This Election Night Could Be Over Early

For anyone who’s blocked out the entire night of Tuesday, Nov. 4, for election-watching, keep this in mind: It might not be necessary.

It’s been 12 years since the outcome of a presidential race was obvious within minutes of the first poll closings, and in that time an entire generation has grown up expecting that every election night will feature the same protracted back-and-forth drama that marked the 2000 and 2004 elections.

But from a historical standpoint, the back-to-back nail-biters that we witnessed this decade are extremely rare. Before 2004 and 2000, the last time the top two candidates were separated by fewer than 100 electoral votes for at least two consecutive elections came in the late 19th century, when the four elections between 1876 and 1888 were won by razor-thin Electoral College margins (twice by candidates who lost the popular vote). And while the Electoral College margin widened in the 1892 Grover Cleveland-Benjamin Harrison rematch, that race was also close (three points) in the popular vote.

After that, genuinely close presidential races went back to being unusual, once-every-two-decades-or-so occurrences – until this decade. In the five elections preceding the 2000 campaign, the best showing by a losing candidate was George H. W. Bush’s 168 electoral votes in 1992. The closest popular-vote margin in that period was also in ’92, a three-way race in which Bush finished with 37 percent (the worst showing for an incumbent president since William Howard Taft in 1912), six points behind Bill Clinton. In the entire 20th century (not counting 2000), there were only five election night barn burners: 1976, 1968, 1960, 1948 and 1916.

The trend in national and key state polling of the last month suggests that, despite the unusually high interest in the campaign and supposedly intense polarization of the electorate, Election Night 2008 will return to the relatively ho-hum form that defined the end of last century, when television networks would routinely find themselves struggling to give viewers a reason to tune in past 9 or so.

Consider 1992, the closest of the five pre-2000 blowouts. Bush had trailed Clinton in every single poll conducted after the July Democratic convention, although the gap had narrowed somewhat in the closing weeks of the race. So there was an expectation heading into Election Day that he’d lose – although since he was the incumbent and since his party hadn’t been defeated in 16 years, observers were still on-guard for a surprise.

But the late afternoon exit polls pointed to a convincing Clinton win, and so did the early returns. The first states to close their polls were (and still are) Kentucky and Indiana. Very quickly, the networks called Kentucky for Clinton, the first time since 1976 a Democrat had carried the state, while they deemed Indiana too close to call – another telling development, since the state hadn’t voted for a Democrat since 1964 and was home to Dan Quayle, Bush’s vice president. Polls in Georgia – another Republican bastion in recent elections – also closed early, and the state was called for Clinton.

At that point, not long after 8, it was obvious that there had been no late surge to Bush. The networks resisted formally calling the election until more states had reported, but it was obvious that Clinton would be the winner. His victory speech in Little Rock came well before most people’s bedtimes on the East Coast.

And that was the most suspenseful of the five pre-2000 election nights. In 1980, a race that had been a statistical tie the week before Election Day, the networks declared Ronald Reagan the winner well before 9 – he ended up beating Jimmy Carter in 44 states, a 489-49 electoral vote drubbing. In 1984, television anchors tried to induce viewers to stay tuned by pointing out that Reagan might win all 50 states; there was no pretending the race was remotely competitive. In 1988, viewers who tuned in at 8 p.m. saw an electoral vote scoreboard that already read: Bush 81, Dukakis 0. In 1996, when Bob Dole somehow cobbled together 159 electoral votes, it wasn’t much more interesting

The shoddy exit polls and botched state projections (i.e. Florida 2000) that undercut viewer confidence this decade should ensure that the television networks will err on the side of patience this Nov. 4.

But it should be very easy very early to figure out if it’s worth staying tuned. As usual, the earliest poll closings will be in Kentucky and Indiana, where parts of each dual time zone state will close at 6 p.m. Eastern time – although no numbers will be reported until the rest are closed at 7. Georgia, South Carolina, Vermont and Virginia will also close at 7.

Just as in ’92, the returns from these states could be decisive. If Indiana is too close to call, it’s bad news for John McCain – and if the state is called for Barack Obama early, it’s probably fatal for McCain. If Kentucky, where McCain now leads by about 10 points, is anything but a quick call for the G.O.P. nominee, this would also signal a long night for McCain. The same is true in Georgia (and South Carolina, for that matter). And if Virginia is quickly called for Obama, it would also suggest that the election is his – just as it would confirm that the much-discussed “Bradley effect” was all hype.

If the news from those states isn’t definitive, better answers should come just after 8, when Pennsylvania’s polls close. Realistically, McCain has no chance of winning the White House without this state; if it is called quickly for Obama, there won’t be much left for McCain to cling to.

Whatever the result, this Election Night will be a historic one. Whether it will be a suspenseful one is another story entirely.

This Election Night Could Be Over Early