Spencer Goidel of Auburn University conducted a study that shows the potential of “choice overload” in elections and the impact it has on voter behavior. The study found that voters who participate in elections often vote for more candidates than those who do not.
The results were published by American Politics Research.
The impact of electoral structure, such as the number of candidates or the type of primary, on voter turnout is a widely debated topic. Goidel’s research focused on Louisiana’s unique “jungle primary” system, in which a nonpartisan, majority-voting primary is held on Election Day. This system can result in a large field of candidates, especially when they are candidates from the same political party.
The research examines the relationship between theories of consumer psychology and the concept of “choice overload,” which states that individuals may face overwhelming choices and make decisions they regret when presented with several comparable options. The aim of the study was to examine the validity of this concept in voting behavior, particularly in multi-candidate elections.
After the 2016 Republican presidential primary, I was surprised to learn that there were no capable candidates. This was a complete shock to me. Goidel was an assistant professor of political science at Auburn University.
A similar development occurred among the Democrats in 2020, which led me to think about the size of the field of candidates and their potential influence on voters.
Goidel took advantage of a natural experiment that resulted from the abandonment and later implementation of the “jungle primary” system in Louisiana. This system was created by the Supreme Court and designed as a tool to compare voter behavior in different electoral systems within the state.
Goidel explained that Louisiana’s primary was no exception in the jungle race, as the state’s voters had a significant number of tuition-paying candidates on Election Day, providing a good opportunity to estimate turnout in a typical primary.
Goidel used a difference-in-differences design to conduct an aggregate-level analysis, comparing enrollment rates in Louisiana and Mississippi, a state bordering Mississippi with a different electoral system, and developed a synthetic control model to construct control districts based on data from three other states.
Goidel used survey data from Louisiana respondents to conduct an individual-level analysis as part of the Cooperative Election Study. This part of the study examined whether voters in congressional districts with a high number of candidates were more likely to not vote in House elections. He combined these two approaches by analyzing the effects of candidate diversity on voter behavior from a population-level perspective and an individual-level perspective.
Both analyses remained consistent, supporting the theory that an increase in the number of candidates in an election means an increase in ballot dumping—even Goidel’s research found that in Louisiana, the reinstatement of jungle primaries in 1888 led to a 5.7 percent increase in ballot dumping per year, compared with the 8 percent increase when primaries were not conducted: When more candidates appeared on a given ballot, many voters also went to vote, even if they were in favor of the majority of them expressing support for the larger number of them who said they would.
These results are reinforced on another level: The analysis showed that in races for seats in the House of Representatives, voter abstention increased by one percent for each additional candidate. As the number of candidates in the House of Representatives increased, the probability of losing voters again increased sharply when the number of candidates exceeded two. In elections with twelve candidates, the probability of losing voters increased from 2.9 to 19.1 percent, indicating voter defection.
Goidel told PsyPost that the “first and most visible way to deal with political discontent is through institutional reforms” that “can increase voter diversity,” but democratizing the process “…is not the answer.” He further warned that democratizing the primary process “could in some ways depersonalize certain parts of the electorate and place the burden on those who support the party, possibly communicating their discontent to them.”
The study presents observations about the influence of candidate density on voting behavior, but its conclusions are flawed. The study examines only Louisiana and its distinctive jungle primary system, and while the results are generalizable to other states with similar electoral systems, such as California and Alaska, they may be incompatible with various other states.
Goidel noted that Louisiana saw a decline in voting in 2020, when mail-in voting was at its peak. He said that if one had internet access and time, voting from home would be a more manageable task. Still, he noted a slight increase in mail-in voting numbers due to the switch to electronic voting.
“A paper I wrote based on a polling experiment on how voters are influenced by the presence of large candidate fields. This paper is currently in preparation and I am interested in studying the effects of primaries for the top two or four candidates in other states.”
The study “More Candidates, Fewer Voters: How a Wide Choice Demobilizes the Electorate” was published on August 13, 2024.
