Will the Idea of Vote-by-Mail Survive COVID-19?

By Maria Carnovale

In November, when millions of Americans will be called to cast their vote in the presidential election, they might not be able to go to the polls. Quarantines, shelter-in-place orders, and the sheer fear of COVID-19 could depress turnout and ultimately affect the most important tenet of democracy: political representation. All federal, state, and local elections in the next few months will be similarly at risk. 

Could the timeworn technologies of pens, pencils, and stamps come to the rescue? Some states have already answered yes, and have instituted new systems to vote by mail. As other states consider this option, vote-by-mail systems could be catapulted onto the national stage.

But building a broader vote-by-mail system needs to be done right. Officials need to address ongoing cybersecurity concerns, and they need to make absentee voting widely accessible to people who have traditionally faced difficulties with mail-in ballots. Otherwise the system will be set up for failure, and this election will be remembered as the 2020 vote-by-mail fiasco.

Let’s consider Oregon’s experience as a model. The state introduced vote-by-mail in 1981 for local elections and expanded it over the years. Registered voters automatically receive a ballot in the mail, fill it out at their convenience, and then mail it back or drop it off at designated locations. Voters can also cast their ballot in person if they prefer. Oregonians report being satisfied with the system, and the low rates of ballot errors show that the voting process is easy to understand. Spurred by this accomplishment, other states have been expanding vote-by-mail. Colorado, Washington, Hawaii, and Utah also mail ballots to all registered voters. On May 8, California announced it would do the same in time for the November elections. In many other states, including the key swing state of Florida, any eligible voter can request a mail-in ballot.

Skepticism and low levels of trust, however, have dampened expansion of vote-by-mail. Some voters and government officials consider the system highly manipulatable: it is hard to confirm the identity of the individual who cast a ballot, the chain of custody is longer and more complex to monitor, and the voter’s privacy and independence is not guaranteed. A handful of scandals associated with absentee voting systems, including illegal harvesting of mail-in ballots in North Carolina and voting irregularities at nursing home facilities, heightened that distrust.

Missing from the current debate on vote-by-mail is the cybersecurity of the broader voting infrastructure, beyond vote counting and recording.

Nevertheless, the evidence on the security of vote-by-mail is encouraging. Proof of voter fraud related to voting by mail is limited. Many voters and government officials apparently feel that the paper trail of votes helps shield the system from potential cyberattacks. “You can’t hack paper,” Oregon’s former secretary of state once enthusiastically claimed.

But missing from the current debate on vote-by-mail is the cybersecurity of the broader voting infrastructure, beyond vote counting and recording. Vote-by-mail’s vulnerabilities lie upstream, in the intricate network of voters’ records that different agencies need to share. In states that use mail-in ballots, the Department of Motor Vehicles often provides digitized voter signatures, which are compared with signatures on ballot envelopes to prevent election fraud. Post offices require voter addresses. These sources of data are vulnerable.

The cybersecurity shortcomings of many state and local governments make them easy targets of malicious actors. Cyberattacks on voter information would undermine individual privacy, and hackers meddling with voter records would harm trust in the electoral system in a particularly sensitive presidential election. Some inexperienced states may not have enough time between now and November to protect the complex network of data required by vote-by-mail, unless they allocate enough resources to address security shortcomings.

It takes time to develop an effective cybersecurity system. States such as Oregon started vote-by-mail with local elections and then gradually expanded it. The system’s infrastructure (including its digital infrastructure) grew with it. The picture is very different for states such as Texas and New Hampshire without previous experience in voting by mail.

Elections are not only about security; they are also about accessibility. Changes in voting systems are rarely neutral to the composition of the electorate. When political preferences differ across groups, these interventions assume a political tinge that ends up overwhelming the debate. Vote-by-mail is no exception.

In normal circumstances—when there is not a highly infectious virus that forces social distancing measures—vote-by-mail has not been clearly shown to increase turnout in many states. What does change is the composition of the pool of voters who cast a ballot. Many studies find that people with nine-to-five jobs find it easier to vote by mail, disproportionately increasing the share of voters with higher income and better education levels. This is not always the case, though: a recent study in Colorado found that vote-by-mail increased turnout compared with in-person voting, especially among traditionally underrepresented groups.

Some states may not have enough time between now and November to protect the complex network of data required by vote-by-mail, unless they allocate enough resources to address security shortcomings.

Vote-by-mail’s impact on turnout in the context of a pandemic is also an open question. Voters could be afraid to venture outside of their homes and into crowded spaces—especially people who are most vulnerable to the virus, such as older adults, those with previous health conditions, and those without the financial security to bounce back from the economic impact of extended and debilitating sickness. In-person voting would underrepresent their political preferences.

Mail-in ballots could help. April’s Wisconsin primary election, which quickly gave expanded options to vote absentee by mail, showed voter turnout in line with historical trends despite the COVID-19 outbreak—with at least 71% of those voters voting by mail, a substantial increase compared with previous years.

However, time is short to address possible problems with inclusiveness. Mail-in voting requires taking numerous steps before election day, such as requesting a ballot and sending it back with enough time for delivery. The process can take multiple trips to a post office, which is a high toll for a single mom working two jobs or an older person in a wheelchair. Those with cognitive and visual impairments might need support in interpreting voting directions, instructions that would typically—at least in theory—be provided by poll workers. Many people with disabilities live alone and with health issues that increase their vulnerability to the virus, so creative ways to deliver ballots and provide support will be necessary.

For instance, Maryland developed an online tool to mark and download ballots for absentee voters. The system was successfully deployed in the state’s 2014 general election. Turnout increased among people who might need extra assistance marking the ballot but could not easily reach a polling station, such as visually impaired voters. The time to obtain a mail-in ballot dropped for those who used the online system. Scaling up experiments such as Maryland’s might be challenging given time constraints, but is probably the only way to ensure that elections in the time of COVID-19 do not disenfranchise underrepresented communities, such as voters with disabilities.

Internet-based solutions are not a panacea; they require appropriate infrastructure. Many urban voters can easily connect to the internet at high broadband speeds, but for many rural voters it’s a different story. People need connectivity in order to request and process absentee voting options online, but large parts of the country are lacking in broadband infrastructure.

Scaling up vote-by-mail experiments might be challenging given time constraints, but is probably the only way to ensure that elections in the time of COVID-19 do not disenfranchise underrepresented communities.

In this game, one size will not fit all. Although vote-by-mail will make voting easier and more secure in most urban areas, it might fail to include voters in underserved areas. There, other options are needed. Increasing polling locations, staffing them with poll workers who are at lower risk of infection, extending early voting periods, and making election day a holiday could be more cost-effective measures in areas that lack the appropriate infrastructure to deploy inclusive and representative absentee systems.

After years of proposals, the spotlight is on vote-by-mail. It can show its potential to make elections accessible in a secure, transparent, and inclusive manner. But for that to happen, a set of ancillary policies will need to ensure that the system does not disenfranchise large parts of the population and that the outcome is regarded as fair and secure. Just doing it will not work; it needs to be done right. Otherwise, nationwide vote-by-mail will be another fatality of COVID-19.