When I published my first book, Code and Other Laws of Cyberspace (1999), a New York Times reviewer wrote,
Lessig plays digital Cassandra: he predicts that the Internet will become a monster that tracks our every move, but that nobody will heed his warning.
I didn’t get it. I believed my arguments about how and why the Internet would evolve to “track[] our every move” were obvious. Maybe too obvious. I naively believed that people would agree, and we’d do something about it. Only after the Internet had become, maybe not a “monster,” but certainly not what we had hoped, and did indeed “track[] our every move” did I come across the review again and realize that, yes, I had indeed been a “digital Cassandra.” I (and many others) had been right, yet none had listened.
Yet today I feel like a democracy Cassandra. Over the past year, in a book I wrote with Matthew Seligman, How to Steal a Presidential Election, and in writing and speaking wherever I can, I’ve described the threats as I see them for the 2024 election. Specifically, how would you play it if your game was to steal it? Those predictions entailed remedies — steps that should be taken now to protect us from those threats then. Yet none of those steps have been taken. On the contrary, the changes that have been made have only made the threats worse.
Don’t get me wrong. I don’t think I’m the smartest guy on the field. I have enormous respect for the lawyers and scholars who have been thinking through the same issues. And growing old — something I’ve been pretty successful at—is the process of learning, through living, humility.
But I can’t escape the thought that something is being missed among lawyers and strategists just now. That’s the Cassandra feeling. And so, to at least give us a way to mark my error, let me state clearly what I think we’re missing and why I think it is so dangerous.
To see the threat, consider what happened in Hawaii in 1960. Richard Nixon was initially declared the winner of the popular vote; his electors were certified as the electors representing Hawaii. But Democrats noticed an error in the tabulation and requested a recount. That recount would take time, extending beyond the day when presidential electors were to meet and vote. So on that day — Electors Day—both the Kennedy and Nixon electors met in the state capitol; both signed documents swearing to be the “electors duly and legally appointed and qualified by and for the State of Hawaii”; both cast their votes for their respective candidates. By the end of the month, the recount was completed. Kennedy was declared to be the winner of the popular vote. The governor issued a new certification. And on January 6, Vice President Richard Nixon counted Hawaii in favor of John Kennedy.
Ok, here’s the key point: If this very same sequence of events happened today, the result would be different. The law would tell Nixon to ignore the votes for Kennedy, and count the state for himself. More troublingly, at least for the Kennedy electors, if indeed they did meet and vote on Electors Day, they would be deemed “fake electors” and, depending on the state, prosecuted for fraud. Again, if the very same sequence of events happened today, the wrong candidate would be declared the winner of the state’s popular election.
That fact is, in my view, among the greatest threats that we face in 2024 — because this truth will incentivize bad behavior, possibly on either side, since it offers a clear way to secure a result that a more complete count, or final resolution of any legal disputes, might throw into jeopardy.
How did we get to this strange inversion of the obviously correct answer — that Kennedy’s votes should be counted? To see how this happened, let’s update the example a bit, to make its point more relevant to 2024.
Imagine on Election Day, 2024, Donald Trump is ahead in the popular vote in Georgia. His lead, however, begins to dwindle as absentee ballots are counted across the state. But late into the process of counting those ballots, a state judge declares certain categories of absentee ballots to be illegal. The judge orders them removed from the overall count. Lawyers for Democrats appeal the ruling, but the other side successfully delays or slows the consideration of that appeal. On December 11, the governor is required to certify the results of the election. As of December 11, given the count of the ballots then deemed legal, Donald Trump is the winner. On December 17, Electors Day, the Harris electors attempt to meet to cast their ballots for their candidate. Officials in Georgia threaten those electors with prosecution for “fraud” because they are not, as of December 17, the duly elected electors of the State of Georgia. After the electors fail to meet and cast their ballots, lawyers for Donald Trump file a motion to dismiss as moot the appeal of the lower court’s ruling that the absentee ballots are illegal. Trump should win that motion, because even if Harris prevailed, the failure of her electors to vote means there could be no votes for her from Georgia for Congress to count on January 6. The Constitution requires electors to meet and vote on the same day; the Harris electors didn’t. The court should thus dismiss the appeal; Trump’s victory would be assured.
There are plenty of ways to fight this hypothetical. An appeal could move more quickly; the Harris electors could be brave and meet and vote regardless of the threat of “fake elector” prosecution. More fundamentally, this very same story could be told the other way around: Imagine it’s Pennsylvania, and it’s Harris who is ahead when a lower court ruling invalidates ballots that would favor Trump. The point is not the specifics. It is much more general: The changes we have made — both in the law and expectations—surrounding a presidential election make the system more vulnerable to manipulation, not less.
Two specific changes have produced this weakness.
First, in our zeal to punish those involved in the efforts leading up to January 6, we have vilified the Trump electors who met and cast their ballots without the certification of their governor. In Georgia, Wisconsin, Michigan, and Arizona, both civil and criminal actions have been brought against those electors for meeting and voting on Electors Day. Those so-called “fake electors,” these prosecutions alleged, were involved in a conspiracy to overthrow the election.
But I’ve looked at many of these claims: Whatever you say about the lawyers who architected the (false) argument to Vice President Pence that he had the authority to pick the slates of electors on January 6 that he would count, the actual electors who met and voted on December 14 did so because they believed they needed to to preserve the opportunity for courts to determine finally who won the popular vote in their state on Election Day. No doubt, the claim that there were enough fraudulent or illegal votes to change the results proved baseless. But in at least Wisconsin and Arizona, there were valid legal arguments that had not been resolved finally by the day that electors were to vote, and whether it was likely or not that those claims would be resolved in Trump’s favor, it was not crazy to believe that Trump had a claim. In Arizona, for example, lower courts had wrongly interpreted the Electoral Count Act to require them to resolve any electoral dispute six days before Electors Day. That legal conclusion was plainly wrong. And in Wisconsin, applying at least the principles articulated by Chief Justice Rehnquist in Bush v. Gore, changes in balloting procedures could easily have raised a constitutional question that the Supreme Court could have resolved in Trump’s favor. Again, I am not saying I agree with these arguments; I am only saying they were not crazy, and that it was right for electors to meet and cast their ballots on Electors Day, so long as there was an open question about which slate of electors had been chosen on Election Day.
Defenders of this war on the Trump electors say that there was no reason for them to swear that they were the duly elected electors, when in fact, at the time they voted, they were not. They could have, as some did, expressly stated that they were acting contingently, depending on the results in ongoing litigation.
That’s true — and in 2024, if this need arises, that’s exactly what any uncertified-but-needing-to-vote elector should say. But the same could have been said about the Kennedy electors in 1960: They, too, swore they were the duly elected electors when they were not. The certificate they signed said that they were the “electors duly and legally appointed and qualified by and for the State of Hawaii.” In neither case should anyone consider that statement “fraudulent” because, in both cases, the documents that were signed were implicitly contingent: They were meant to have a legal effect if and only if their candidates were declared to have won the election on Election Day. If, for whatever reason — a recount, a declaration that certain ballots should be counted or should not be counted, a ruling that a recount procedure violated the Equal Protection Clause, whatever— the results flipped, then them voting assured that Congress would have the chance to count the votes for the actual winner of the election in their state. And if the results did not flip, then the National Archive would have an extra set of papers to store in its historical vault.
The second change that has produced this weakness is Congress’ fault. In December 2022, Congress surprised many by passing the Electoral Count Reform Act. That law corrected some obvious flaws with the infrastructure guiding the counting of electoral votes. Some of those changes were compellingly good and needed. But some only made the problem worse. Specifically, Congress shortened the time that states have to resolve election disputes by directing the governor to certify the results six days before the electoral college votes, and by giving Congress no way to account for any changes after that certification. Everyone knew in 2022 that litigation takes time; one of the weaknesses in the litigation in 2020 was the number of courts that short-circuited any real analysis of the underlying claims because time was too short. However, the ECRA didn’t address that problem in the way some — like Senator Rubio — had suggested by moving Electors Day later in the process. Instead, Congress did the opposite: It took an already shortened timeframe and made it even shorter.
No doubt, with the help of organizations like Protect Democracy, states have done a great deal to improve their processes to resolve disputes and contests more quickly. But these changes will not change the incentives of candidates in the middle of a close contest: The rules now give the prevailing side an irresistible incentive to slow the process of resolving any dispute about the election. If the lawyers for the leading candidate can run the clock past the mandatory certification date, then their candidate will prevail. We have thus created a system where the lawyers’ incentive is not to resolve claims in a way that could convince the public that the election was decided honestly and fairly. Instead, the system amplifies the incentive to fudge any contest, to run the clock out.
There is no good reason to force the states to truncate their process for resolving a close election. Both sides need to give the public confidence that legitimate claims were considered fairly and responsibly. If lawyers make illegitimate claims, they can — and should—be disciplined. There is no reason that a state can’t be deliberate about resolving its election contest. The only real deadline is when Congress determines to count the electoral votes — as Hawaii, in 1960, demonstrated. Thus, the problem we’ve created for the states is both improper and unnecessary. Insult, meet injury.
Yet even today, there is something Congress could do. Congress could amend the Electoral Count Reform Act in a small but critical way, thereby removing this perverse incentive completely.
Here’s a proposed new section to the ECRA that would do that, while giving legal claims about the winner of an election a chance of being resolved fairly:
3 U.S.C. §14: If on the date that electors are to meet to cast their ballot, there is ongoing litigation in any state or federal court about which slate of electors was appointed on Election Day in that state, then electors who might plausibly be determined to have been appointed on Election Day shall meet and cast their ballot as specified in Section 7. If those electors have not been certified by the state’s executive, they shall mark their ballots as “Contingent,” but otherwise follow the procedures of Sections 8–13. If litigation is resolved finally in favor of the appointment of these contingent electors before January 3, then the state’s executive shall certify, and Congress shall count, the votes of the contingent electors, unless both Houses deem the votes not regularly given.
This provision would not disrupt the ordinary process in the vast majority of states. It would be operative only in states with ongoing litigation about which candidate won. But most importantly, the provision would remove any incentive that lawyers would have to slow the process of resolving election disputes. The presumptive result of the ECRA would govern unless, by January 3, a court ruled differently. The change would almost double the time courts would have to do their work. It would remove any need to truncate legal analyses, giving both sides ample opportunity to have their claims heard and resolved fairly.
I get that it is impossibly difficult to imagine Congress acting before November 5. If there’s any chance to flip that likelihood, it is the recognition that both sides could play this game. There is no reason for either Republicans or Democrats to enter this election with a strategic incentive to skew the results. We should instead tweak the rules, to remove that incentive, and give all Americans the confidence that any contest was resolved fairly.
