McCain’s chances, in essence, boil down to the polling being significantly wrong, for such reasons as a Bradley Effect or “Shy Tory” Effect, or extreme complacency among Democratic voters. Our model recognizes that the actual margins of error in polling are much larger than the purported ones, and that when polls are wrong, they are often wrong in the same direction.
However, even if these phenomenon are manifest to some extent, it is unlikely that they are worth a full 6-7 points for McCain. Moreover, there are at least as many reasons to think that the polls are understating Obama’s support, because of such factors as the cellphone problem, his superior groundgame operation, and the substantial lead that he has built up among early voters.
[/quote]
Afx114, I’ve thought a lot about this, so please forgive the length of this post, but I want to take the time to respond to your comments. I’m no statistician, but I agree with you that the bulk of my argument comes down to my gut feeling that the polls may be “off”. By how much, I don’t know. None of us will know the answer to that question until tomorrow. But I’m guessing that this year the polls are even more inaccurate than they have historically been. At best, polls are just guessing games and estimates of future behavior based on currently expressed sentiments. And generally, those estimations are correct.
We can discuss the Bradley Effect and the extent of its impact till the cows come home. Clearly something was at work during the primaries, because Obama was indicated to be neck and neck with — or leading — Clinton in several big states, and the results often belied the polls. Somehow, for whatever reason, they were *inflated* for Obama. I’m not implying any ill will on the part of the pollsters. But for whatever reason, the polls taken during the primaries were not always very reliable indicators of Obama’s actual performance, and in some cases were just plain crazy-off. And yet, most people seem to have conveniently forgotten or have chosen to ignore this phenomena that was at work during the primaries.
I read an article today that posited that people may not have been racist, but simply objected to Obama on entirely race-neutral grounds, yet were reluctant to say so for fear of being branded a racist. That’s not racism; that’s white guilt playing on their conscience. Many white people may feel compelled to vote for Obama out of sympathy or a sense of reparations: “Our country has made it so hard for black people so get ahead for so long, how can I not vote for this clean-cut, intelligent black man? Oh, but he does want to ‘spread the wealth, and that’s just not what America is about. And that Wright character is a bit unnerving…and who is Bill Ayers? I’d love to vote for Obama, he seems like such a nice young man, but I just don’t know if I can do it.” And so they tell pollsters that they will vote for Obama. Or that they haven’t decided.
An interesting story is also told by a priest who is hearing about accounts from all over the country: people who are literally coming to his parish to CONFESS that they aren’t voting for Obama. Anecdotal evidence, to be sure. But what if it’s indicative of something larger that the polls simply aren’t catching?
And there is a second phenomena that I suspect may be occurring, and it literally blows my mind that the media and the polling firms never discuss this. Unlike in previous elections, each party could pretty much count on unity going into the general election, regardless of how fervently their voters supported a losing candidate. I supported Edwards and truly disliked John “Lurch” Kerry, but I held my nose and voted for the guy anyway.
That’s not the case this year. There could be hundreds of thousands, if not millions, of bitter Hillary supporters who intend to either stay home tomorrow, or even worse, vote for McCain (because that actually takes a vote away from Obama and gives it to his opponent). How many of these so-called PUMAs (Party Unity My Ass) are there? Well, as late as August they represented 28% of Hillary’s 18 million supporters. That’s about 5.4 million Democrats.
Now, here’s the rub that fascinates me: when pollsters make their blind calls — after getting as many as 7 or 8 hang-ups — they get at least two pieces of information: who the person intends to vote for and what their party affiliation is. Then the pollsters ratchet their results up (or down) to reflect what thy anticipate to (a) that party’s representation across all registered voters in that particular state; (b) the likelihood of that person going to the polls; and (c) the estimated voter turnout in that party.
In this election, many polling firms have stated that they are boosting Democratic party affiliation and turnout by double-digits (over the GOP) to reflect record-breaking Democratic turnout.
But even assuming that turnout is record-breaking for Dems, there seems to be one not-so-tiny fly in the ointment: this year, moreso than ANY other year in recent memory, I don’t think it’s accurate to assume that all Dems will necessarily vote for Obama. So pollsters’ party ID models may be fundamentally flawed this year. Significantly? We don’t know.
What I don’t see happening is polling firms ratching Democratic party identification DOWN to reflect the fact that there are a helluva lot of bitter Hillary voters out there who aren’t going to toe the line. The polls simply assume that everyone is going to jump onto the wagon and head to the party.
Let’s not forget: Obama lost 5 of the last 7 Democratic primaries, even after he was declared the presumptive nominee by the media and voters were told that it was mathematically impossible for Hillary to win. And yet, millions of Hillary voters dragged themselves to the polls in the final days of the primary season, to vote for a candidate who stood NO chance of winning. Why? Were they intensely-devoted to Hillary? Or just passionately anti-Obama? I just know I’ve never seen anything like it. And my guess is that there are still plenty of these people still lurking around the polls, not listening to the media’s insistence that Obama’s a shoo-in.
So let’s crunch the numbers: In August, Clinton had 5.4 million supporters (we’re not talking registered voters; we’re talking committed people who actually voted, so they are extremely likely voters) who refused to vote for Obama or vowed to vote for McCain in the GE. The 2000 election had a spread of 500,000 votes. The 2004 election had a spread of 3 million votes. Regardless of what Obama says, I can’t imagine that he is going to re-draw the red/blue map in some fantasy blowout/landslide. This country is still way too polarized. We will be looking, I believe, at another very close race over the same 2-3 battleground states. And if you take a close look, you will see that both McCain and Obama are now in a statistical dead heat in the majority of the battleground states.
If half of the bitter Hillary voters abstain or go with McCain, we’re talking about 2.5 million votes that Obama simply doesn’t get. That could be statistically very significant in this race, as it’s hard to imagine either candidate winning by more than 3 million votes. Even if McCain only gets 10% of these bitter folks, that’s 1.8 million votes. In a contest that could come down to a few hundred thousand votes in each battleground state, those defections could be critical. Remember: Obama’s electoral lead is premised on leads within the margin of error in almost all battleground states.
And yet NO one in the media talks about the PUMA effect!! Everyone just blithely assumes that a giant Obama lovefest is underway. And this, incidentally, seems to be irritating the PUMAs even more, as they believe that they have consistently been ignored and marginalized by the DNC, the Obama campaign and the media.
So, bottom line, I wouldn’t stake my life on the polling methodology this year. And I guarantee you those schleps betting on Intrade can’t even spell “methodology”, much less explain how it relates to the polls upon which they are basing their wagers. All they know is Gallup says Obama + 7, and Zogby says Obama +5.
I bet if you reviewed some of the wagers placed on Intrade you would find that this time last year the chance of a recession was almost nil because the sheep were going on the phony, published, flawed data issued by our government. I always look deeper, and looking deeper it was clear that the government data was just wrong because it made insane assumptions. I put the chance of a recession at 90% last year, and I’m putting the chance of a Depression at 100% next year (if there’s a Depression Index on Intrade, let me know — I could make a handsome profit). If I had followed the Intrade bunch, I’m sure I would have still been holding onto my dead weight house that has lost 250K since I dumped it last year, because the sheep thought in 2006 that the market was going up, up, UP!!
Lastly, I think it’s a mistake to assume that increased turnout only benefits Democrats. The GOP is a stubborn bunch, a quiet bunch, and they may be operating more out of fear of an Obama presidency than any *hope* that this country can truly be fixed. And in my experience, fear is often a bigger motivator than hope. Like the bitter Hillary voters in West Virginia who showed up and cast their votes by a 40% margin for Hillary, even as she was being urged to drop out, you just have to wonder if the GOP will mobilize and vote on principle for the same reason: because they refuse to be told that the choice is out of their hands.
Am I predicting a McCain victory? Of course not. I am predicting a hard-fought Obama win or a remotely-possible McCain upset. Gonna be close.
But an Obama landslide just seems pretty far-fetched to me.
And if I’m wrong, I’ll gladly admit that I need to consider paying a lot more attention to polls 🙂
