When I was in college I had a rather eccentric roommate who, every year around this time, would lose himself in an arcane ritual known to his friends as ?Bowl Calculus.? The enormous whiteboard upon which we had scribbled a half-semester's worth of timelines, circuit diagrams, equations, and outlines would be wiped clean and - in a fit of productivity ? he would descend into a caffeine fueled haze of rushing yards, coaches' polls, rankings, revenue models, and injury reports. I, who never really followed football, merely mourned the loss of my whiteboard, but when all was said and done he would emerge with a prediction of what schools were going to which bowls in a just a few months.
The actual Bowl Championship Series was never all that far off from the elaborate calculation though what specific edge or advantage it gave him was never terribly clear to me. What my roommate did with college football, however, pundits, pollsters, and political junkies do with the vast sums of data that pour out of the presidential election. The advantage gained from that is very clear indeed.
Political predictions offer a glimpse into state of the race as it is and transform the presidential election from a single moment into a political spectator sport that spans months. Analysts and pundits compete to ensure that their model is the most accurate, hoping to attract viewers, gain influence, and establish themselves as a go-to source for insight into the political process.
Most analysts anyway.
The flip side of political prediction is that it can be self-fulfilling. As Politico writers John Harris and Jim Vandehei commented:
A candidate who is perceived to be doing well tends to get even more positive coverage (about his or her big crowds or the latest favorable polls or whatever). And a candidate who is perceived to be doing poorly tends to have all events viewed through this prism.
In short, predictions can influence coverage, create the impression of momentum, and shift the dynamic of entire campaigns. A candidate who is perceived as doing well not only gets more positive coverage, but that coverage often translates into his doing well. It is a cycle that most candidates find extraordinarily hard to break once they find themselves on the wrong side of it... as John McCain is.
Barack Obama has secured sufficient leads in Pennsylvania, Virginia, and Michigan to push his campaign over the top by most estimates. Indeed every major electoral prediction now forecasts a win for the Illinois Senator come November 4.
Almost.
To be sure, there are a wide range of predictions which are further complicated by how votes are awarded. Electoral-Vote.com, which marks states as strong, weak, barely, and tied, is the most optimistic for the Illinois Senator, awarding him 306 electoral votes between its ?strong? and ?weak? categories. Politico, in contrast, forecasts the smallest number of votes for Obama of the mainstream predictions, figuring the Democratic nominee for 234 electoral votes before the swing states enter into things. Other sites run the gamut between the two but the average across Electoral-Vote, Politico, Real Clear Politics, CNN, MSNBC, and the New York Times places 282.5 electoral votes in Obama's column compared to 160.33 in McCain's.
And then there is Fox News. Consensus be damned, Fox has its own take on the electoral spread which has no resembelance whatsoever to the numbers generated from the polling and predicting community at large. Fox figures Obama for just 183 electoral votes, 51 fewer than the very-cautious Politico was willing to presume. Simultaneously, Fox predicts 189 votes for McCain, 15 more than CNN awarded in McCain's most optimistic mainstream forecast. Another glance at those numbers should reveal something else. Unlike every other major electoral map, Fox's actually has McCain ahead by 6 electoral votes.
The average of the other six analyses shows Obama ahead by 122 votes.
There is, however, nothing inherently wrong with coming to an independent conclusion. That Fox's analysis does not line up with the mainstream conclusions is, in and of itself, unremarkable. What is remarkable is how Fox arrives at those conclusions. Among the states Fox feels comfortable ?calling? for McCain are Indiana and North Carolina. An average of polls taken throughout the election shows McCain up by half a point in Indiana and down by a third of a point in North Carolina.1
The closest state awarded to Obama under this metric is Washington State, where the Democrat leads by 12 full percentage points. Curiously absent from Obama's column is Colorado (+5.7), Florida (+2.5), Iowa (+10.8), Michigan (+9.0), Minnesota (+9.1), Missouri (+0.9), New Hampshire (+7.3), New Mexico (+7.5), Nevada (+3.2), Ohio (+3.9), Oregon (+13.3), Pennsylvania (+9.3), Virginia (+5.7), Wisconsin (+10), and of course North Carolina where it is Obama who leads McCain.
Obviously Obama is not going to win every state where the average polling favors him but if Fox is prepared to award North Carolina, where McCain trails by 0.3% to the Republicans surely Oregon and Iowa, where the Democrats hold a double digit lead belong in Obama's column.
Indeed Fox's election map may be the single most significant contribution the network is able to make to the McCain campaign in the final days before the November 4 election. As this column asserted earlier this week, the perception that the race is already run is among the surest ways to crush voter turnout. With media outlets kicking around phrases like ?historic landslide,? Fox's more conservative viewership is unlikely to exhibit much enthusiasm on election day. By creating the perception that the race is close, Fox will likely boost turnout rates amongst its audience and that may be enough to tip the balance for McCain in some key swing sates.
Fox's bias is almost a tautology in political circles and something that is known and accepted by many millions of Americans. It remains rare, however, for the network to so far overstep the bounds of journalistic integrity that their bias is not just evident, but mathematically proved.
Facebook
Twitter