Defending the Status Quo or Seeking Change? Electoral Outcomes, Affective Polarization, and Support for Referendums

Defending the Status Quo or Seeking Change? Electoral Outcomes, Affective Polarization, and Support for Referendums (DOI: https://doi.org/10.1017/S0007123425000365 ) was written by Bjarn Eck and Emilien Paulis in 2025. It was published in the British Journal of Political Science.

The authors study the differential behaviors of electoral winners and losers. "In established European democracies, we believe it is unlikely that citizens who lose the elections ... would abandon democracy... [T]hey might be more inclined to support additional instruments to influence decision making, such as referendums."

The authors further argue that affective polarization is a moderator for this causal effect.

The authors conducted a survey in 2021 and 2022 across these nations:

• Belgium (stratified across Flanders and Wallonia)

• Czechia

• Denmark

• Finland

• France

• Germany

• Greece

• Ireland

• Netherlands

• Poland

• Spain

• UK

Each nation's (or subnational group's) sample was forced to be representative by demographic quotas on age, gender, education, and region. Most have a sample size of at least 1200.

Data was also collected in Bulgaria and Italy, but due to difficulty in categorizing electoral winners and losers, this was excluded from analysis. Bulgaria was headed back to general elections after no coalition could be formed, and Italy was led by Draghi's grand coalition.

Support for referenda is measured by two 5-point Likert scales.

Affective polarization is measured by a 101-point thermometer question for each party. These measurements are then divided by 10 and transformed into a "spread-of-scores" metric as , where v_p is a party's vote share and the weighted average of like for an individual i is calculated as .

Controls include internal political efficacy, political interest, self-reported ideological position, gender, education, age, and occupation status.

A linear model is fit with nation (or subnational group) fixed effects.

The authors find that electoral losers are more likely to support referenda. Non-polarized electoral winners are also more likely to support referenda, but polarized winners are much less likely to support them.

CategoryRicottone