Obama’s current numbers are his best case
Undecideds will break heavily for McCain
An interesting memo from the McCain campaign this morning confirms what I have suspected for the last couple of weeks: mainly, that Obama’s polling numbers reflect all that he can expect to get on election day. Not a lot of time for me to completely explain the analysis here, so here’s the quick-and-dirty version.
The campaign’s memo cites as support for this the historic difference between projected African-American vote and the actual African-American vote on election day: that support for democrats is typically under-reflected in polls prior to election day. Obama is polling in the very high 90s among African-Americans, making it very unlikely that his support is understated there.
What’s more, I firmly believe that most of the “undecideds” aren’t really that undecided and are actually McCain voters who tell the pollsters that they are undecided for any number of reasons: they may not trust pollsters, they believe that their vote is none of anyone’s business, or they feel their vote is not the “popular” choice and so they pretend to be undecided. Doubtful that any member of the Obamanation has trouble telling pollsters who they are for (they’re probably waiting by the phone, hoping for a pollster to call so they can pump up Barry’s numbers and sing his praises). I think it is also doubtful at this point in the process, less than a week out and after months of campaigning and $600 million in expenditures by Obama, that there are really 5-8% of the nation that hasn’t made up their mind. The number of “true” undecideds has got to be much, much smaller. Among these, if Obama (who, in one sense, has the easiest job of convincing undecideds or “soft” Republicans to come to his cause because of the current negative mood toward the current administration and government in general, yet, in another sense, the hardest job of making the sell because of his inexperience, etc.) hasn’t convinced them yet, it is doubtful he will. They would have already bitten on the “change” platform or have reconciled their doubts about his experience. This leads me to believe that even those “true” undecideds will break heavily for the most “comfortable” of the two candidates: McCain.
If you believe that most undecideds will swing for McCain, you realize that Obama’s polling numbers are his probable best-case-scenario on election day and not very likely to be better than they are now come Nov. 4th. Thus, any state in which Obama is currently less than 50% is in serious play. Factor in current trends showing McCain on the rise and with 6 more days for those trends to play out, I think McCain’s team has it exactly right when they say that this will be a very, very close race come next Tuesday.
UPDATE: Dick Morris makes a similar point.
UPDATE II: Dave in Florida makes the same point and includes the analysis that I did not have the time to breakdown.