Explore the scholarly record.

Abstract

Abstract Voting is a cornerstone of democratic legitimacy, yet surveys often overestimate electoral participation. One explanation for this discrepancy is social desirability bias (SDB), whereby respondents overreport their intentions to vote to align with democratic norms. We use a list experiment to examine the extent to which citizens misreport their intention to abstain from voting. The study draws on a probability-based sample of Finnish citizens ( $$N = 655$$ ) and replicates the analysis in a convenience sample ( $$N = 1276$$ ) in the context of the 2024 European Parliament election. In the probability-based sample, 18.8% of respondents directly report an intention to abstain, compared with an experimental estimate of 16.0%, providing little evidence of aggregate-level social desirability bias. In the convenience sample, we find modest evidence that abstention intentions are underreported. Subgroup analyses, however, suggest divergent reporting patterns: women and older voters tend to underreport their intent to abstain, whereas men and younger voters tend to overreport it. As a result, aggregate estimates may conceal substantial reporting biases operating in opposite directions across groups. These findings highlight how apparently unbiased aggregate survey estimates of turnout intentions can mask important subgroup-level distortions and suggest that the perceived social desirability of electoral participation may vary across groups.