higher respondent satisfaction. Additionally, a greater number of parties reduces intergroup animosity and the winner-loser gap in citizen attitudes. Proportional representation also diminishes the winner-loser gap, while plurality systems do not alleviate levels of affective polarization. However, some findings differ from those of previous studies. Ranked-choice systems decrease the winner-loser gap in perceived electoral fairness, whereas proportional representation reduces the gap in satisfaction with democracy. The relationship between the number of parties and the winner-loser gap depends on the electoral system. Only proportional representation and ranked-choice systems, coupled with five (but not fewer) political parties, reduce the winner-loser gap. Contrary to theoretical expectations, plurality systems exhibit a lower level of affective polarization than ranked-choice systems.
Implications
The findings of the study suggest that electoral institutions play a role in shaping political attitudes and partisan animosities, aligning with several prior studies that established a correlation between them. Nevertheless, the study also yields unexpected results, such as, when holding the number of parties constant, plurality electoral rules resulting in a lower level of affective polarization compared to ranked-choice systems. Only when combining ranked-choice voting and proportional representation with a high number of parties does it lead to a reduction in winner-loser gaps in perceived fairness and satisfaction with democracy. This underscores the significance of considering both the number of parties and electoral institutions in attitude formation. The authors draw the conclusion that diminishing the prevalence of plurality voting could potentially have mitigating effects on partisan polarization.
Questions left unanswered
While the authors highlight that their contribution enhances the internal validity of research on the effects of electoral institutions on public opinion, they also raise several unanswered questions. Firstly, it remains to be studied whether the results hold when respondents are confronted with real-world parties, as opposed to teams used in the experiment. Another potential avenue for future research is to explore the extent to which systems with coalition governments differ in their impact on partisan animosities compared to single-party governments.
Methods and Analysis
Was the study and its analyses pre-registered?: Yes
Did the study rely on proxy variables to measure polarization?: Yes
The authors used a variant of the dictator game instead of a feeling thermometer.
Were standard p-value thresholds used (p<.05 or 95% Confidence Intervals that don’t overlap zero)?: Yes
- Largest p-value presented as significant: 0.05
Were correlational results interpreted with causal language?: No
Limitations / Weaknesses
While the authors tested whether respondents understood each election rule in their survey instrument, the question remains whether voters are aware of the consequences of each of those rules for election outcomes. Since the experiment was conducted in the US, respondents might have been highly likely to comprehend the logic of plurality voting, as it is practiced in most US elections. However, despite understanding the rules themselves, they might be unaware of the effects of proportional representation or ranked-choice voting on real-world election outcomes. For example, while partisans might support proportional representation in principle, upon learning that it reduces their chances of gaining a majority in parliament and forming a government, they might become more affectively polarized toward the out-party.
Open Data & Analyses
Does the article make the replication data publicly available?: No
Does the article make the replication analysis scripts publicly available?: No
Article Citation
Fischer, S., Lee, A., & Lelkes, Y. (2021). Electoral Systems and Political Attitudes: Experimental Evidence (SSRN Scholarly Paper 3803603)
Bibtex
1
2
3
4
5
6
7
8
@online{fischer2021,
title = {Electoral Systems and Political Attitudes: Experimental Evidence},
author = {Fischer, Sean and Lee, Amber and Lelkes, Yphtach},
date = {2021-05-12},
doi = {10.2139/ssrn.3803603},
url = {https://papers.ssrn.com/abstract=3803603}
}