So Who's Winning This Thing?
Andrew Romano Real Clear Politics.com
The current Real Clear Politics electoral map; it reflects the latest polls and is not a prediction of the outcome
Forget my conversation with the sardonic, Southern-fried Fred Thompson. Forget my coverage of the gassy St. Paul protests. And forget my behind-the-scenes profile of the man who's prepping Sarah Palin for the spotlight. When I returned to New York from Minnesota last week, I discovered over the course of dozens of conversations with actual human beings that there's only one question on the minds of people who aren't paid to obsess over Election 2008: So who's going to win this thing?
Never mind that my business card says "reporter" and not "Nostradamus." When I gave an honest answer--"I have no idea"*--people pounced. "What do you mean you have no idea?" they said. "You've been covering this stuff for, like, a year." They had a point. So instead of avoiding the question, I've decided to launch Stumper's postconvention coverage with an in-depth look at where the race stands right now. Because while polls are terrible at predicting what will happen two months hence, they're actually pretty good at showing how people would vote if the election were held today. (After all, that is the question pollsters ask their victims.) And judging by the latest stats, the race has reached its most interesting and perplexing point since... well, ever.
From the end of the primary season on June 3 until the shortly before the start of the Democratic Convention late last month, the Real Clear Politics average--a blend of the most recent half dozen or so national match-ups between Barack Obama and John McCain--told an essentially static story: despite never breaking the magical 50 percent mark, Obama led McCain by a steady three to six points for months. But two back-to-back conventions--which typically mark the point when the public begins to pay attention--have scrambled those jets. After arriving in Denver tied with McCain for the first time since late May--45.1 percent for Obama to 43.9 percent for McCain--Obama appeared to depart with a five-point boost. By Sept. 2, in fact, he was ahead of McCain 49.2 to 42.8--his largest share of the vote to date and his widest average lead (at 6.4 percent) since late June. Now, however, that lead has been neutralized. With the festivities in St. Paul finally finished, the race has reverted to a tie--46.7 to 45.5. Only this time it's McCain who has the advantage in the RCP average--his first since Hillary Clinton hung up her spurs.
In other words, McCain--yes, that McCain--is winning, at least on a national level.
Should Obamans be worried? Absolutely. But that's not because the latest polls have revealed something shocking about the election. Instead, the new numbers simply confirm what expert observers knew all along--that the race would get really, really close once the public finally tuned in. And before you Democrats start shopping for cyanide--and/or before you Republicans start booking rooms for Mac's inaugural bash--there are three caveats to consider.
First, while McCain's convention bump is real, it's far too early to tell whether he's actually ahead of Obama. That's because the Real Clear Politics average includes one poll--the new USA Today/Gallup--that shows something dramatically different from the rest of the post-convention surveys: McCain leading 54 percent to 44 percent, the largest edge for either candidate since Obama mounted a few double-digit margins in mid-June. Given that the two other soundings taken over the same time period resulted in a dead heat--Rasmussen, 48-47; CNN/ORC, 48-48--it's prudent, for now, to assume that Gallup is a bit of an outlier until other polls confirm its 10-point split.
The second caveat relates to the reason why Gallup shows such a wide margin. While most polling outfits at this stage focus solely on registered voters (i.e., anyone who answers the phone and is registered to vote), Gallup has also turned its attention to 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"). Screening for likely voters is tricky business, especially this year. For starters, the old screening models--which tend to favor tried-and-true Republican demographics--may not apply to an election in which Team Obama is investing massive resources in turning out subgroups (young people, African-Americans) traditionally underrepresented at the polls. But the more important thing to remember is that shifts in the likely voter pool correspond to fluctuation in voter enthusiasm--and thus, according to statistical studies, vastly "exaggerate the volatility of voter preferences."
Here's the deal. Thanks to Sarah Palin and the party in St. Paul, "there has been a very substantial jump in the percentage of Republicans saying they are more enthusiastic about voting in this election, from 42% a week ago (after the Democratic convention, but before the Republican convention) to 60% today," according to Gallup--a leap that has narrowed the "enthusiasm gap" between the parties from "19 points in the Democrats' favor a week ago to only seven points today." Gallup credits McCain's newfound national lead to this burst of enthusiasm. But the problem is that while the GOP's joviality is an important development--it means that McCain will have an easier time turning out his base in November--revved-up Republicans only account for about 45 percent of the electorate. That's not nearly enough people to boost McCain to an actual 10-point lead. What a group that size can do, however--especially when they get excited--is skew Gallup's assessment of who's likely to vote further to the right than usual. And that, according to polling expert Nate Silver, is exactly what they've done: "Republicans, especially evangelical conservatives, are pumped now, after having been indifferent toward John McCain for most of the election cycle. They may be picking up the phone when a pollster calls [where] they had been screening out the call before, perhaps to the extent that they are biasing the sample." Case in point: the same Gallup poll shows McCain ahead 50 percent to 46 percent among registered voters--i.e., everyone Gallup called, as opposed to only the ones who expressed extreme enthusiasm. That's probably the more accurate result.
Our final caveat? Presidential election are fought on a state-by-state basis--not in the national polls. Here, the picture doesn't look quite as rosy for McCain. According to Real Clear Politics, Obama currently leads in each of John Kerry's 2004 states, including top McCain targets Michigan (+4.3 percent), New Hampshire (+0.3 percent), Pennsylvania (+5.0 percent) and Wisconsin (+7.2 percent). He also leads by healthy margins in a pair of Bush states: New Mexico (+4.3 percent) and Iowa (+9.0 percent). If Obama can hold these advantages until Election Day, he'll wind up with at least 263 electoral votes--seven shy of victory. That's where the red states of Virginia and Colorado come in. At this point, Obama's leading in the latter (+0.4 percent) and tied in the former. Win either one and the White House is his. At this point, RCP gives the Democrat 273 EVs (Colorado, no Virginia) to McCain's 265; the prediction whizzes at FiveThirtyEight.com are even more optimistic, projecting additional Obama victories in Virginia and Ohio for a final score of 304 to 234. Which means that while McCain is "winning" nationally, Obama is ahead in the electoral college.
That said, we should probably add one more caveat to the list: no battleground state polls have been released since the second day of the Republican Convention. If the national surveys are right and McCain has in fact received a five-point post-St. Paul bounce, that enthusiasm will almost certainly trickle down. That means that Obama's narrow leads in New Hampshire and Colorado could vanish, and that McCain could pull ahead in Virginia. Palin has already taken Alaska and Georgia (evangelicals) off of Obama's wish list and put Wisconsin, Oregon and Washington--implausibly--on McCain's. If the post-convention state polls fluctuate as much as their national counterparts have, we'll be right back where we began: |
