Poll: Americans are divided on what causes problems in US elections

Amid a flurry of changes to voting rules and regulations across the country, the American public is divided over whether the bigger problem with elections in America is that it is too hard to vote or that it is too easy to cheat, according to a new CNN Poll conducted by SSRS.

Partisans are on opposite sides of that question, and are divided over the effect several prominent rule changes would have on the fairness of elections in the United States.

And the debate over how elections should be carried out in this country is happening while about 3 in 10 Americans express doubts that President Joe Biden was elected legitimately, despite the lack of evidence of fraud or wrongdoing in last year’s presidential election.

The poll finds 46% of Americans feel the bigger problem in elections is that the rules are not strict enough to prevent illegal votes from being cast, while 45% say the bigger issue is that the rules make it too difficult for eligible citizens who want to vote to cast a ballot. That marks an increase in concern about it being too hard to vote compared with a CNN poll in March, when 39% felt that way.

In the interim, bills that would make the laws around voting more restrictive have advanced in closely contested states such as Arizona, Georgia, Florida, Michigan and Texas. At the same time, some less competitive states are making changes to ease access to voting, including in Delaware, Kentucky, Virginia and Washington.

Public opinion is sharply divided by party over whether election rules are too restrictive or too loose, with 76% of Democrats saying that the rules make it too difficult to vote and 87% of Republicans saying that the rules are not strict enough to prevent illegal votes from being cast. The shift since March comes across party and demographic lines, but appears largest among independents. Last month, 56% said the bigger problem was that the rules weren’t strict enough. Now 44% say the same, and the share saying the rules make it too hard to vote has climbed from 31% to 44%.

There is little consensus over whether several proposed changes would make US elections more fair, less fair or not have much effect. Two have clear majorities saying they would make elections more fair: Nearly two-thirds say elections would be more fair if the rules ensured that in-person early voting is available outside of normal business hours and on weekends (65%) and if voters were required to provide photo identification before they cast a ballot (64%). About half say elections would be more fair if two frequently discussed changes to voter registration practices were made: 51% say elections would be more fair if eligible voters were automatically registered to vote when they turn 18 (just 17% say that would make elections less fair), and 48% say they would be more fair if voters could register at their polling place on Election Day (23% say that would make elections less fair).

There’s a closer divide over the impact of allowing election officials to send absentee ballot applications to every registered voter in a state. More say allowing that would make elections more fair (42%) than say it would make them less fair (36%), with 20% saying it wouldn’t change the fairness of US elections.

And two rules are more widely seen as decreasing elections fairness than increasing it: Limiting access to ballot drop boxes only when polls are open (41% less fair, 34% more fair) and making it illegal to provide food or water to people waiting in line to vote (39% less fair, 13% more fair, 46% no difference).

Democrats and Republicans are on opposite sides of nearly all of these measures. The only rule change that a majority in both parties say would make elections more fair is ensuring that in-person early voting is available outside of business hours and on weekends (79% of Democrats, 52% of Republicans say so). On others, there is a wide gap.

While 71% of Democrats say sending absentee ballot applications to all registered voters would make elections more fair, 69% of Republicans say that would make them less fair. Most Republicans say limiting access to ballot drop-boxes to those hours when polls are open would make elections more fair (53%), most Democrats say it would make them less fair (63%). Democrats divide over whether photo ID requirements would increase (40%) or decrease (35%) the fairness of elections, but Republicans solidly see them as a measure to increase fairness (90%). And while 58% of Democrats see bans on providing food and water to voters waiting in line as decreasing the fairness of elections (58%), most Republicans say it would make no difference at all (65%).

Although there is no evidence of fraud or wrongdoing in the 2020 election, many still say that Biden did not legitimately win enough votes to become president. The poll suggests the share of Americans who believe that falsehood has held roughly steady since just before he took office in January at 30%. Those doubts are concentrated among Republicans, 70% of whom say they do not think Biden won enough votes to be president. But the share of Republicans who falsely say there is solid evidence that Biden did not win has dropped from 58% in January to 50% now.

The methodology and weighting for this poll have been modified compared with CNN polls conducted before 2021. Interviews conducted on cell phones made up 75% of the total, up from 65% in prior surveys. Dialing extended over six days rather than four days, allowing for more effort to be made to contact those who are not easily reachable. Demographic weighting was adjusted to account for more discrete education categories broken out by race, and a geographic weight was applied to ensure representative distribution by population density. In addition, results were weighted for partisan identification and lean among independents, with targets computed using an average of the current poll plus four recent CNN polls.

The new CNN Poll was conducted by SSRS April 21 through 26 among a random national sample of 1,004 adults reached on landlines or cellphones by a live interviewer. Results for the full sample have a margin of sampling error of plus or minus 3.6 percentage points.