Tuesday, October 9, 2012

Is Obama Losing?

Interesting. A recent Pew research poll shows Obama trailing Romney by 49% to 45%. At around the same time, a CBS Research poll shows that Romney has closed a massive, nearly two-digit gap among women voters at 47% to 47%.

Is Obama losing? Did one shoddy debate performance blow it all?

No way.

First, I would be very silly to argue that there is some conspiracy behind the poll numbers. It sounds just as silly coming from a liberal as a conservative, and there's no merit to the argument anyway. But there is always, in every close election, one or two poll aberrations that show the trailing challenger ahead by two to four points. We saw it with Bush vs. Kerry in 2004, and interestingly, it came after the first of their televised debates as well.

What's going on here? Well, not a conspiracy, but something known as response bias. Here's how it works. Say you get a phone call from a pollster, and you are feeling down in the dumps about your candidate because he's trailing in the latest polls. In that case you're far more likely to tell the poll worker to go fuck off, because you don't feel like answering any depressing questions. But let's say that he's just given a great debate performance, and now you're fired up. By contrast, the other guy's supporters are down in the dumps over their candidate's poorer performance, and are more likely to ignore the telephone polls. The result is that more excited people answer the poll questions than non-excited ones do. This skews the results, and it appears, at least temporarily, that the polls have significantly shifted. This explains the Pew results, as well as the CBS poll results among women because (and I hate to be sexist here, but what I'm about to say would be agreed upon by the fairer sex as well), women are more emotional.

But let's not uncork the champagne just yet. On either side. The Pew research poll is just one among many.  The Rasmussen poll shows Obama and Romney tied. But then, it's always shown Obama and Romney either tied or giving Romney a slight lead! The reason for this is that Rasmussen does not call any cell phone subscribers, thus leaving a disproportionate number of young people off its tally, skewing the results in favor of Romney. But if Romney's had a bump in the polls, why doesn't Rasmussen then show a lead rather than a tie? Gallup polling has given Obama a one-point lead in its previous polls, and now it's showing as tied as well.

Clearly, the debate showed Obama is mortal. Previously, haters of Our Trophy President have felt as though he were invincible, and that 46% of the electorate is the best they could ever hope to achieve. It made them depressed. Obama's invulnerability is still true, but now his opponents feel energized enough to think that maybe The Big Joke (as I sometimes call Romney) could actually pull off an upset. The depression is temporarily over.

Ah, but not so fast! If one evaluates the debate by preparedness and eloquence of performance, Romney clearly won last Wednesday. But if one evaluates truthfulness and the ability to avoid making any serious gaffs, Romney lost big time. Say what you will about Obama's stammering and pauses, he made no gaffs. And Romney? He just couldn't help himself. In addition to telling the debate moderator, to his face, that he would cut funding for his employer (head slap!) he then insisted that he loved Big Bird, whose salary he would also have cut! (D'oh!) Already, people in Big Bird costumes are showing up at every Romney campaign rally. That's negative six points for romney, plus six points for Obama. Also, PolitiFact and FactCheck.org had already been picking apart the Romney/Ryan ticket so much that conservatives had taken to claiming conspiracy theory for those news sources as well. (Even the normally level-headed Jeff Wagner on WTMJ radio said that PolitiFact was to politics what the Lengerie Football League is to the NFL.) But they took almost a full week to pick apart all the lies and distortions that Mitt Romney told, brazenly, during his debate performance. With clarity, poise, and eloquence, he outright lied.

In other words, Romney lost on Instant Replay.

No matter what, it is clear that one Pew poll does not an election make. If multiple polls appear showing Romney with a clear-cut lead, then it's time to worry. Until then, it's just one poll. Nate Silver still calculates a 75% chance that Obama will win. Our President still leads in swing states, and that primes the stage for an electoral college victory, even in the unlikely event that Mitt Romney somehow wins the popular vote. And as excited as current polls might make Romney supporters today, they'll be plunged right back into deep depression tomorrow, after Obama re-takes the polling lead, and then defeats Romney on Tuesday the 16th.

They'll even go back to claiming that the polls have been tampered with.


Eric

*

No comments: