More evidence of an Obama bounce, other thoughts on the fall election

As Andres reports below, MSNBC and the Wall Street Journal have released a new poll showing it 47-41 Obama-McCain, very similar numbers to previous polls over the the last two weeks. Some thoughts on where the race seems to be now:

There should no longer be any doubts about Obama's general election appeal, or his ability to put together a winning coalition - four major polls now show the same thing - an African American with a funny name is clearly defeating a celebrated and universally known American war hero, who in this race is more incumbent than challenger. While we have a long way to go, consider this passage from the new MSNBC analysis, which shows Obama winning among Hispanics, women, white women, Catholics, independents and blue-collar workers:

In the head-to-head matchup, Obama leads McCain among African Americans (83-7 percent), Hispanics (62-28), women (52-33), Catholics (47-40), independents (41-36) and even blue-collar workers (47-42). Obama is also ahead among those who said they voted for Clinton in the Democratic primaries (61-19).

Yet among white men — who made up 36 percent of the electorate in the 2004 presidential election — Obama trails McCain by 20 points, 55-35 percent. “That is the reason why this election is close,” Hart notes.

In addition, McCain leads Obama among white suburban women (44-38), group which makes up about 10 percent of all voters that Hart calls “absolutely critical” for both candidates in the fall.

However, Obama has a seven-point advantage (46-39) among all white women. How important is that lead? Newhouse explains that Republican candidates always expect to win white men by a substantial margin, but it is white women that usually decide the race. “If a Republican wins among white women, we usually win that election,” he says, noting that George W. Bush carried that group in 2000 and 2004.

 

Among Hispanics McCain is showing surprising weakness Obama surprising strength - As we've written before, McCain is now 15 points net below Bush's 2004 numbers with Hispanics. This shift with this community, voting at much higher numbers than 2004, could end up swinging four states Bush won to the Democrats - CO, FL, NM, NV - and perhaps making AZ and TX competitive.

McCain is not ahead in a single state Democrats won in 2004 - New polling out this week shows Obama now leading McCain in MI and WI, two states he had previously been ahead in. One of the arguments the McCain camp has been making is that he has the ability to play on Democratic turf. NDN has long believed this argument to be more spin than reality, as the it is hard to believe that in this year of a very damanged GOP brand with a weak, wobbly candidate at its top that McCain could break the lock of the 19 states equaling 248 electoral votes Democrats have won in each of the last four elections. With the only two (MI, WI) of these 19 states McCain had been leading in moving to Obama, the race is now moving to nine states Bush won in 2000 or 2004 - CO, FL, IA, MO, NC, NH, NM, NV, VA and perhaps other states like AZ, MT and TX (see these electoral college maps NDN has produced to help visualize all this).

I am not in any way suggesting that states like MI and WI won't be contested by McCain, but the notion that there is a clear opening for him in these 19 states with 248 electoral college votes is more spin than reality. Look for the candidate time and TV ads to begin moving to these other states, which are the true 2008 battlegrounds now. For more on the emergence of a new post-Southern Strategy electoral strategy for the Democrats see our recent magazine article, a 50-Year Strategy.

We are seeing a very new electorate emerge in 2008 - Every poll has to make assumptions about the composition of the electorate to produce its results. Given the huge increase in turnout this year of African-Americans, Hispanics, Millennials and women, and an enormous shift in party ID towards the Dems, this electorate will look very different from any electorate any pollster has ever seen before. This electorate will not only be much more non-white than any electorate we've ever seen, it will be much less white and male than any electorate we've ever seen, and have fewer Republicans than we've seen since at least 1982, perhaps even 1966 (see this essay for more on the growing power of minorities in American politics).

It will also be important to not overstate the role of independents in the race. In the hyper-partisan era of Rove and Bush the number of unaffiliated voters has dropped, with independents only making up 26 percent of the electorate in the last two elections, significantly down from the general rule of thumb of one-third D/R/I. Bush has created more partisans, and from 2004 to 2006, more votes shifted in the two parties than it did among independents, as partisans now outnumber independents by 3:1. Democrats owed their victory in 2006 more to what happened with the two party's partisans than what happened with independents.

Thus, one of the key numbers to watch this year is how well each candidate is doing with their own partisans. If Democrats continue to outnumber Republicans in the electorate by 10-15 points as they do now, it will be just as important for Obama to keep 92-95 percent share of Democrats as it is for him to win independents. And the same is true for McCain. If his weakness with the GOP base causes his share with his own partisans to drop below 90 percent, he will have a very, very hard time winning this election. See this post for more on the declining clout of independents.

Remember that Kerry won both independents and moderates in 2004 and still lost. Bush did much better with his partisans than Kerry did with his. For both Obama and McCain, their partisans will outnumber independents in their own coaliton by at least 3 to 1, and thus at least as much attention needs to be paid to these voters as the media's holy grail of the independents. Thus, in this election I think you will be seeing much more attention being paid to each party's base - for different reasons - than in past elections. Given the intensity and much more highly networked Democratic base, this is a big, big problem for McCain.

One question pollsters should start asking this cycle should involve the likelihood to take an action on behalf of a candidate. My guess is that Obama supporters are twice as likely to do something for their man than McCain supporters, which in this networked age when a supporter can do so much more than ever before, could become a huge differentiator in the fall election.

Overall, in our much more partisan and networked age, when the barrier to enter into politics has been so lowered, the partisans in both camps have become much more important than they were in late 20th century politics. There are more of them then before, and with all the new tools, there is much more they can do to help their candidate - money, advocacy to their social networks and neighborhoods, voting.

Thursday Update - Gallup has a look at all this today. While the number of independents in their analysis is the traditional one-third, they look at what the Party ID shift means for the fall. In their current polling Obama is only getting 78 percent of Democrats. Kerry got 89 percent. If nothing else changes in the election if Obama simply matches Kerry's number with Democrats - as one would expect he would - he will win with 51-52 percent of the vote.

Thurs 415pm Update - Josh Marshall posts a new WI poll showing double digit leads for Obama in a head to head with McCain and and the Dems in Party ID. Further evidence that the McCain "playing on Democratic turf" argument isn't holding up very well.

Thurs 6pm Update - Chris Cillizza has own take on the Party ID shift and the collapse of the GOP brand.

Fri am Update - This am EJ Dionne examines the Party ID gap and the new emerging dynamic in this year's election.