Possible Bradley Effect: The biggest worry has been the persistence of the Bradley Effect in key battleground states -- particularly PA, where a Bradley Effect would negate Obama's perceived lead. Despite my confident assertion that the Bradley effect is dead, it hangs like a ghost over the polling. Notably, the fact that the Bradley Effect appears to have been diminishing over time must be balanced against the fact that African-America candidates have run in environments which are more favorable to "diversity" overall. Especially in places like Ohio and Indiana, where the demographic is less favorable overall to Obama, how much of a Bradley Effect still persists in a race that is polling fairly close?
Republican Rebound: One advantage the Ds had was the fact that the Rs were not very engaged. As the intensity has mounted, this has changed. A lot of Rs that threatened to stay home if McCain won are now asking themselves if they really mean it or if they can go out and vote for Palin and against Obama.
In other words, even excluding the race issue, how many independents that have generally voted R over the last few elections and polled as leaning Obama will, when push comes to shove, decide to give the Rs one more chance because there is enough different about McCain to make them hope for a different result than under Bush/Chenney or because they can persuade themselves that Palin will keep McCain "honest." There is a generation of voters who so automatically associate "Democrat" with "evil liberal" that despite being mad at the Rs for the economic meltdown, the war, etc., will they just find themselves unable to vote for a "liberal Democrat?" Or, more likely from my perspective, find themselves unable to stay home on election day no matter how unhappy they are with McCain and/or the Republican brand.
Embedded voter suppression: It has gotten far more ugly and obvious, but it remains a real problem -- especially where Republicans have long held control of the state voting mechanism. It is important to recognize that the nonsense over Acorn and supposed irregularities in Obama fundraising is not merely to lay the groundwork for lawsuits (which are generally ineffective) or for preemptive whining to rally the base for subsequent elections by building the mythology of how 2008 was "stolen." It is also to provide psychological immunization for increasingly open acts of illegal voter suppression (by which I do not mean merely making use of every legal means of making it more difficult for people to vote, I mean the crap like threatening potential voters with arrest). By creating the accusations that it is really the Dems that have tampered with the vote, it provides justification for ANYTHING Republicans do to "counter" the "voter fraud."
I will add that I think the Obama campaign has pretty much done everything it can to adjust for these factors and will not be taken by surprise. These are the uncontrollables that create risk. From everything I have seen, the Obama campaign has intelligently prepared for all of these, and responded with precision planning and execution unequaled in modern Democratic campaigns. It remains to be seen whether it will be enough.
Every objective measure continues to point to an Obama win. There are now, in fact, many fewer undecideds than usual on election day, and in a number of critical states Obama's margin passes the remaining undecided count (depending, of course, on who is polling). Moving beyond the polls, Obama continues to draw huge crowds of supporters and armies of volunteers in numbers far beyond McCain. There is no sign of complacency in the campaign, and the tightening of the race over the last few days is within the usual margins of previous elections. Ds continue to enjoy huge voter registration advantages. And, of course, the general environment works very much to the D advantage.
Possible "Reverse Bradley" and The "X" Factor: There has been considerable talk about whether a "Reverse Bradley" may happen. i.e., Voters in split households or in communities that are heavily Republican will say they are voting for McCain when they intend to vote Obama. This is difficult to track, but is plausible given the extreme split by age (with younger folks living at home potentially breaking heavily for Obama but not wanting to say so in front of parents) and by gender. In recent weeks, Obama has polled far higher among independent white women, the ones Palin was supposed to attract. I find it plausible that in Ohio, Indiana, and a number of other places with traditional "value voters" that Obama will do unexpectedly better and largely because of white women.
Why? Several reasons having to do with the unique nature of this campaign and the demographic in question. First, among working class households in traditionally R areas, and particularly among older voters, it is women rather than men who pay the regular bills and manage the household budget. Traditionally, these women voters have been more sensitive on pocket book issues. It is seems reasonable to me that many more women than men in these swing areas will find the question of who they trust more on the economy and health care compelling.
Second, the "Oprah" factor. A large number women in the target demographic get their opinions from Oprah, the View, the Today Show and other "chick shows" rather than from Rush and Sean Hannity. While these programs have not traditionally had significant political impact, they may very well this time. Whereas men in the target demographic may be much more likely to agree with Michael Savage that they need to get out the vote for McCain to stop the liberal agenda, women in the same demographic will give much more weight to the fact that they liked what they saw of Obama and Michelle on The View or on Ellen.
Third, in the target demographic, women are much more likely than men to avoid argument and therefore not state their real opinion to pollsters if they perceive that it is a likely to engender conflict with their spouse.
Voter turnout numbers continue to look huge. People are lining up and staying in line for early voting. Conventional wisdom holds that voter turnout favors Dems, because the traditionally Dem demographic is the one with the lowest overall vote turn out.
Many different routes to Dem victory. Dems are not tied to a single state. Dems can lose Ohio and Florida, but still win with Nevada, Colorado and Va. For McCain to win, he must limit R loses to the state of Iowa and take every other state Bush took in 2004. Alternatively, he must grab something from the Dem side of the ledger. That's a tall order, given that McCain's chance of flipping either PA or NH seem fairly small at this point, and the other states the Rs hoped would be contestable: MN, WI, and MI seem well out of reach.
One day more . . . .