Presidential Polls - How Voters Voted

Nearly one in five voters said they changed their minds in the twenty-four hours leading up to their vote or in the voting booth on election day, November 3, 2020. At the same time, four in five had their minds made up already and two percent were not sure. Apparently, there was a "middle" in this election that thought long and hard about which candidate they were going to vote for, and then flip-flopped in the waning hours of one of the most important elections in the modern era.

There were significant differences among various demographics, for instance; voters aged 39-49 (30% yes/68% no) were more likely to have changed their mind at the last moment than were both the youngest voters aged 18-29 (22% yes/75% no) and the oldest voters aged 65+ (2% yes/98% no). Generationally, Millennials (born 1980-1995; 27% yes/70% no) were more likely to have changed their minds about who to vote for president than Baby Boomers (born 1946-1964; 3% yes/95% no), who altogether were more decisive with their votes.

Men (27% yes/72% no) were more likely to change their vote for president than women (9% yes/89% no), and weekly Walmart shoppers were more likely to change their minds in the twenty-four hours leading up to the election (27% yes/72% no) than were voters who never shop at Walmart (11% yes/87% no). We observed the same pattern with Amazon shoppers-weekly Amazon shoppers (32% yes/66% no) flip-flopped and altered their votes at the last minute more often than voters who never shop at Amazon (9% yes/88% no).

Our political polling indicated political affiliation was a significant factor, too. Independents (11% yes/87% no) changed their minds less easily than Democrats (21% yes/77% no) and Republicans (19% yes/79% no). Also, Hispanics (31% yes/66% no) altered their voting decision in higher numbers than Africans Americans (17% yes/81% no), whites (16% yes/83% no) and Asian voters (12% yes/84% no). Urban men (34% yes/64% no) were very likely to flip-flop, while suburban women (8% yes/90% no) were much more decisive about their votes for president in 2020.

The perception among many pundits was that Trump was such a polarizing figure that there were few voters who were on the fence or could change their minds very late in the game. However, our polling prior to the election showed a small yet important group of voters were still open to the possibility of voting for Trump even just days before the election. Our newest data confirm that a considerable number of voters changed their minds in the twenty-four hours leading up to the election. Did this benefit Biden? It's hard to say conclusively, but when we drill down and look at those who did change their minds, versus who they ultimately voted for, slightly more changed their minds for Biden (19%) than did for Trump (16%).


Zogby Analytics Poll Methodology
US Nov 3rd Election Voters
11/4/20 - 11/12/20

Zogby Analytics conducted an online political polling survey of 5180 adults who voted in the November 3rd Presidential Election.

Using internal and trusted interactive partner resources, thousands of adults were randomly invited to participate in this interactive survey. Each invitation is password coded and secure so that one respondent can only access the survey one time.

Using information based on census data, voter registration figures, CIA fact books and exit polls, we use complex weighting techniques to best represent the demographics of the population being surveyed. Weighted variables may include age, race, gender, region, party, education, and religion. The party breakdown for this survey is as follows: 37% Democrat, 35% Republican and 28% Independent/unaffiliated.

Based on a confidence interval of 95%, the margin of error for 5180 is +/- 1.4 percentage points. This means that all other things being equal, the identical survey repeated will have results within the margin of error 95 times out of 100.

Subsets of the data have a larger margin of error than the whole data set. As a rule we do not rely on the validity of very small subsets of the data especially sets smaller than 50-75 respondents. At that subset we can make estimations based on the data, but in these cases the data is more qualitative than quantitative.

Additional factors can create error, such as question wording and question order.


About Zogby Analytics:
Zogby Analytics is respected nationally and internationally for its opinion research capabilities. Since 1984, Zogby has empowered clients with powerful information and knowledge critical for making informed strategic decisions.

The firm conducts multi-phased opinion research engagements for banking and financial services institutions, insurance companies, hospitals and medical centers, retailers and developers, religious institutions, cultural organizations, colleges and universities, IT companies and Federal agencies. Zogby's dedication and commitment to excellence and accuracy are reflected in its state-of-the-art opinion research capabilities and objective analysis and consultation.