The COVID-19 pandemic led to a variety of changes in voting methods across the country in 2020, and the ways people could cast ballots varied widely from state to state.

Many states maintained traditional voting procedures, but some expanded early voting, and others automatically sent all registered voters a ballot that could be returned by mail.

New data from the Current Population Survey's November 2020 Voting and Registration Supplement released today show us in detail the ways voters cast their ballot at the national and state levels and allow us to compare voting methods and voter turnout across elections. Since these estimates are derived from survey responses, they differ from official reports from each state and estimates from other data sources.

In the 2020 election, 69% of voters nationwide cast their ballot nontraditionally — by mail and/or before Election Day. This is the highest rate of nontraditional voting for a presidential election (Figure 1) since questions regarding voting method have been included in the survey.

By comparison, about 40% of voters cast their ballots by mail and/or prior to Election Day in 2016.

Much of the surge in nontraditional voting was due to an increase in mail-in voting.

In 2020, 43% of voters cast ballots by mail and another 26% voted in person before Election Day. In 2016, 21% mailed in their ballots and 19% voted in person prior to Election Day.

The degree to which voters used nontraditional methods in the 2020 election varied across demographic groups by age, education, sex and race/ethnicity (Figure 2).

• A larger share of voters age 65 and over voted nontraditionally (77.7%) than those in younger age groups (69.7% or less).
• Those with a bachelor’s degree or higher also voted early or by mail at a higher rate (74.5%) than those with lower educational attainment (63.2% for high school graduates).
• Women (71.2%) used nontraditional voting methods at a higher rate than men (67.4%).
• Of the racial groups, the highest share of nontraditional voting was among Asian alone, non-Hispanic voters (82.4%). A larger share of voters age 65 and over voted nontraditionally (77.7%) than those in younger age groups (69.7% or less).
• Those with a bachelor’s degree or higher also voted early or by mail at a higher rate (74.5%) than those with lower educational attainment (63.2% for high school graduates).
• Women (71.2%) used nontraditional voting methods at a higher rate than men (67.4%).
• Of the racial groups, the highest share of nontraditional voting was among Asian alone, non-Hispanic voters (82.4%).

Educational attainment appears to have played a greater role in White alone, non-Hispanic voters’ decision to cast their ballot in a nontraditional way in 2020 than in 2016 (Figure 3), an outcome not evident for other race and Hispanic origin groupings.

Among race and Hispanic groups, the growth in the gap in the use of nontraditional voting based on education was statistically significant only among the White alone, non-Hispanic population. 

In 2016, among the White alone, non-Hispanic population, there was a 5.1 percentage-point gap in the use of nontraditional voting between those with less than a bachelor's degree (37.7%) and those with a bachelor's degree or more (42.8%). That gap grew to 9.4 percentage points in 2020: 63.4% of those with less than a bachelor's degree voted nontraditionally, compared with 72.8% of those with a bachelor's degree or higher.

Use of nontraditional voting also varied across states. Map 1 displays nontraditional voting rates in the 2020 election at the state level.

In states such as Washington and Oregon, which have sent all registered voters a mail-in ballot for multiple electoral cycles, over 95% of voters cast their ballot by a nontraditional method.

Other states continued to rely on traditional voting methods. For example, only 15.1% of voters in Alabama and 18.5% in Mississippi cast ballots nontraditionally (the two percentages did not differ statistically). 

Map 2 displays the change in estimated nontraditional voting rates by state from 2016 to 2020, reflecting the varying degree to which states adapted their voting procedures prior to the 2020 election.

Some states that relied heavily on nontraditional voting in 2016 continued to do so in 2020. For example, the nontraditional voting rate in Colorado was 90.4% in 2016, compared with 97.2% in 2020.

Other states that did not use nontraditional voting methods extensively in 2016 continued to make limited use of those methods in 2020. For example, in Oklahoma, 17.0% of voters cast ballots early or by mail in 2016, compared with 23.2% in 2020.

A third group of states saw large shifts in their use of nontraditional voting between 2016 and 2020. In New Jersey, for example, 7.9% of voters cast ballots early or by mail in 2016, compared with 92.3% in 2020, when all registered New Jersey voters received a ballot in the mail.

In addition to changes in method of voting, the 2020 election saw among the highest levels of turnout recorded in recent elections. Is the use of nontraditional voting methods associated with the change in voter turnout?

As shown in Figure 4, people living in the quartile of states that used nontraditional voting methods at the highest rate in 2020 also saw a larger change in turnout between the 2016 and 2020 elections than that of people living in the remaining states.

Among those living in states that used nontraditional voting methods at the highest rate in 2020, turnout increased by 7.7 percentage points compared with the 2016 election. For example, in Hawaii, 97.6% of voters cast ballots early or by mail in 2020 and turnout increased by an estimated 17.0 percentage points.

The 2020 election will long be remembered for having taken place during a pandemic. As the above findings indicate, the methods of voting used by those casting a ballot were significantly different from prior elections, though the magnitude and nature of the changes varied from state to state.

Zachary Scherer is a statistician in the Census Bureau’s Social, Economic, and Housing Statistics Division.