This piece in TWS is fascinating. What we have is a bimodal distribution in the polling; and bimodal by polling organizations. There are essentially two different groups of professionals looking at two very different electorates. One group says it’ll be more like 2010, the other that it’ll be more like 2008.
In my 30 years as a political junkie I’ve never seen a scenario in which both sides seem to firmly believe they’re going to win – maybe even going away. National polls point to a Romney win, some states’ polling point to a possible electoral college win for Obama. Typically the national polls would have to be closer for there to be a chance of a national/electoral college split decision. So either someone’s wrong or the polls will converge in the final 5 days.
Unless it’s 1876 again: a 4% popular vote win that lost the electoral college. I read someplace yesterday that the consensus is the GOP stole that election, but only after the Dems had stolen it first via suppression of the black vote. The particulars of the 1876 corrupt scenario don’t seem likely to repeat this cycle.
My best guess (and my bias – I am a Republican) is that the pollsters who are actually measuring the specific variable (party ID or sentiment) with a dedicated poll and then comparing it to benchmark historical data are more likely to be accurate than the pollsters who tweak their LV screen and make some assumptions about the mix of the electorate.
My math says that for Obama to win he’d have to (1) do better at gotv than last time, a perfect storm for him; while at the same time (2) the GOP would have to show no improvement in gotv; and (3) undecideds would have to break more heavily for the incumbent.
I dunno. Strikes me as unlikely. I know GOP turnout will be higher for Romney than it was for McCain and maybe as high as the wave of 2010. I suspect matching the messianic excitement of 2008 will prove to be a near impossible task for Obama. And I think the small number of remaining undecideds will break 2:1 for Romney. So it ends up something like 52% and 295 electoral votes.
But who can be sure? I could see it tipping either way, with outcomes ranging from narrow BHO win to solid Romney victory. We’ll know in a few days. Sure is interesting. Some pollsters will have egg on their face, and one or the other side’s afterparty is going to be interesting.