We’ll Need Vote-by-Mail in November. And It Could Be a Legal Nightmare.

States are suddenly scrambling to expand vote-by-mail for this year’s election. Exhibit A is Wisconsin, which for this week’s primary election has received more than 1 million absentee ballot requests, quadruple the state’s 2016 numbers, a previous high. The move to vote-by-mail, by and large, is a positive development, given that the Covid-19 threat will remain or return as a menace in November, and in-person voting could pose serious health risks for voters and poll workers alike.

But while absentee ballots can help keep people safe and expand voting access, they come with a drawback: a greater chance of litigation. Simply put, there are more things that can go wrong with vote-by-mail compared with in-precinct voting. And history shows that a major fight over an election’s outcome is more likely to come in the form of challenges to absentee ballots.

This doesn’t mean states shouldn’t make the shift to vote-by-mail. But in addition to hiring and training enough staff to be able to handle new procedures and ordering enough supplies and equipment, states — and Congress — must prepare for the increased risk of disputes over the results of the November election.

What can go so wrong with absentee voting? When you go to a neighborhood polling place, you meet face-to-face with a poll worker who signs you in and then lets you cast a ballot that immediately joins the pool of ballots to be counted after the polls close. True, there might be a problem when you interact with the poll worker: For example, a question about your signature compared with the one in the poll book. But those problems usually can be sorted out as part of the face-to-face interaction. Once you’ve satisfied the poll worker that you are qualified to vote and you get an ordinary ballot to cast, you no longer can become unqualified or your ballot left uncounted.

By contrast, voting by mail requires additional steps, all of which are susceptible to problems that do not exist when voting in person. First, the voter needs to get the absentee ballot in time to send it back by the required deadline. Administrative delay at the local election office, or a postal problem, can cause the voter not to receive the ballot with enough time for its return. This has turned out to be a serious problem in Wisconsin’s current primary, prompting litigation in federal court. The same could easily happen in November.

What’s more, although some states deem an absentee ballot eligible if it is postmarked by Election Day, other states require the absentee ballot to arrive at the local election office by then. In the latter states, there is a real risk that a voter who mails an absentee ballot will have it disqualified if it arrives too late. In 2018, more than a quarter — 27 percent — of absentee ballots that were rejected across the country were disqualified for this reason; in Florida, the rate was even higher, at least one-third.

Even if an absentee ballot arrives on time, it can be rejected for a variety of other reasons. Every absentee ballot must be submitted inside a special envelope, and on that envelope, the voter must provide essential information: name, address, signature and often additional identification (like a driver’s license number). Absentee ballots are routinely invalidated because of innocent clerical errors voters make in filling out the envelopes. Sometimes the voter doesn’t even do anything wrong, but an employee in the local election office innocently misreads what the voter has written.

In 2018, nationwide 91.8 percent of all absentee ballots sent in by voters to local election offices were counted, leaving 8.2 percent not counted — a significant disqualification rate. For the last presidential election, in 2016, states reported their numbers somewhat differently, but still the rates of disqualification are alarming. For example, Georgia reported counting 93.6 percent of absentee ballots returned, meaning that 6.4 percent were not counted. New York reported counting 90.7 percent of returned absentee ballots, with 9.3 percent uncounted.

These ballots largely are rejected not because the voter is unqualified, but because of errors in the process of the voter’s submission. Especially when it is possible to argue that the process defect was not the voter’s fault, there can be a sympathetic argument for counting the valid voter’s ballot despite the process problem. In fact, that sympathetic argument can be dressed up with constitutional principles and judicial precedents. And in a situation where the voter did not want to vote absentee in the first place, but did so solely because of health reasons caused by Covid-19, the argument for counting a valid voter’s absentee ballot despite an innocent process defect might seem even more compelling — or, at the very least, worth pursuing in court if it might make the difference between winning and losing the election. The result: lawsuits.

This heightened risk of litigation over absentee ballots is not just theoretical. Most of the most significant vote-counting disputes in recent decades involved fights over absentee votes. Consider just a few of many examples.

The 1984 election for Indiana’s 8th Congressional District saw an ugly fight in court, and in Congress itself, over absentee ballots that arguably deserved to be counted even though they had not been properly “witnessed” as required by state law. In the 2008 U.S. Senate race in Minnesota, between incumbent Norm Coleman and challenger Al Franken, absentee ballots were disqualified for various procedural reasons — some improperly so, others in accordance with state law. Still others were counted even though they were comparable to ballots that had been rejected, thereby generating an equal protection issue that took courts eight months to resolve (while keeping the crucial seat vacant in the middle of the financial crisis). Finally, absentee ballots were the focus of disputes in the 2018 midterms, both in Florida, where they came close to putting another U.S. Senate seat in doubt, and in North Carolina, where a House seat was left vacant for nine months because of an improper “ballot harvesting” operation.

Maybe you’re thinking: So what if we have a fight over counting absentee ballots in a few places this fall? We survived Bush v. Gore, after all. But a vote-counting dispute today could actually be much worse than that searing fight 20 years ago.

For one thing, not only is the body politic far more polarized; the Supreme Court has become engulfed in this hyperpolarization. Since Bush v. Gore, the court has repeatedly divided 5-4 in election cases: Citizens United (campaign finance), Shelby County (the Voting Rights Act) and Rucho (gerrymanders). Rightly or wrongly, Democrats view these 5-4 rulings, with Republican appointees in the majority, as having tilted the electoral playing field in favor of Republican candidates. Again, in the pending Wisconsin election, the court split 5-4 along partisan lines over the rules for counting absentee ballots—an ominous sign for what might occur this fall.

Moreover, since Senate Majority Leader Mitch McConnell kept a Supreme Court seat vacant after Justice Antonin Scalia’s death, depriving President Barack Obama of the opportunity to appoint a replacement, progressives have viewed the current 5-4 balance of power on the court as itself illegitimately stolen. If the current Supreme Court were to divide 5-4 in awarding the election to President Donald Trump, Democrats would see the presidency as well as the court as illegitimately stolen from them. Democrats reluctantly accepted Bush v. Gore as the product of a system that gave the court the power to do what it did. But if a “stolen” court were to steal the presidency, that combination could be too much to bear — a large proportion of the electorate might not see two of the three branches of government as legitimate.

It is also possible that a vote-counting dispute would not stop at the Supreme Court, but would go all the way to Congress. Without going into the details of why the legislative branch is so unprepared for this situation, a topic I’ve delved into previously, suffice it to say that the country could face a full-blown constitutional crisis if a voting dispute held up the results of the presidential election and Congress were unable to settle it before the winner is supposed to be inaugurated for the new term, beginning at noon on January 20, 2021.

So, what should be done? There’s no question that, for public health reasons, expanding vote-by-mail is a wise decision for states to be making right now. But states — especially battlegrounds in the presidential election — should clarify as soon as possible the rules that their own courts are supposed to use in litigation that might arise over counting absentee ballots. It is not enough that state law has rules for casting ballots. There needs to be clarity on whether ballots can still count if something has gone wrong in the process of casting of them, especially if the problem is not the voter’s fault. State legislatures have tended to leave these vote-counting matters for their courts to decide in the heat of the moment, and states have learned the hard way what happens when the counting rules are ambiguous. The lesson could be even harder this year, with the stakes perceived to be so high.

For example, it is one thing that Pennsylvania law provides that absentee ballots must arrive in local election offices by 8 p.m. on Election Day. But what if Philadelphia election officials miss their own deadline for sending absentee ballots to voters, so that the voters themselves — through no fault of their own — have no chance for complying with the deadline? In this situation, does Pennsylvania law still require invalidation of any absentee ballot from Philadelphia postmarked on or before Election Day but arriving afterward? It is important to answer this question in advance of the election, before an actual dispute arises.

Various groups have drafted guidance for election administration generally or, state judges and federal judges specifically, to help prepare for and handle election disputes, including those specifically involving vote-by-mail. (Full disclosure: I helped draft the guidance for election administration.) The more these efforts can reduce judicial disagreement over vote-counting disputes, the less likely that a dispute will spin out of control. It is harder for a candidate and his or her supporters to condemn a 9-0 U.S. Supreme Court decision than a 5-4 one.

In addition, Congress must better prepare for the possibility that it might face a dispute over the results of presidential election. Doing so requires revisiting the law that governs this possibility, the Electoral Count Act, which is dangerously ambiguous in its essential provisions. It matters less the particular way that Congress eliminates this statutory ambiguity than that it does so one way or another. Think of this statutory clarification as the electoral equivalent of stockpiling enough ventilators in the event of a pandemic. It has been easy for Congress to ignore the statutory ambiguity because the risk of a disputed presidential election reaching Congress has been low. But we know now the horrible cost of failing to prepare adequately for a catastrophe that has a low probability.

The coronavirus pandemic has already caused havoc to human lives and the economy. While a move to more widespread use of absentee ballots might be a salutary way to preserve voting given the outbreak, it will not come without its own risks. The one thing we cannot handle, on top of everything else, is an election dispute that leaves the country without a clear presidential winner by noon on January 20. The states and Congress must figure out now how to prepare for this possible situation. The alternative is like not worrying about a pandemic until it is too late.