Turnout among individuals torn between two or more political parties is about 4.5 percentage points lower than among non-ambivalent citizens, according to a 46-country analysis, though the effect varies among countries with different electoral circumstances.
According to an article published April 1 in Party Politics, party ambivalence — evaluating two or more parties positively at the same time — leads would-be voters to stay away from the ballot box, particularly in countries that are polarized or have parliamentary systems, voluntary voting regimes and fewer parties jockeying for victory.
The findings represent the most comprehensive analysis yet on the link between ambivalence and voter turnout, according to author Semih Çakır, and could give future researchers new tools for understanding party ambivalence across national contexts.
"Party ambivalence is more prevalent than we think, especially in multiparty contexts," Çakır told The Academic Times. "And political strategists should be more creative about their strategies to persuade the ambivalent without demobilizing them."
Previous research has investigated the impact of a range of psychological factors on voting behavior, Çakır, a doctoral student at Université de Montréal, said. But most studies focusing on the relationship between ambivalence and turnout are limited to the U.S. context, meaning the findings might not apply cross-nationally.
"While single case studies could be insightful, the American context — with its bipartisan presidential system —is arguably a very particular one," Çakır wrote in his paper.
Additionally, he added, "as previous work has been mostly confined to the United States, a two-party system, we know little about how to measure ambivalence in multiparty systems."
To formulate measures of two-party and multiparty ambivalence, Çakır drew on the Integrated Module Dataset from the 2019 Comparative Study of Electoral Systems. The CSES compiles national-level survey responses related primarily to voting and turnout.
Çakır used data from 125 elections in 46 countries around the world, constructing two gauges of ambivalence.
The two-party ambivalence metric — measured from a non-ambivalent low of -5 to an ambivalent high of 10 — treats ambivalence between the two largest parties in a country as a function of the similarity and intensity of respondents' evaluations of them. If a respondent rates the two parties both highly and similarly on a 0-to-10 point likeability scale, for example, that respondent is relatively high in two-party ambivalence.
Çakır's multiparty ambivalence metric incorporates information on as many as five of a country's largest parties and ranges from -10 to 10, with lower scores indicating a single strong preference over other parties and higher scores indicating being more ambivalent toward all available parties.
Ambivalent voters have been shown to be less certain of their vote choice and more prone to persuasion during election campaigns, placing them high on the priority lists of campaign strategists.
"It has been shown in the U.S. context that ambivalence is linked to split-ticket voting … more vote switching, [and] late decision-making during campaigns," Çakır said. "So, ambivalent citizens are at the center of party strategies for persuasion. However, we don't really know about whether party strategies to attract the ambivalent lead to demobilization of this segment."
To gauge just how prevalent ambivalence is among voters around the world, Çakır plotted the distributions of two-party and multiparty ambivalence scores.
Drawing on over 147,025 individual observations available from the CSES dataset, he found that about 12% of citizens in the 46-country sample were ambivalent according to the two-party metric, compared to 40% who logged scores indicating that they strongly favored one of the two largest parties.
By the multiparty measure, approximately 9% of respondents were ambivalent toward all the largest parties in their countries, a percentage Çakır noted was "very conservative" considering that a full ambivalence score of 10 is possible only when a respondent gives a likeability score of 10 to all five parties in a context where five parties exist.
"Around 50% and 25% of the sample is ambivalent towards at least two parties and at least three parties, respectively," he said. "When we compare the effect sizes, it is clear that multiparty ambivalence indeed yields more sizeable effects despite being a more conservative measure, which also draws attention to the presence and importance of ambivalence in multiparty contexts."
Çakır used random intercept logistic models to calculate the impact of two-party and multiparty ambivalence across the sample of countries being studied.
Controlling for a variety of individual-level variables including respondent age, gender, education and party identification, he found that moving from -5 to 10 on the two-party ambivalence scale is linked with a 4.5 percentage-point decrease in the average predicted turnout. The effect was robust to macro level controls for countries with presidential systems, compulsory voting regulations and a high effective number of parties.
There is an 8 percentage-point difference in the average turnout likelihood between a hypothetical voter who totally favors one party and dislikes the others versus a voter who is highly ambivalent about all available parties, according to Çakır's analysis of multiparty ambivalence.
But these sample-wide generalizations mask important variations between individual countries, he discovered. While the average effect size of two-party ambivalence on turnout is 4.5 percentage points, for example, it actually differs substantially across countries.
"The effect of changing from -5 to 10 on the [two-party] ambivalence scale could be as high as 28 percentage points in Bulgaria and as low as 3 percentage points in Germany, or nil in other countries such as Brazil, France, Israel, Peru and Sweden," Çakır wrote.
Similarly, he found, multiparty ambivalence impacted turnout in about half of the sample, with an effect size that varied across countries. While the results ruled out the chance that ambivalence increases voter turnout, it raised questions about what was driving the observed variation.
Further analysis revealed that ambivalence has no statistically significant effect on turnout in contexts where party polarization is low, and that the negative relationship between the two variables becomes significant only where competing political groups are ideologically distinct from one another.
Likewise, Çakır found, ambivalence exerted its negative impact on turnout only in countries with parliamentary systems as opposed to presidential ones. That could be because presidential systems in countries such as the U.S. tend to emphasize individual political candidates during elections rather than the parties they represent.
Çakır also discovered that ambivalence has a negative impact on turnout only in countries with relatively few electoral parties with significant vote share. The effect size became insignificant moving from countries with five to countries with 10 such parties.
In countries with compulsory voting regimes, the effect of ambivalence on turnout was largely neutralized, a phenomenon Çakır said was likely because the policy increased the potential costs of not voting.
Taken together, Çakır said, the variables impacting the effect of ambivalence on voter turnout should prompt reflection on the part of political strategists hoping to drum up support for their favored parties.
While parties often deploy polarizing rhetoric to cut down on voters' ambivalence and secure their votes, he noted, the strategy can backfire by simply demobilizing them instead. That result could be counterproductive from a democratic standpoint.
"Considering that swing voters are important for political competition, democratic accountability and responsiveness, demobilizing the ambivalent would have negative consequences [for] democratic quality in the long run," Çakır said. "I think parties should be wary of this."
The article "Does party ambivalence decrease voter turnout? A global analysis," published April 1 in Party Politics, was authored by Semih Çakır, Université de Montréal.