On the eve of the 2020 US Presidential Election, Redfield & Wilton Strategies released a report which identified several key reasons why the election carried a degree of uncertainty despite Joe Biden’s consistently high leads in the polls. In particular, we highlighted that polls do not show how members of the public are voting (or not voting), but, rather, how members of the public say they will be voting.
Six weeks after the election, our research team re-contacted 2,000 US adults who had been polled in the build-up to 3 November, and found a startling 40% of respondents admitting, yes, they had told someone in person that they would vote for a candidate other than the one they actually ended up voting for. Male respondents (46%) were much more likely than female respondents (35%) to say they had done so.
A slightly greater proportion (43%) of those who said they voted for Donald Trump also said they told someone in person they were going to vote for another candidate, than those who said they voted for Joe Biden (40%). Just over a quarter (28%) of those who did not vote told someone in person that they were going to vote.
Moreover, a significant minority (29%) of the US public admitted saying somewhere online that they were going to vote for a candidate other than the one whom they actually ended up voting for. It is unclear whether this group were deliberately lying, or whether they simply changed their mind later in the campaign. Moreover, the specific online context was not stated (such as an anonymous online survey, a publicly visible comment on social media, or a private message to a friend or acquaintance). Ultimately, many Americans shared information about their voting intention online which was inaccurate.
A third (33%) of those who said they had voted for Donald Trump also admitted having posted somewhere online that they were going to vote for a different candidate, compared to 28% of respondents who said they had voted for Joe Biden. Among those who say they did not vote, less than a fifth (19%) say they had misrepresented their voting intention online. Those aged between 25 and 34 years old were particularly likely (42%) to say they were going to vote for someone other than the candidate they ultimately voted for.
High levels of misreported voting intention may be related to the fact that less than half (46%) of Americans said they would feel “very comfortable” sharing with friends and family how they voted in the Presidential Election. A further fifth (22%) would feel comfortable, but not very comfortable. Meanwhile, a total of 18% said they would be somewhere on the uncomfortable spectrum.
Younger people aged 18 to 24, were significantly less likely (32%) to feel “very comfortable” sharing their voting intention with friends and family than those aged 25 or older (43% to 52%). A similar proportion of Joe Biden supporters (50%) and Donald Trump voters (48%) would feel “very comfortable” disclosing their vote with family and friends.
In mid-December, a similar percentage of Americans felt “very comfortable” sharing how they voted compared to the proportion in seven polls during September and October who said they would feel “very comfortable” (41% to 45%) disclosing their voting intention.
In the weekend prior to the election, 24% of respondents agreed with a statement suggesting they would feel comfortable lying about their voting intention.
Altogether, by December 16 a clear majority (61%) of Americans said that most, if not all, of their family and closest friends knew how they voted in the Election. A further 28% said that some of their family and closest friends know how they voted, while only 11% said that none (or close to none) of their family and friends knew how they voted.
Respondents aged 18 to 24 were less likely (51%) to have disclosed their eventual vote (or the fact that they did not vote) with all their family and friends. An equal proportion of Donald Trump voters (65%) and Joe Biden supporters (66%) said their family and close friends were aware of how they voted.
Post-Election, Americans were more likely than in several polls conducted before the Election (53% to 57%) to have told most, if not all, of their family and closest friends how they voted. The US Public were also somewhat less likely in December, than prior to the Election (15% to 19%), to not have told any family or close friends how they voted.
Moving beyond friends and family, a strong majority (64%) agreed that they would feel comfortable sharing how they voted if they were asked, unprompted, by someone they just met. Only 15% disagreed that they would feel uncomfortable, and 18% neither agreed nor disagreed. Male respondents (68%) were somewhat more likely to feel comfortable informing a stranger of how they voted than female respondents (60%).
Prior to the Election, between 55-58% agreed that they would feel comfortable sharing how they intended to vote if asked unprompted, by someone they just met. Therefore, after the Election, moderately more Americans reported being willing to disclose how they voted than they were in the months leading up to November 3.
Indeed, while most (55%) of those polled on 16 December said they felt equally comfortable sharing how they voted with others than compared to how they felt in the months prior to the election, almost a quarter (23%) felt more comfortable, which underlines that many were wary to disclose their voting intention in the partisan environment of the campaign season (and unsure of how many people were indeed on their side). 11%, however, said they feel less comfortable sharing their vote than they did before the election.
A higher proportion of respondents who say they voted for Joe Biden (28%) felt more comfortable sharing how they voted than respondents who said they voted for Donald Trump (20%). At the same time, 14% of Donald Trump supporters felt less comfortable disclosing now that they voted for Trump as compared to in the period before the election.
Ultimately, a very significant proportion of the American public inaccurately stated their 2020 Presidential Election voting intention prior to November 3, both online and in person. Any historical account of the 2020 Presidential Election is incomplete with taking into account this significant degree of distrust and mendacity among the public.
To find out more information about this research contact our research team. Redfield & Wilton Strategies is a member of the British Polling Council and abides by its rules.