The results are in (or are they?)
Following up on yesterday’s post—if electability is the sort of thing that floats your boat, then you absolutely should take a gander at the latest CNN/Opinion Research poll. Yes, it’s a national poll, so all appropriate caveats apply, but the results are striking all the same:
As interesting to me as the four first-place finishes for Edwards are, I am also intrigued by the margins—in three of the four races, Senator Edwards does better than Senators Clinton and Obama by statistically significant margins. The difference between the three is within the three-point margin of error only in the match-up with Rudy Giuliani (all three Democrats would beat Giuliani, according to the poll).
Senator John McCain is the strongest of the Republicans in this poll—fighting Clinton and Obama to a statistical draw—but, as noted, even he falls significantly short in the head-to-head with Edwards.
I was going to beat this drum a bit more, but Ian Welsh has already done a really fine job of it, so let me just cut to him:
Sorry to quote at such great length, but I wanted to make sure you all read that middle part—yes, that part.
While I am not going to tell you to avoid Clinton or Obama based on sex or race—because I don’t think you should—if you are talking about electability, I think Ian makes a fair point about the biases that still lurk in the dark recesses of the American psyche.
It’s called the Bradley effect, and I am old enough and grew up in the right part of the country to remember whence it got its name. The short read: popular LA Mayor Tom Bradley, and African American Democrat, ran for governor of California in 1982 against the white Republican George Deukmajian. The polls consistently showed that Bradley would win, but he did not. Post election analysis revealed that, when polled, white voters over-reported their willingness to vote for a black man.
And that is far from the only example of this effect.
Do I like that in 2007/8 America we still have to talk about this? Hell no! Do I think that there is still a statistically significant percentage of the electorate that will tell a pollster that a minority or woman would be just fine with them, but when in the privacy of a polling booth know that there is just no way they would ever pull the lever for a non-white non-male? Sadly, yes, I do.
Again, please, do everything you can to fight against the Bradley effect—vote for the person that most appeals to you—but if you are evaluating the polls, please don’t ignore this sad paradox, either.
All of which is to say that, if we were to take the Bradley effect (or some likely anti-woman corollary) into account, the CNN poll likely underreports Edwards’s electability advantage over his two chief rivals.
I would also add that some social scientists believe that the Bradley effect is amplified when the opposing candidates are otherwise poorly differentiated. In other words, when you give the racist or sexist voters the “all things being equal” equation, it gives them the opportunity to focus on the more obvious, demographic difference.
Which is all the more reason to look for a candidate that is not looking to hedge bets or split the difference on the major issues of the day. Being tentative or triangulating just invites the bias problem to again “surprise” the pollsters. Candidates—unfair as it is—that are faced with an extra level of immutable bias should take that to heart, and change their message accordingly.
And that, to make lemonade out of lemons, could be a great thing for Democrats because it brings the core values of the party and electability closer to being the selfsame thing.
(cross-posted on The Seminal)
On the Democratic side, Edwards performs best against each of the leading Republicans. In addition to beating Huckabee by 25 percent and McCain by 8 percent, the North Carolina Democrat beats Romney by 22 percentage points (59 percent to 37 percent) and Giuliani by 9 percentage points (53 percent to 44 percent).
As interesting to me as the four first-place finishes for Edwards are, I am also intrigued by the margins—in three of the four races, Senator Edwards does better than Senators Clinton and Obama by statistically significant margins. The difference between the three is within the three-point margin of error only in the match-up with Rudy Giuliani (all three Democrats would beat Giuliani, according to the poll).
Senator John McCain is the strongest of the Republicans in this poll—fighting Clinton and Obama to a statistical draw—but, as noted, even he falls significantly short in the head-to-head with Edwards.
I was going to beat this drum a bit more, but Ian Welsh has already done a really fine job of it, so let me just cut to him:
Ok, enough with the BS "let's not talk about electability" idea that seems so prominent in the left-wing blogosphere. This isn't 2004, Kerry isn't the "electable" candidate preparing to cruise himself into the ground. It's almost 2008, we've got a different crop of candidates, and the most electable of the three top candidates is Edwards. This has been clear in poll after poll, the latest of which is CNN's poll, which shows Edwards crushing Republicans.
. . . .
Edwards is also the most liberal (or progressive, if you prefer) of the three of them. Democratic primary voters are supposed to be left-leaning, but they seem to support the most centrist candidate of the three -- Hilary Clinton, the woman who won't even say she'd shut down torture without exception.
Now, as long as we're talking turkey and breaking taboos, let's say the rest of what needs to be said.
Clinton has the highest negatives of any Democratic candidate, by a large margin. She's also a woman. Everyone plays up how that's an advantage, and sure, Americans claim they'd vote for a woman. But there's a well known polling bias on such social issues: people don't want to say they're sexist on the phone, but we all know sexism hasn't gone away. Some of Clinton's theoretical support in a general election is probably phantom popularity. It might only be a few percent, but given she already has razor thin margins against many Republicans, that could be the difference between victory and ignomious [sic] defeat.
And then there's Obama. Bill Clinton wasn't America's first Black President. Obama, on the other hand, would like to be. I fully expect a chunk of Obama's support would simply evaporate at the polling station, because a lot of Americans, no matter what they say, aren't voting for a black man. Shoot the messenger if you choose, but everything I know about America tells me America is still riddled with racism.
Edwards is male, southern and telegenic. He has run a populist campaign. He is probably as left wing as someone can be in the US and still run for President. He has been a friend to unions and to the poor. He has had the guts to admit he was wrong on the war and while his anti-war platform isn't as strong as I'd like (he should commit to pull out) it's better than Clinton's or Obama's.
He's electable. Of the big 3 candidates he's the most progressive.
. . . .
This isn't 2004. Voting your beliefs (the poor and middle class are getting screwed) and choosing the most electable candidate aren't in opposition to each other this time.
So what I'm asking Democratic primary voters is to take a good hard look at Edwards again. Stop accepting the media's narrative of Edwards as "the number 3 guy". Look at the numbers, look at his positions and realize that this time you can have it all -- you can have a progressive candidate and you can have a nominee who will absolutely wipe the floor with the Republicans.
Vote your heart, but by all means also vote electability. And don't let political correctness blind you to political realities. Because the country simply cannot afford another 4 years with a Republican president.
Sorry to quote at such great length, but I wanted to make sure you all read that middle part—yes, that part.
While I am not going to tell you to avoid Clinton or Obama based on sex or race—because I don’t think you should—if you are talking about electability, I think Ian makes a fair point about the biases that still lurk in the dark recesses of the American psyche.
It’s called the Bradley effect, and I am old enough and grew up in the right part of the country to remember whence it got its name. The short read: popular LA Mayor Tom Bradley, and African American Democrat, ran for governor of California in 1982 against the white Republican George Deukmajian. The polls consistently showed that Bradley would win, but he did not. Post election analysis revealed that, when polled, white voters over-reported their willingness to vote for a black man.
And that is far from the only example of this effect.
Do I like that in 2007/8 America we still have to talk about this? Hell no! Do I think that there is still a statistically significant percentage of the electorate that will tell a pollster that a minority or woman would be just fine with them, but when in the privacy of a polling booth know that there is just no way they would ever pull the lever for a non-white non-male? Sadly, yes, I do.
Again, please, do everything you can to fight against the Bradley effect—vote for the person that most appeals to you—but if you are evaluating the polls, please don’t ignore this sad paradox, either.
All of which is to say that, if we were to take the Bradley effect (or some likely anti-woman corollary) into account, the CNN poll likely underreports Edwards’s electability advantage over his two chief rivals.
I would also add that some social scientists believe that the Bradley effect is amplified when the opposing candidates are otherwise poorly differentiated. In other words, when you give the racist or sexist voters the “all things being equal” equation, it gives them the opportunity to focus on the more obvious, demographic difference.
Which is all the more reason to look for a candidate that is not looking to hedge bets or split the difference on the major issues of the day. Being tentative or triangulating just invites the bias problem to again “surprise” the pollsters. Candidates—unfair as it is—that are faced with an extra level of immutable bias should take that to heart, and change their message accordingly.
And that, to make lemonade out of lemons, could be a great thing for Democrats because it brings the core values of the party and electability closer to being the selfsame thing.
(cross-posted on The Seminal)
Labels: 2008 election, Barack Obama, electability, Hillary Clinton, Ian Welsh, John Edwards, the Bradley effect
0 Comments:
Post a Comment
<< Home