Is John McCain--gasp!--beating Barack Obama in the race for the White House?

According to the latest USA Today/Gallup poll, the answer is yes. (No word yet on what the survey says about the existence of Santa Claus. [Insert rimshot here.]) Released at 2:16 this afternoon, the new Gallup numbers give McCain a 49 to 45 percent edge over his rival from Illinois. It's a remarkable result, if only for its rarity. The math is pretty simple. Since the start of April, 61 national polls have been released. Five show a tie. Fifty-two put Obama ahead. Only four--including today's, the first since the Democratic primary battle ended in early June--give McCain the lead. Over at McCain's Crystal City headquarters, we suspect that Steve Schmidt and Co. are breaking out a bottle or three of Maker's Mark--if only because the stats arrive at the perfect moment to fuel an emerging media storyline that downplays the political potency of Obama's overseas adventures by asking, as Timesman Adam Nagourney does today, "where's the bounce?" "The question is why — given how sour Americans feel about President Bush and the Republican party, and the perception that Mr. Obama is running a better campaign than Senator John McCain — the senator from Illinois is not scoring even higher in national opinion polls," he writes. And that was B.G. (Before Gallup).

Still, how should normal people--those of us who aren't paid political partisans--interpret this surprising sounding? It's a fascinating question--especially given that Gallup also released a tracking poll earlier today showing Obama enjoying his largest lead (48-40) in months. To most sane individuals, that absurd 12-point discrepancy--the polls were conducted by the same firm, after all, over the same two-day period (July 25-27)--is reason enough to refrain from putting any stock whatsoever in individual opinion surveys. This response makes some sense. (Standard disclaimer: rely on the polling averages at RealClear Politics and Pollster--which show Obama ahead by about four points overall--for a more accurate snapshot of the current state of the race.) But a closer look at today's twinned Gallup results reveals a plausible explanation for the contradiction--and may, in fact, say a lot about where the 2008 contest is headed.

The thing to remember is that the two polls were actually targeting two different audiences. While Gallup's tracking poll focused on 2,674 registered voters (i.e., anyone who answers the phone and is registered to vote) the USA Today survey homed in on a more elusive group of 791 likely voters (i.e., respondents most likely to show up on Election Day, according to "how much thought [they] have given to the election, how often they say they vote and whether they plan to vote in the election in November"). Likely voter models far from flawless, but they do help measure how voter enthusiasm may affect an election. In this case, according to Gallup polling director Frank Newport, the number of likely GOP voters was higher than in June's USA Today survey--mainly because Obama's international trip, the media's "laudatory" coverage and McCain's vigorous attacks "may have had the side effect of energizing Republicans." The result? A 10-point swing for McCain since June.

The lesson of Gallup's contradictory surveys is hardly set in stone. But one thing seems clear: McCain fares better when voters on the right are given a visceral, immediate reason to resent Obama--like his journey to Europe and the Mideast, which many saw as presumptuous--and are therefore more eager to vote. Most polls of registered voters released since June show Obama stuck below 50 percent--a reminder, says Democratic pollster Peter Hart, of the particular obstacles he faces. “Here’s a 46-year-old African American with a narrative that is very unusual and that few other Americans can relate to,” he told the Times. “Add to that the fact that he has had four years in the United States Senate and very little international experience. That’s a large leap for the American public to make.” But while some respondents refrain from voting for Obama for those very reasons, they're not always enough to inspire respondents to vote against him--which is why the same surveys show McCain mired in the low 40s with 10 percent still undecided. It's only when voters are given something polarizing to cling to--a criticism, an image, an event--that they can tilt the outcome toward the Republican nominee (as the new USA Today numbers demonstrate). Simply put, inchoate doubts about Obama won't lift McCain to victory; specific attacks may. This afternoon, I called the Arizona senator's current anti-Obama onslaught more harmful than helpful. Today's mismatched pair of Gallup polls may be an early sign that I was wrong--and that McCain won't be cutting back on the swipes any time soon.

UPDATE, 11:40 p.m.: Reader M.E. makes a smart point:

The same Gallup/USA poll had a slightly larger registered 900 voter sample that has Obama leading 47-44, which wasn't cited in your article.  By dropping 16% of Obama registered voters via the likely voter screen, with such 'unlikely' voters including first time registrants, and only 2% of McCain's registered voters, the gallup lv model gives McCain a 49-45 lead. In other words, the Gallup likely voter model decided that Obama voters were eight times more likely than McCain voters not to show up at the polls. As others have noted, this is an absurd result in an election where poll after poll has shown that greater enthusiasm is on the Democratic side.  Are likely voter models that weigh against first time voters, and in many cases also against minority voters, likely to be accurate in this election?

My answer: probably not. But the point of this post wasn't to debate the accuracy of Gallup's likely voter screen and the results it produced; there's no way to know now, 99 days from Nov. 4, who will show up to vote. (If I had to guess, I'd say that Obama has a slight lead--maybe two percent overall.) Instead, the point was to use the USA Today poll as a sort of hypothetical scenario--what would have to happen, at this point in the race, for McCain to win by four points? The answer I arrived at, reasoning backwards from the result, is that Republicans and Republican leaners have to be so revved up to vote against Obama that they turn out in droves. Otherwise--as the surveys of registered voters show--Obama wins. Which is why I suspect McCain will continue to fuel the fire.

Also: a lot of readers have remarked on my "sarcasm" in the opening paragraphs. Just to be clear, I was mocking how the media is responding to results--with shock!--not responding that way myself. I can see how this would be unclear, given that I'm a technically a member of the despised MSM myself. But I think you'll see that the rest of the item, far from being sarcastic, is actually a pretty sober analysis what this new poll--which is getting a lot of attention because it deviates from 57 of 61 released since the start of April--might mean for McCain and Obama going forward. I'm not at all surprised that this race is essentially tied.