The Shy Respondent and Propensity to Participate in Surveys: A Proof-of-Concept Study

Abstract

The Democratic presidential support among voters was overstated in 88% of national polls in 2016 and 93% in 2020. The "shy voter" phenomenon from British electoral politics was one explanation offered in the 2016 elections. Although a similar pattern occurred in 2020 election polls, the evidence does not support misrepresentation of voting intent by Trump supporters as the explanation. However, an alternative hypothesis of a self-selection bias against Trump voters in pre-election surveys has been proposed. Moreover, if these Trump voters were less likely to participate in surveys due to a psychosocial predisposition, then their absence might not be corrected by sample weighting based on demographics and party affiliation. This study explores whether there is a segment of the population with a personality or behavioral predisposition that makes them want to avoid polls ("shy respondents") and whether this affects their likelihood of survey participation and voting. As part of a national survey of motivators and barriers to survey participation, we had a proxy measure of survey shyness: "I prefer to stay out of sight and not be counted in government surveys." We compare this stated predisposition to both willingness to participate in surveys and likelihood of voting. We find this "shy respondent" measure is related to stated willingness to participate in future surveys. Although "shy respondents" were less likely to vote than others, a majority "always" vote in presidential elections. Collectively, "shy respondents" who are unlikely to participate in surveys represent about 10% of "likely voters" in presidential elections. This survey shyness is also related to political alienation measures, which may lead to underrepresentation of more populist leaning respondents in pre-election surveys. The relationship of survey shyness to demographics is generally slight so demographic weighting is unlikely to correct for underrepresentation of "shy respondents" in pre-election polls.