Election Defeat Reduces Social Trust Among Polarized Canadian Voters

A new analysis of Canadian Election Study data shows that voters whose preferred party loses a national election report lower generalized trust in other people. After controlling for age, gender, education and employment, the researchers found that electoral losers were about 3.8 percentage points less likely to say that most people can be trusted than supporters of the winning party. When the comparison is limited to the two largest parties, the gap expands to roughly six percentage points. The decline is strongest among voters who entered the election confident of victory and who hold strong negative feelings toward opposing parties, with a trust gap approaching ten percentage points.

Psypost

The study examined responses from more than 30,000 Canadians across six federal and eleven provincial elections between 2004 and 2024. Survey participants first reported their intended vote before the election and later answered a standard question measuring social trust. The authors also collected affective-polarization scores by comparing respondents’ warmth toward their own party with their coldness toward rivals. Statistical models focused on subsamples ranging from 17,570 to 27,813 individuals, allowing the team to isolate the effect of election outcomes from demographic factors.

Psypost

The association between losing and reduced trust was moderated by affective polarization. Among voters with low animosity toward opponents, election loss did not correspond to a meaningful trust decline. By contrast, highly polarized voters—those who strongly disliked rival parties—showed the largest gaps, especially in the 2019 and 2021 federal elections, which the authors identified as the most polarized periods in the two-decade span.

Psypost

These results suggest that the informational signal of a national election can reshape everyday perceptions of societal honesty, but primarily for individuals already entrenched in partisan hostility. If widespread, such trust erosion could undermine civic engagement, cooperation on collective problems, and the broader health of democratic societies.

Psypost

The authors caution that the observational survey design cannot prove causality; unmeasured factors such as economic conditions or candidate characteristics may also influence trust levels. The study lacked a neutral control group, as all citizens experience election outcomes simultaneously. Future research could employ experimental methods—such as behavioral games—or expand the analysis to cross-national samples to test whether the observed patterns hold beyond Canada and translate into actual discriminatory behavior.

Psypost