How we did this
Pew Research Center conducted this study to understand Americans’ views of several election proposals, as well as the potential impact of the coronavirus outbreak on the presidential election this November. For this analysis, we surveyed 4,917 U.S. adults in April 2020. Everyone who took part is a member of Pew Research Center’s American Trends Panel (ATP), an online survey panel that is recruited through national, random sampling of residential addresses. This way nearly all U.S. adults have a chance of selection. The survey is weighted to be representative of the U.S. adult population by gender, race, ethnicity, partisan affiliation, education and other categories. Read more about the ATP’s methodology.
Here are the questions used for the report, along with responses, and its methodology.
Over the past two months, the outbreak of the novel coronavirus has had a devastating impact on nearly all aspects of life in the United States. And now, most Americans expect it will disrupt the presidential election in November.
With just over six months until Election Day, two-thirds of Americans (67%) – including 80% of Democrats and Democratic-leaning independents and half of Republicans and Republican leaners – say it is very or somewhat likely that the coronavirus outbreak will significantly disrupt people’s ability to vote in the presidential election.
The national survey by Pew Research Center, conducted April 7 to 12 among 4,917 U.S. adults on the Center’s American Trends Panel, finds broad public support for giving voters the option of voting by mail – and less widespread but growing support for conducting all elections by mail.
Overall, 70% favor allowing any voter to vote by mail if they want to, including 44% who strongly support this policy. About half of the public (52%) favors conducting all elections by mail. The share supporting this proposal has increased 18 percentage points since 2018.
While most Americans expect the coronavirus to affect the presidential election, majorities are at least somewhat confident that it will be conducted fairly and accurately (59%) and that all citizens who want to will be able to vote (63%).
But as is the case with most attitudes about the conduct of elections and proposals on voting, partisans are sharply divided in these views. Large majorities of Republicans are confident that the election will be conducted fairly and accurately (75%) and that all citizens who want to vote will be able to do so (87%).
Democrats are considerably less confident: 46% are confident in the fairness and accuracy of the November election, and just 43% are confident all citizens will be able to vote if they want to.
As states begin weighing options for conducting elections amid the coronavirus outbreak, there is broad support among the public for both making voting by mail widely available and automatic voter registration.
Democrats overwhelmingly favor allowing any voter to vote by mail if they want to (87%), including nearly two-thirds who strongly favor this measure (63%). Republicans are split: 49% support allowing universal voting by mail, 50% oppose this.
Automatic voter registration is supported by 69% of Americans, including a 42% plurality who strongly support the election proposal. Democrats (84%) overwhelmingly favor automatically registering all eligible citizens to vote, compared with about half of Republicans (53%). Still, among the four election reforms asked about on the survey, the partisan gap is the smallest on automatic voter registration.
Of the four proposals asked about on the survey, there is only one with majority opposition: removing people from registration lists if they haven’t recently voted or confirmed their registration. About four-in-ten adults (42%) say they favor removing citizens from registration lists, while 56% oppose this. Republicans are much more favorable toward this policy than are Democrats (63% vs. 24%, respectively).
Public support for conducting all elections by mail has increased sharply since the fall of 2018 – from 34% then to 52% today – with most of the change coming among Democrats.
Today, 69% of Democrats strongly or somewhat favor conducting all elections by mail, up from just 40% two years ago. About a third of all Democrats and Democratic leaners (34%), including 44% of liberal Democrats, strongly favor this proposal.
At the same time, Republican support for the measure has ticked up only slightly. About a quarter (26%) favored conducting all elections by mail in 2018, compared with roughly a third (32%) who say they support the proposal today.
Currently, there is a 37 percentage point partisan gap in views of conducting all elections by mail. In 2018, partisan differences were far less pronounced (14 points).
As Americans look ahead to the presidential election this November, nearly eight-in-ten adults (79%) say it is at least somewhat important for the losing candidate to publicly acknowledge the winner of the election as legitimate, including a majority (54%) who say this is very important. This represents a slight increase since November 2016, when 74% said it was at least somewhat important for a losing candidate to publicly concede.
The increase has been driven by Republicans, who are 19 percentage points more likely to say that a losing candidate publicly acknowledging the winner as legitimate is somewhat or very important than they were four years ago. The proportion of Republicans who say this is very important has nearly doubled, from 32% to 60%.
Among Democrats, slightly fewer now say that it is somewhat or very important for a losing candidate to publicly concede than said this prior to the 2016 election. However, about three-quarters of Democrats (76%) say this is important, including half who say it is very important.
Most favor vote-by-mail access, automatic voter registration
The share of Americans who favor conducting all elections by mail has risen 18 percentage points from 34% in the fall of 2018 to 52% today. Democrats and Democratic-leaning independents account for most of this overall increase.
Just prior to the 2018 midterm elections, four-in-ten Democrats said they favored conducting all elections by mail. Today, nearly seven-in-ten Democrats (69%) say this, including 34% who strongly favor vote-by-mail elections.
Among Republicans and Republican leaners, rise in support is much more modest: About a third of Republicans (32%) now say all elections should be conducted by mail, up 6 points from two years ago.
Support for removing people from registration lists if they have not recently voted or confirmed their registration is slightly higher than it was in 2018 – 42% now say this, up from 37%. This increase is attributable to a 10 percentage point rise among Republicans over the past two years, from 53% in 2018 to 63% today. Just about a quarter of Democrats (24%) continue to say they favor this proposal.
Public support for another election proposal – automatic voter registration for all eligible citizens – has also risen slightly since 2018: 65% said they favored the policy two years ago, 69% do so today. Democrats remain significantly more likely to support this policy than Republicans (84% vs. 53%).
Support for conducting all elections by mail has risen among several groups – but not Republican men. Two years ago, similar shares of men and women in each party expressed support for vote-by-mail elections. Today, there is now a gender gap in views within the GOP.
In 2018, roughly a quarter of Republican men (24%) and GOP women (28%) said they favored conducting all elections by mail. There has been no change in support for the proposal among Republican men over the last two years, but support has risen by 13 percentage points among Republican women – from 28% then to 41% today.
Among both Democratic men and women, support for the proposal has risen by roughly 30 points over this period. Today, about seven-in-ten say they strongly or somewhat favor conducting all elections by mail today, up from about four-in-ten in 2018.
Broad support for vote-by-mail option, especially from Democratic-leaning groups
While the shares of Americans who support conducting all elections by mail has increased, there is even greater support for “allowing any voter to vote by mail if they want to.” Today, 70% of adults say they favor this, while just 30% oppose it.
There are some demographic differences in the magnitude of support for access to voting by mail. Black and Hispanic adults (79% in each group) are more supportive of the proposal than white adults (66%). Younger adults are also more supportive of this than older adults – though majorities in each age group favor allowing any voter to vote by mail if they want to.
Nearly three-quarters (74%) of those with a college degree or more education support expanding access to voting by mail, including 53% who say they strongly favor the proposal. A somewhat smaller share of those with no college degree (67%) also favor allowing any voter to cast their ballot by mail.
But these demographic divides are dwarfed by the partisan gap in attitudes about voting by mail. Nearly nine-in-ten Democrats and Democratic-leaning independents (87%) support allowing any voter to vote by mail if they want to, including 63% who strongly support the proposal.
Republicans and Republican leaners are divided: 49% say they favor allowing any voter to vote by mail if they want to, while 50% oppose the proposal. Conservative Republicans are far less supportive of a vote-by-mail option than their moderate and liberal counterparts (41% vs. 62%, respectively).
Most Republicans in states where many vote by mail favor expanding it
Adults who live in states where relatively large shares vote by mail are more supportive of expanding access to this method of voting than those who live in states with lower rates of voting by mail – a dynamic that is driven in large part by differences among Republicans.
For example, about two-thirds (68%) of Republicans who live in states where about a third or more of voters cast their ballots by mail say they favor allowing any voter to cast their ballot by mail if they want to. This compares with roughly four-in-ten of those who live in states with low rates of voting by mail.
There is less variation in views among Democrats. More than eight-in-ten Democrats and Democratic leaners – regardless of their state’s rate of 2018 mail voting – say they favor allowing any voter to vote by mail if they want to. However, Democrats in states with higher or more moderate rates of mail-in voting in 2018 are slightly more likely to say they strongly support allowing any voter to vote by mail than those in states where fewer vote by mail.
About six-in-ten Republicans favor removal of inactive voters from registration lists
There is broad support among Americans (69%) for automatically registering all eligible citizens to vote. By contrast, fewer (42%) support removing people from registration lists if they have not recently voted or confirmed their registration.
There are partisan and demographic differences in views of these two policies. Overall, 47% of white adults support removing people from registration lists if they have not recently voted or confirmed their registration; a slightly greater share (52%) opposes this policy. Smaller shares of Hispanic (37%) and black (26%) people support this approach to removing people from registration lists. In addition, younger adults are far less supportive of this proposal than older adults.
There is a large, 39 percentage point partisan gap on this question: 63% of Republicans support removing people who have not recently voted or confirmed their registration from registration lists, compared with just 24% of Democrats.
While there is more widespread support for automatically registering all eligible citizens to vote, differences in views by partisanship and race persist.
A large majority of Democrats (84%) favor automatically registering all eligible citizens to vote. Support is less widespread among Republicans: 53% favor this policy, while 47% oppose it.
Larger majorities of black and Hispanic adults (81% of each) than white adults (64%) are in favor of automatic voter registration.
Fewer than half of Democrats confident November election will be conducted fairly and accurately
Overall, about six-in-ten Americans (59%) say they are at least somewhat confident the November presidential election will be conducted fairly and accurately, while a similar share (63%) express confidence that all citizens who want to vote in the election will be able to.
But there are wide partisan gaps in these views – with Republicans much more confident than Democrats in the fairness and accessibility of the election. While three-quarters of Republicans and Republican leaners say they are at least somewhat confident the upcoming election will be conducted fairly and accurately, just 46% of Democrats and Democratic leaners say the same (with a majority saying they are not too or not at all confident in this). The partisan gap is even wider when it comes to ballot access. Nearly nine-in-ten Republicans express at least some confidence that “all citizens who want to vote in the election will be able to” – including 41% who say they are very confident about this. By comparison, just 43% of Democrats say they are at least somewhat confident all citizens who want to vote will be able to do so, while 56% say they are not too or not at all confident about this.
Among Democrats, confidence that the presidential election in November will be conducted fairly and accurately varies by ideology as well as by race and ethnicity. About half of conservative and moderate Democrats (52%) say they are either somewhat or very confident that the election will be conducted fairly and accurately, compared with 37% of liberal Democrats who say this. Similarly, while about half of conservative and moderate Democrats (53%) are somewhat or very confident that all citizens who want to vote will be able to do, 31% of liberals hold this view.
White and Hispanic Democrats express higher levels of confidence than black Democrats that the election will be conducted fairly and accurately: About half of white (48%) and Hispanic (50%) Democrats say this, compared with just 35% of black Democrats.
However, the gap between white and black Democrats is nearly reversed when it comes to the question of whether all citizens will be able to vote. About half of Hispanic (51%) and black (48%) Democrats are somewhat or very confident that they will, while just 38% of white Democrats say this.
Republicans are much more united in their levels of confidence in the election. Nearly identical proportions of conservative (74%) and liberal or moderate (76%) Republicans say they are somewhat or very confident the election will be conducted fairly and accurately. And nine-in-ten conservatives say they are confident that all citizens will be able to vote, while about eight-in-ten liberal or moderate Republicans (82%) say this.
Confidence in November presidential election fairness, accessibility linked to expectations of COVID-19 disruptions
Those who think the coronavirus outbreak is likely to significantly disrupt voting in November are less likely to express confidence the election will be conducted fairly and accurately or that all citizens who want to vote will be able to do so. To some extent, this reflects partisan patterns of opinions on these questions – but the pattern holds within both partisan groups as well.
Democrats and Democratic leaners who think it is not likely that the election will be disrupted are 12 percentage points more likely to say they are confident the election will be conducted fairly and accurately than Democrats who think it is likely the election will be disrupted. Among Republicans and Republican leaners, there is a 7-point gap on this question.
And while nearly six-in-ten Democrats (57%) who don’t expect the election to be disrupted say they are somewhat or very confident that all citizens who want to vote will be able to, just 40% of those who say it’s likely the election will be disrupted express confidence that all citizens who want to vote will be able to. Republicans who think it’s not likely the election will be disrupted are 12 points more likely to say they are confident that all citizens will be able to vote than those who are less certain the election will proceed with no significant disruptions.