Robert Clay Photography / Jason Kolenda / The Atlantic

Last week, Wisconsin voters faced a dismaying choice: Exercise their right to vote or protect their health and safety. As the coronavirus spread, so many voters applied for absentee ballots before the April 7 primary that election officials couldn’t send them out in time. The legislature had refused to postpone the election; the courts turned down Governor Tony Evers’s attempt to change the rules by executive order. So voters who hadn’t received their absentee ballot had to go to the polls despite the pandemic; the plight of poll workers went unaddressed. Wisconsin’s case was described as “a narrow technical question about an absentee ballot,” according to the 5–4 majority opinion of the U.S. Supreme Court. It was not. It was a matter of life and death.

The outcome—voters in masks standing in long lines at the polls—was all the more tragic and senseless because states have the option of conducting elections entirely by mail. Five states already do: Colorado, Hawaii, Oregon, Utah, and Washington. Critics of voting by mail, most notably President Donald Trump, have portrayed the system as insecure and skewed. “Tremendous potential for voter fraud, and for whatever reason, doesn’t work out well for Republicans,” Trump recently wrote on Twitter. I am a political scientist who has studied Oregon’s pioneering system in great detail, and my research suggests that these complaints are baseless.

States with upcoming primaries this year should convert them into vote-by-mail elections. Even with Bernie Sanders’s withdrawal from the Democratic presidential race, many important state and local primary contests are yet to be decided. Every state should adopt the system before the November general election. Wisconsin’s experience should not be repeated.

In Oregon, where I live, vote by mail has been the sole method by which elections have been conducted since 2000. All registered voters are sent a ballot approximately three weeks before Election Day. If a voter no longer lives at the address on record, the ballot is not forwarded. After marking the ballot, the voter puts the ballot in a secrecy envelope, which contains no identifying information. The secrecy envelope is then placed into a second mailing envelope. Voters are required to sign the outside of this mailing envelope, which they then mail or take to one of the numerous drop-off sites in the area. (In a further convenience for voters, Oregon has recently moved to prepaid postage for these mailing envelopes.) When the ballots arrive at the local county-elections division, the outside signature is compared with the original registration signature on record at the office, to protect against fraud.

These safeguards are robust. As it happens, my two children—then college-age—have twice made an error on their mailing envelope (they signed each other’s mailing envelope or used an illegible scrawl instead of the original registration signature). In both cases, the county elections division caught these errors, contacted them to rectify the problem, and my children were still able to vote.

The introduction of vote by mail in Oregon has been gradual. In 1987, the legislature made vote by mail an option for local and special elections, and a majority of counties then adopted vote by mail for local elections because the administrative costs were lower than operating polling stations. In 1993, vote by mail was used for the first time in statewide ballot measures.

In 1995, the Republican-controlled state legislature passed a bill that would have adopted vote by mail for all types of elections. The impetus was the high level of absentee voting (22 percent) in the 1994 general election, which had delayed the certification of certain races for many weeks, including a close U.S. House seat. The reason for this delay in 1994 was that in so-called hybrid elections, officials needed to make sure that no one voted both at the polls and by absentee ballot. This was a rare and unlikely occurrence, but in a close election, this possibility had to be examined.

John Kitzhaber, Oregon’s Democratic governor, vetoed the vote-by-mail bill, arguing that it was too early for Oregon to adopt such a drastic electoral reform without further study. The opportunity for further experimentation soon arrived when Senator Bob Packwood resigned. The December 1995 primary and the January 1996 general election to fill his seat were conducted by mail. Turnout was high. The Oregon League of Women Voters led a successful petition drive to put a vote-by-mail measure on the 1998 general-election ballot. This ballot measure passed, winning 67 percent of the vote.

By then, leaders of both major parties had cooled to the idea, amid uncertainty about whether vote by mail helped one side over the other. Yet there is no evidence that vote by mail favors one party over the other. Vote by mail simply makes it easier to vote. It helps older and rural voters, who are more likely to be Republican, by reducing the impact of unanticipated difficulties, such as illness, bad weather, or long travel time to the polls. Young voters, who are more likely to be Democrats, often face inflexible work schedules or lack adequate transportation to the polls. For all types of voters, vote by mail simply reduces the likelihood that an unexpected crisis will prevent a voter from getting to the polls. As a personal anecdote, I remember a general election before the adoption of vote by mail that three members of my department could not vote in. One had a serious bike accident the day before, another discovered at 5 p.m. that his car battery was dead, and I was in the hospital having a baby. When three political scientists cannot vote on Election Day, the system might just need adjustment. Recent events reveal a major virtue of vote by mail: No one has to wait for hours in long lines during a pandemic!

While vote by mail is generally popular, not everyone in Oregon and elsewhere has supported it. Some critics have voiced concern about the possibility of undue influence on the voter, as there is no guarantee of a secret ballot. A spouse or partner could theoretically attempt to coerce a voter into voting a certain way. In Oregon, doing so risks a criminal penalty. Others have focused on the potential for voter fraud, when a blank ballot is stolen by another, either to negate the vote or used to vote twice. The scanning of the signature on the mailing envelope, as described above, should assuage such concerns. Oregon has sent out more than 100 million ballots since 2000, with just a dozen or so cases of proven fraud.

Finally, some observers have emphasized the problem of “ballot harvesting,” when nefarious persons go door to door in neighborhoods known to tilt heavily toward one side, offering to turn in ballots—and then simply discard them. Such a scheme by a Republican operative in North Carolina led to the overturning of a 2018 congressional election. All of these concerns should not be ignored, but they also apply to existing absentee voting in all states. Despite these concerns, no evidence has surfaced that vote by mail in Oregon has led to pernicious undue influence or voter fraud. According to the National Vote at Home Institute, other universal vote-by-mail states have also had only a few cases of voter fraud. The last absentee-fraud case in Washington was in 2010, before that state shifted to a 100 percent vote-by-mail system. Colorado had one absentee-voting prosecution in 2012 and one in 2017—one before and one after adopting the all-mail system. If anything, voters may be more alert to fraudulent schemes—and more careful about whom they give their ballot to—when everyone is voting the same way.

A change in voting systems nationwide could change the dynamics of campaigns; when Amy Klobuchar and Pete Buttigieg dropped out of the Democratic presidential campaign the day before Super Tuesday, some people had already cast absentee votes for then. Perhaps such voters would be wise to mail or drop off their ballot a few days later in presidential-primary years. Yet when many voters cast their ballot before Election Day, candidates and their campaigns must plan their activities accordingly. In Oregon and other universal vote-by-mail states, there are no last-minute media blitzes. Instead, appeals are typically targeted at reaching voters one or two weeks before Election Day.

Previous research on vote by mail has generally supported the argument that the vote-by-mail format increases turnout, although not for every type of election. My previous research on the overall turnout in 61 Oregon elections from 1980 to 2016 shows that vote by mail increases turnout modestly, particularly in special and presidential elections. Often involving one vacant seat or one ballot measure, many special elections are what political scientists call “low-stimulus elections.” In such a situation, even politically aware voters might find it less important to go to the polling place for a single issue, but the vote-by-mail format makes it easier for them to vote on even one ballot measure or race. Presidential elections are high-stimulus events, and they typically motivate first-time voters. Those inexperienced voters might have had difficulty locating their polling place in traditional elections, but have no such problems under vote by mail.

Vote by mail is the simplest way to avoid repeating the nightmare of Wisconsin’s primary. States should take care in adopting it, so that protections against voter fraud and undue influence are available. But legislators in Wisconsin and elsewhere do not have to invent these protections as they go along, as Oregon had to; they can copy what five states have already done. This can be done in a few months, and certainly in time for the November election. Given our current health crisis, we have no other choice.

The switch doesn’t have to be permanent. Once the pandemic has waned, state lawmakers might choose to return to their former system—but they might well find that the public prefers voting by mail.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.