Today’s Rasmussen poll is the first time since Obama grabbed the Democratic nomination more than 2 months ago that John McCain has held any type of lead in either of the two tracking polls. It’s a 1% margin (47% to 46%) and only after leaners are included (it’s a 44% tie before), so this is a purely symbolic and statistically insignificant development - just as Gallup’s national poll was last week. But in the world of perception-driven and momentum-infused campaigns, it is a development nonetheless. It comes after a week in which McCain launched negative ads and in which Obama’s post-trip bounce faded in statistically significant margins in both tracking polls. And it provides a data point for media reports trying to assess the impact of McCain’s attacks.
In more tangibly worrisome news for Democrats, today’s Wall Street Journal discusses the Bradley effect and whether Obama’s support might be overstated in polls because of his race. The WSJ interviews respected Democratic pollster Peter Hart who “estimates that 10% of current Democrats and independents who say they support presumed [Obama] may not be giving a fully honest answer, at least based on their responses to broader questions about race.” The article goes through mechanisms polling groups are putting in place to monitor any racial bias, including recording the interviewer’s race and asking race-related questions that aren’t related to the presidential election. I was certainly not aware that such precautions were being taken, and if the Bradley effect does indeed play a role this year it could impact exit polls.
As many of you probably remember, the Bradley effect was thrown in the discussion in the hours after the New Hampshire primary as an explanation for the polls’ collective mistake. I immediately explained that made very little sense and future primaries confirmed that as there never was a consistent discrepancy between polls and results in Clinton’s favor: Obama’s results were worse than the final polls hinted at in California, New Jersey and Connecticut, but they were better in Indiana, Missouri, Virginia, Maryland, Oregon, Connecticut and North Carolina - and there was little difference in Texas and Pennsylvania.
None of this is to say that the general election will have the same dynamics as the Democratic primary. It is true that voters who might be ashamed to tell a pollster that they will not be voting for Obama because they are afraid of looking racist are more likely to be Democrats or independents (otherwise why would they fear the interviewer might think they are shunning Obama out of racial bias), but that does not mean that their behavior in the general election will be the same as that in their primary, in which Democratic voters unwilling to support Obama were voting for another Democrat and thus would be less likely to be fear that their act is transparently racist. Furthermore, the universe of registered Democrats voting in a general election is much larger than the primary electorate.
Meanwhile, a number of presidential state polls were released:
- In Arizona, PPP releases the first poll since June and finds McCain leading in his home state 52% to 40%.
- Respondents were also asked about a potential 2010 Senate match-up between McCain and Governor Napolitano. McCain led 50% to 43% (he trailed in such a poll taken in the fall of 2007).
- In Florida (polling history), a poll taken by a Republican polling outfit McLaughlin & Associates for the Florida Chamber of Commerce shows McCain ahead 45% to 40%.
- In New York (polling history), the Siena poll has Obama ahead 44% to 26%, an improvement since the previous poll.
- Oklahoma, finally, does not suffer the fate of other Western red states as McCain remains very strong, leading Obama 56% to 24% in the latest Sooner poll.
There has been a lot of speculation lately that McCain could be forced to defend his home-state because of the demographic trends, the conservative base’s distrust towards their Senator and the fact that McCain’s showing here on Super Tuesday was disappointing. This past Friday, a poll conducted in Arizona’s biggest county (which accounts for half of the state’s population) found McCain’s lead down to 5% - 10% short of Bush’s margin in 2004. However, PPP’s poll should help reassure Republicans. The best Democrats can hope for is that McCain feels compelled to spend time and money here. But as long as the Republican is ahead by double-digits, it is unlikely either candidate will invest any resources in Arizona.
The PPP poll is skewed towards Repubs, look at the breakdown, and there were 8% more Repub. than Dems, with only a 100,000 voter registration difference, this does not equate to a 8% difference.
Logan,
The 2004 exit poll had Republicans making up 44% of the electorate and Democrats making up 30%, so 8% is already a significant swing towards Democrats. (It’s true that the GOP advantage in registration is about 4%, not 8%, but self-identification is not necessarily the same as registration status.)
It is true that Bradley effect could be in effect, and if it is it will be very detrimental to the Obama campaign, and exit polls will be against him. However, I’m not sure that it will have that much of an effect for several reasons. First of all, it seems that many voters are able to use other reasons on why they are NOT supporting Obama which could be cover racist sentiments, like they are not voting for Obama becaues he isn’t experienced enough or is too risky. This is probably a good excuse for Clinton voters who may have racist reasons but are afraid to state it. I would say that extremely few voters are racist and won’t actually vote for Obama but tell pollsters that they actually will because of being afraid as being seen as racist.
The second reason is that many African Americans and young people (Obama’s most powerful groups) are un-registered or not likely voters, and this gives Obama an opportunity. If he can really make huge surges in the black and youth turnout, it will likely be an invisible increase for Obama that can cancel out the Bradley effect.
And of course, the WSJ strongly indentifies with the GOP (even without Murdoch) They are very good and fair when it comes to buisness news but in the political area they are conservative (even without Murdoch) and so the editors have an incentive to make Democrats worried by talking about a Bradly effect that would affect Obama adversly.
On the state polls, it is apparent that Obama is weakening, which is why he is now putting in a pushback so that he is at least neutral in the eyes of voters again. Of course McCain would be ahead in Florida if the polling firm is GOP affliated. Yes, both Oklahoma and New York are safe states, no suprise here. Maybe a little worrying that Obama is under 50% in New York, but an 18 point lead is standard for Democrats in the state. And of course, while Oklahoma leans Democratic at the state level, at the federal level it is overwhelming Republican.
On Arizona specfically, it is no surprise that McCain is diogn better. As I have said before, if McCain was not from Arizona, it would be a lean GOP state, not a safe state, but favorite son status alone is probably enough to keep Arizona for McCain. Obama could take the state if he really wanted to, but it would be a waste of money, and therfore no one is going to spend alot in Arizona. One interesting aspect is that McCain is ahead of Napolitano, considering that if McCain was running for reelection, it would be because he lost the Presidency, and something like that would probably dampen enthusiasm for him greatly. Unlike John Kerry in the 2008 Senate race who is facing token opposition, if McCain was facing Napolitano he would probably be in the race of his life, maybe even enough that he decides to retire rather than run for reelection in the Senate after a dissapointing presidential loss.
Have you Republicans and skeptics forgotten the “Doug Wilder Effect”?
I am a bit worried about the Bradley effect, and accusations about the race card could expand it. However, I agree with Jaxx that most people who won’t admit that race is a factor already are saying they won’t vote for because it’s risky or something.
No big mac, the Bradley effect was said to influence Doug Wilder’s election too. Wilder won the 1989 VA gubernatorial election, but the margin was very small-only half a % point. Most polls leading up to the election had him leading by 5-10 points.
Yeah Joe is right, the Bradely effect means that the amount of support that a black candiate gets is lower in the actual election than it is in polls because people are not willing to admit racial cocerns and choose the black candidate to be safe, even through in reality they won’t pick him or her. The Bradely effect doesn’t mean that the candidate will flat out lose an election, its that if they do win, it will be by a narrower margin. Wilder was luckier than Bradely in that the effect was less and he was far enough in the polls to win the election despite racism. Looking at the polls now, Obama is close, and if thier is a Bradley affect he will lose the general election big time. However there are alot of outlets that are more acceptable than racism. For example, some white voters say they won’t vote for Obama because they think he is a Mulsim or that he isn’t American enough. While these slurs are definilty disgraceful, they are (unfortuantly) not as taboo as racism, and it’s a great outlet for people who are truly uncomfortable with voting for a black man a excuse to say they are not voting for them. Therefore the Bradely affect is nonexistant or at least extremely minimal. I will say that if the Rev. Wright controversy never happened or if Obama was never the target of internet smears saying he is a radical Mulsim then the Bradley effect could very well be huge, as there would be no “acceptable” excuse for voters who have racist sentiments to express for thier disapproval of Obama. Of course Obama probably would have been much stronger if these problems didn’t show up, even includinig the increased Bradley effect, but I highly doubt that thier will be a Bradely effect with so many other (and more acceptable) slurs to use as an excuse. In the end, people do want to be able to express thier opinions as much as they can as their is a certain amount of pride in having your opinion recorded by an reputable polling firm.