US Election: Electoral maths justify Obama's short odds
Paul Krishnamurty explains the electoral college and mathematical mountain facing Mitt Romney...
"Were Romney to lose either Virginia or Colorado, he would need Ohio plus at least one more upset."
Earlier this week, I explored the strange contradiction that shows Mitt Romney ahead in most national polls, yet a [3.1] outsider to become the Next President. Nothing has changed in the meantime, but it's worth looking at the numbers that are driving the odds. The more one examines the electoral college and the swing states that determine it, the more the markets make sense. If Barack Obama wins enough key states, he could very feasibly lose the popular vote yet remain President.
Nothing, of course, is by any means settled. Several of these swing states remain on a knife-edge and just a small shift in opinion or differential turnout could produce either a Romney Presidency or an Obama landslide. Consider the following scenarios:
The electoral college
The Presidency is decided by an electoral college, with each state awarded a number of votes that varies depending on their size - i.e.California 55, Texas 38, South Dakota 3. In order to win, each candidate needs 270. According to the latest Realclearpolitics electoral map, 39states are rated either solid, likely or leaning to either candidate, leaving 11 'toss-up' states. To illustrate the current state of play, assume there are no upsets amongst the first 39, which would mean a running total of Obama 201, Romney 191, before the toss-ups are decided.
Current betting forecasts the following results in the toss-up states
Obama wins Ohio (18 votes), Nevada (6), Pennsylvania (20), Iowa (6), New Hampshire (4), Wisconsin (10), Michigan (16) = 281
Romney wins Colorado (9), Virginia (13), North Carolina (15), Florida (29) = 257
If all 11 toss-ups are won by the current favourite, Obama would win 281 to 257
It is unlikely, of course, that every favourite will deliver although even some of our 11 toss-ups are looking pretty one-sided. So let's reduce our list to the six states where the favourite is currently no shorter than [1.4]. If we award Nevada, Pennsylvania, Wisconsin, Michigan and North Carolina to their favourites, our new running total is Obama 253. Romney 206, leaving the following six tight contests.
Ohio - Obama [1.45], Romney [3.1]
Iowa - Obama [1.52], Romney [3.1]
New Hampshire - Obama [1.58], Romney [2.58]
Florida - Romney [1.5], Obama [2.9]
Colorado - Romney [1.88], Obama [2.04]
Virginia - Romney [1.94], Obama [2.0]
Here we see the scale of Romney's task. Without a big upset somewhere else, Obama is just 17 electoral college votes short of victory and favourite in three of six remaining states. Ohio alone takes him to the winning target. Victory in all six would mean a landslide by a margin of 332 to 206, comfortably beating the 90.5 vote handicap, which is currently a [4.0] chance.
Alternatively, were Ohio to be the only upset, Romney would become President, with 275 to Obama's 263. The betting, however, suggests Romney has a mountain to climb in Ohio and is no position to focus all his attention on the state, given that he only marginally holds favouritism in two of his four toss-ups. Were Romney to lose either Virginia or Colorado, he would need Ohio plus at least one more upset. Were he to lose Florida - where the RCP poll average predicts a small 1.8% advantage - it is almost certainly game over.
With so many more plausible routes, it is easy to see why Obama is such a hot favourite. Nonetheless, before steaming into short odds-on bets on the basis of these numbers, beware that polls are often wrong, margins are tight and big upsets are frequent in politics.