In election law, as in so many things, the word “reform,” when associated with a new idea, is usually a sure sign that mischief is afoot. A case in point: early voting reform. This innocuous sounding but insidious idea, which has led some 32 states to allow voting to commence as much as six weeks before Election Day, is a ticking time bomb at the heart of our electoral system. It is only a matter of time before it is revealed as a devastating error, a blue ribbon commission is established to determine what went wrong, and the idea is tossed onto the ash heap of history, along with other crackpot ideas that have gained currency, such as same day voter registration.

But, for now, the time bomb quietly ticks.

Under early voting regimes, which come in various shapes and sizes, a state abandons the traditional neighborhood-based Election Day and instead allows voters to cast their ballots long before Election Day at voting stations set up expressly for that purpose, through mail-in ballots, or perhaps other means. This option is available essentially to all voters, not merely to those who might need to vote early because they are absent from the country or their district due to military service or for other reasons. The point of early voting reform is not to accommodate those citizens who would find it impossible to vote on Election Day. Rather, early voting advocates have argued that it encourages more people to vote.

Maybe it does, though that is far from certain, and academic research to date has not borne that out. But the price we pay for this accommodation is steep and largely hidden. The widespread use of early voting in this year's election means that voters are not voting based on the same information. The traditional Election Day ensured that all Americans went to the polls having followed the same race, having heard the same debates, and having had the opportunity to weigh the same facts. The elected candidate therefore shared a broad mandate, and all Americans owned the outcome of the election equally, for good or ill. If there was buyer's remorse after the election, at least we knew we had all made the same choice based on the same information.

With early voting, there is no longer a single electorate. There are many electorates. There is the electorate that voted in September just after the conventions, the electorate that voted in October before the debates, and then the more informed electorate that voted on Election Day. The vote count on election eve is no longer a snapshot in time reflecting our collective judgment. It is more like a "moving average"—an aggregation of what different Americans thought at different times based on different information.

What's wrong with that? For one thing, it means that candidates have less of an opportunity to build a case for their own election and convey it to the American people. Voting starts earlier and earlier each election cycle, and so the candidates have to rush their campaigns along, with some voters gong to the polls not long after the conventions.

And it means that critical events in the campaign, indeed the most substantive events, such as the debates, are viewed and accounted for by a smaller and smaller part of the electorate. By the time Mitt Romney exceeded expectations and scorched Barak Obama in the first debate this year, many thousands of Americans had already cast their votes. And if current "reform" trends continue, far more ballots will be cast by this point in future elections.

But that's not the worst part. A far more troubling scenario would emerge if an event took place late in the campaign that fundamentally changed—or should have changed—the voters' calculus. A war breaks out. A scandal erupts. A grainy video surfaces that reveals a candidate's past act of corruption or depravity, or worse. Not an everyday occurrence, to be sure, but also not so far fetched in the wake of everything that came to light during the John Edwards case, for example, or things we sometimes learn of our former presidents after they are elected and serve. What would happen if a major revelation or development occurred after tens of millions of Americans had already voted, but millions more had yet to cast their ballots—perhaps too few to alter the outcome already determined by millions of early voters? Lending new meaning to the concept of the October surprise, an election decided in such circumstances would be tragically tainted.

The traditional American system for elections is imperfect, of course, but it has survived the test of time. Today's multiplicity of reform experiments intended to expand voter participation—early voting is just one of many examples—are destabilizing that traditional system and are laden with unintended consequences. Elections are serious things. We tamper with them at our peril. There are some glaring weaknesses in our electoral system, but the traditional civic act of voting in neighborhood polling places on Election Day isn’t one of them.

Robert K. Kelner is chairman of Covington & Burling LLP’s Election and Political Law Practice Group in Washington, D.C. The views expressed are his own.

Next Page