Today the RCP average of polls shows +6.8% for Obama, and has been over 7% for several days, even at 8.2% a few days ago. The polls included in their current sample range from Obama +2% to +14%. Obama is at 49.5% and McCain at 42.7%.
I wonder if Obama may have hit a ceiling, at least in polling numbers. It’s a cliche that we have a polarized electorate. So, how low can a generic Republican, or a reasonably competent Republican candidate, get? 40%? Maybe. I’d argue a little higher. Recall that 40% is a McGovern or Goldwater level. Let’s say 42% as a GOP minimum (or a Dem minimum for that matter). Right now undecideds are at 7.8%. How much will that number decrease, in the polling before election day? I don’t think all that much. Obviously the undecided and ‘other candidate number will be about 1, maybe 2, percent when the ballots are cast. But, in the polling, I bet it won’t go much below 6 or 7 percent. Let’s say 6.5%.
So, if we accept these numbers, McCain might be expected to keep polling at 42% and the undecideds at 6.5%, leaving Obama with a “ceiling” of 51.5%. (I’m not positing anything negative about Obama, nor any Bradley effect, or anything other than a polarized electorate and a reasonably competent GOP candidate).
If Obama stays around 50-51% for the next 18 days, with a lead of 6-8%, I predict we will see news stories asking “What is wrong with Obama? Why can’t he break ahead of the (approximately) fifty percent mark? Why can’t he seal the deal? We saw this in the primaries against Hillary Clinton? Blah blah blah.”
One final note: In our system, an 8% popular vote lead can be expected to lead to an electoral landslide. See elections of 1988 and 1996. On theother hand a 2% popular vote lead can result in a very tight “one swing state could have made the difference” result. And since polling is an inexct science, 2% versus 8% leads can be hard to tell apart.