The Polls Are Tightening. What Else Is New?

We’re in the last week of a political campaign, which means it’s time for the trailing campaign to declare that the race is suddenly breaking in his favor.

On Tuesday night, John McCain’s pollster churned out a press release (dressed up as an internal campaign memo) in which he asserted that “the McCain campaign has made impressive strides over the last week of tracking” and that “the campaign is functionally tied across the battleground states … with our numbers IMPROVING sharply over the last four tracks.”

For weeks, the Drudge Report and the other McCain-friendly outlets have been touting the narrative of a tightening race; and now, in the last week, independent national tracking polls have found Barack Obama’s lead wilting ever so slightly, from an average advantage of about eight points down to about six in the most recent data.

But it shouldn’t shock anybody if the polls tighten (or continue to tighten) between now and Election Day, because they almost always tighten in situations like this, when one candidate takes a pronounced lead into the homestretch. What’s more, though, is that history shows that this kind of late-in-the-game adjustment never amounts to much.

There was, for instance, the close of the 1992 campaign, when George H. W. Bush, who’d trailed by double digits in virtually every poll conducted after the Democratic convention in July, suddenly found himself pulling within five points of Bill Clinton in some polls.

A Gallup poll released on Oct. 30 of that year (the election was Nov. 2) showed Clinton ahead by just two points, 40 to 38 percent – a 13-point drop from just three weeks earlier. A day later, ABC News pegged the gap at just four points.

Naturally, Bush and his campaign, much like the McCain campaign with its memo this week, began shouting that the race had fundamentally changed and that they were poised for a come-from-behind victory.

“People have been looking for a reason to vote for the president,” was Bush strategist Charlie Black’s spin. “It’s late, but we’ve still got plenty of time. If the numbers continue to move at the rate they have since the last debate, we’d win by a point or two.”

Bush himself began railing against the “faithless Republicans and faithless Democrats who wrote me off two months ago” and declared that Clinton “can feel it slipping away.”

It was all a mirage, of course. Clinton’s expansive lead throughout the fall was always an illusion, the function of a large number of Republicans and Republican-leaning voters either refusing to back Bush in polls or simply telling pollsters they wouldn’t turn out. As Election Day neared, their enthusiasm increased and they returned to the G.O.P. fold, thereby cutting the fat out of Clinton’s lead.

The Gallup and ABC numbers of the last week of the ’92 campaign were the best Bush could muster. Other polls still showed him lagging by bigger margins – 7 and 10 points in two polls released around the same time – and a closer look at the battleground states found Clinton in commanding position to win more than 270 electoral votes. Yes, the race was tightening some, but not in a way that really mattered that much. On Election Day, Clinton won a 43 to 37 percent popular vote verdict, with 370 electoral votes.

The late tightening trend was also evident in 1988, when Michael Dukakis, whose hapless campaign had fallen 15 (or more) points behind Bush in mid-October, crept to within five points in an NBC survey the weekend before the election.

“People making up their minds are making it up for Dukakis,” was the spin from Dayton Duncan, the campaign’s press secretary, while an enthused Dukakis told a Colorado crowd, “The race is tightening up. You sense it, you hear it, you taste it here in Colorado and all over the country.”

Dukakis lost by eight points, a 426-112 electoral vote landslide.

In 1996, Bob Dole lagged nearly 20 points behind Bill Clinton for much of the summer and fall. But even his moribund campaign witnessed a late uptick in support. In the last week of that race, most polls found Clinton’s lead shrinking to around 10 points, but one – the Zogby poll – showed Dole edging within four, a result that prompted the G.O.P. nominee to declare (what else?) that the late-deciders were breaking his way and that he was on course for a narrow upset victory.

But Dole, like Dukakis and Bush in ’92, was trounced, losing to Clinton by eight points in the popular vote and 379-159 in the Electoral College.

Even Walter Mondale got in on the act in 1984, although the onetime VP was in such dire position against Ronald Reagan that he couldn’t even point to significant tightening in any polls.

No matter: Mondale pointed out how large the crowds at his rallies were (without thanking the union bosses who had bused them in) and thundered that “the pollsters and the slick magazines are trying to tell us that the election is over. Well, I’ve got a little message for them – Public opinion polls don’t count at all. People vote, public opinion polls don’t vote, and we’re going to win.”

He lost 49 states.

The Polls Are Tightening. What Else Is New?