The Volokh Conspiracy

Mostly law professors | Sometimes contrarian | Often libertarian | Always independent

Electoral College

The Worst Defense of the Electoral College Yet

Will a popular vote produce more fraud?


My first post is already making waves, with Rick Hasen at the Election Law Blog describing it as the worst defense of the electoral college yet. In his view, the idea of widespread fraud in a popular vote is "ridiculous," given "the tens of millions of voters who would have to be involved." We simply "do not see such election fraud on such a massive scale"—and in any case, "going after one swing state" would be a much easier task for fraudsters.

As I noted, the partisan-fraud concern is just one point in favor of the electoral college, not a slam-dunk argument for keeping it. There may be lots of other, stronger reasons at stake. But I'm still not persuaded by Hasen's claim that a result-altering fraud is unthinkable. The 2000 election had a popular-vote margin of roughly 500,000 (or 0.5%), not tens of millions. The 1960 election had a margin of roughly 100,000 votes (or 0.16%). In some ways, we should expect closer margins in a true popular-vote world, because both campaigns would be laser-focused on getting to 50.1% (or a plurality, etc.)—as opposed to getting to 270, with the popular vote merely an afterthought.

Hasen is entirely correct that the margins in a single state can be much closer. According to the official 2000 tally, the Florida election was decided by 537 votes, or 0.009% of Florida's votes. But Florida was a state whose legislators and local officials often came from different parties and were often at loggerheads during the recount. An election-swinging 1000 votes might have been harder to fake or suppress in Florida 2000 than an extra 10,000 votes would be in each of the ten reddest or bluest states in 2032, depending on our institutions and partisan incentives.

It's also true that we don't see fraud on such a scale right now. But as others have pointed out, right now such fraud wouldn't really help: in a deep-red or deep-blue state, it would risk prosecution and jail time for relatively little benefit. That isn't very strong evidence that such fraud would stay unthinkable in a world where one-party states could bring home the White House. (No one ever calls a single House campaign like NC09 a "Flight 93 election.")

The electoral college is a guardrail, not a guarantee. There can be good reasons for taking down guardrails, when they prevent us from going where we want to go. But we shouldn't assume that, just because we've never plunged over the cliff before, we won't do so in future.