Or is this still fundamentally a tight race?
Two weeks ago, I wrote an article that said “forbidding a landslide (ie a victory in the popular vote of more than 5%), the 2008 map is unlikely to look significantly different to the maps from 2000 and 2004”. I still think that is true, but the last two weeks have seen a marked movement in both the polls and the betting markets towards Barack Obama. Real Clear Politics (fresh from a little flamewar with Nate Silver at FiveThirtyEight.com) has Obama up by 5.7% at time of writing, InTrade has seen McCain drop below 30% briefly in the last 24 hours, and Betfair is pricing Obama and McCain at 1.36 and 4.0 (last traded price) respectively.
Now to be fair to John McCain, there are scenarios that could plausibly see him win from this defecit. A scandal about Obama, an Israeli attack on Iranian nuclear facilities, a Bradley effect, or two knock-out debate victories – any one of these would easily reduce a sub-6% lead to rubble, but most punters and pollsters seem to be accustomising themselves to the idea that the lead Obama currently enjoys could just as easily grow into double figures as be overwhelmed by the Republican.
The Electoral College magnifies (and sometimes distorts) a candidate’s lead in the popular vote. With the caveat of Ross Perot’s influence, Bill Clinton only beat George HW Bush by 5.7%, and yet took 370 Electoral College votes in 1992. An 8.5% lead over Bob Dole in 1996 saw him win 379 ECVs to Dole’s 159. If McCain’s campaign unwinds, given it already faces a fundraising defecit and the best Democratic GOTV movement in the party’s history, might we see Obama drastically exceed the expectations of the Spreadbetting markets such as SportingIndex that currently offer a Buy-Sell price of 313-307?
The difference between a 5% lead and an 8% lead nationally would likely hand to Obama states that I had considered likely to go for McCain. Colorado and Nevada might be decicive if the national polls are within 2%, but with leads of greater than 6% nationally, then Big States like Virginia, North Carolina, Florida, Ohio, and even Georgia could elect the Democratic candidate. These six states alone are worth 91 Electoral College votes. Double-digit leads would see Missouri, Indiana, and even Texas look vulnerable – another 56 ECVs.
My essential thesis is that the difference between a 4% lead and a 10% lead (i.e. the extremes of a 7% lead with 3% MoE) could be almost 150 Electoral College votes. For those with the nerve to play the spreadbetting markets, a buy-price that sluggishly follows the polls might prove good value. With two more Presidential Debates focussing on domestic policy to come, the question becomes will Obama seek to finish in style, or coast to a safe but moderate victory?
UK POLLING NEWS (17:20): The News of the World is carrying an ICM poll of 192 Labour-held marginals in tomorrow’s paper, showing that the Conservatives would win a healthy majority. The voting intention published of this sub-set of constituencies is Con(43%)-Lab(34%)-LD(15%). Full results here, and the newspaper is reporting the poll here.