George F. Will: Farewell, Election Day

The sentiment expressed by a sly bumper sticker this year (EVERY DISASTER IS A CHANGE) is a cousin of this axiom: Most improvements make matters worse. That axiom is pertinent to this election season because, for many years now, improvers have been toiling to perfect voting procedures.

One result is that by last week, five weeks before what is anachronistically called Election Day, and before any debates between the presidential and vice presidential candidates, and before Congress acts on the unprecedented legislative proposal for coping with the financial crisis—before all these events, Americans in 24 states and the District of Columbia, which have 50 percent of the nation's population, began casting their ballots.

Dislocation of the calendar is disorienting. Next year's World Series will end in November. This year, what the nation thinks of as the November elections actually began in September.

Seven presidential elections ago, in 1980, only 5 percent of the votes were cast before what really was Election Day. If this year, like 2004, produces an increase in early voting, close to 30 percent of the votes will have been cast before Nov. 4. In some states, more than 30 percent will have been. In at least five states, a majority of the votes will be cast early, including the swing states of New Mexico (51 percent early in 2004) and Nevada (53 percent). In Washington state, it will be more than 70 percent. In Oregon, which for a decade has voted entirely by mail, the figure will be virtually 100 percent.

In its rate of voter participation, Oregon ranks high among the states without same-day registration, and it spends 30 percent less on elections than it would were it still maintaining statewide voting sites and machines. Oregonians overwhelmingly adopted mail voting by a referendum, and overwhelmingly approve it.

So what is wrong with early voting? Even leaving aside the large matter of increased potential for fraud in voting by absentee ballots, there are two costs to early voting.

First, for tens of millions of early voters, the campaign process of informing and persuading is effectively truncated. Now, there is evidence that early voters are more partisan and informed than other voters and hence are less likely than the rest of the electorate to be swayed by events late in an election season. Nevertheless, early voting increasingly affects the rhythms of campaigns, forcing the front-loading of arguments.

An early reason for absentee voting was to prevent the disenfranchisement of people—Civil War soldiers in the field—who could not get to polling places. Today, however, as John Fortier of the American Enterprise Institute notes in his book "Absentee and Early Voting: Trends, Promises, and Perils," the academic consensus is that mail and absentee-ballot voting "has little or no effect on voter turnout except in low-turnout elections." This might be partly because dispersing voting over many weeks complicates voter-mobilization efforts. Furthermore, those unusually partisan and informed people who take advantage of early voting options are not typically the poor or people who, absent those options, would be nonvoters.

The second problem with early voting is that one of its supposed benefits is actually a subtraction from civic health. The benefit is that it makes voting easier—indeed, essentially effortless. But surely the quality of the electoral turnout declines when the quantity is increased by "convenience voting."

A word describes most of the people who will vote only if a ballot is shoved through their mail slot: "slothful." What kind of people will not bestir themselves to exercise their franchise if doing so requires them to get off their couches and visit neighborhood polling places? People who are barely interested, and hence probably are barely informed.

The requirement that voters go to a polling place is a slight filter that has the negative function of screening out people who are almost completely uninterested. But the requirement also has a positive virtue.

The great national coming-together that Election Day has been and should be is a rare communitarian moment in this nation of increasingly inwardly turned individualists who are plugged into their iPods or lost in reveries with their iPhones. It is one thing, and an admirable thing, to privatize airports, turnpikes and many other government entities and operations; it is not admirable to scatter to private spaces, and over many weeks, the supreme act of collective public choice. The coming of the public into public places for the peaceful allocation of public power should be an exhilarating episode in our civic liturgy.

With political excitement at an amazing boil this year, election officials in some communities are hoping that a surge of early voting will reduce the possibility that unusually heavy turnouts on Nov. 4 will cause local polling mechanisms to buckle under the strain. Good grief. Has the approach of Election Day—the fact that 2008 is divisible by four—taken these officials by surprise?

Elections are government projects, so perhaps it is utopian to expect them to be well run. Still, it is time for second—or in some cases, first—thoughts about the fading away of Election Day.