With two months to go, can we say anything useful about who will win the 2012 presidential election? Examination of polling and previous results leads me to the following prediction: Even if most of the close states break for Gov. Romney, President Obama should be narrowly reelected with between 275 and 290 electoral votes (against the 270 needed to win).
There are, of course, many caveats on such a prediction, two of them being that this may not fully reflect any post-convention “bounce” that Obama received nor any gain from the fallout of Romney’s clumsy remarks following the murder of American diplomats in Libya.
Battleground states
America elects a president state by state, with the winner gaining all the electoral votes of the state. That system is widely thought to favor rural areas (more precisely, states with a small population) because of the addition of two electoral votes to what each state is entitled to by population. Thus, while California has 66 times the population of Wyoming, it only has 18 times the electoral votes.
However, in practice, this “rural bias” seldom has much effect on the campaign because many rural states are very one-sided in their voting and therefore not worth the time of a national candidate to fight to gain just a few electoral votes. The democratic nominee is not going to try to contest for Wyoming’s three electoral votes when in 2008 the state voted 2 to 1 for the Republican candidate.
Instead, where campaigns tend to focus is on the so-called “battleground” or “swing” states: larger states where the population is evenly divided between the two major party candidates. Here, an extra effort might yield a significant gain in electoral votes.
Sometimes it is implied that these states are the same from election to election, but that is not the case. If we consider a “battleground state” to be one where the margin of victory was 5 percent or less, the only state that was in that category for the 2000, 2004 and the 2008 election was Ohio, and in 2008 it’s margin of victory was 4.5 percent.
Indeed, some 17 different states have had at least one close race in the last three elections.
A number of states (Iowa, Minnesota, Nevada, New Hampshire, New Jersey, Oregon, Pennsylvania and Wisconsin) were closer than 5 percent in 2000 and 2004, but wider than that in 2008.
What happens, of course, is that a given state is not independent of the national trend. A significant victory for the Democrats (such as 2008) pulls the margin of victory in most states toward the Democrats. That was true in Iowa, where the margin of victory was less than 1 percent in the close elections of 2000 and 2004, but nearly 10 percent in 2008.
This is relevant, because if we think that 2012 will be a closer election than 2008, we’d expect states might revert back to long-term trends. So it would be reasonable to Iowa to be much closer this time around.
Three groups of states
In analyzing the race, it is useful to divide the election into three groups of states: those where the Democrat is well ahead, those where the Republican is well ahead, and the list of battleground states in between those marks.
As of early September, polling data for all states was not publicly available. However, as you might expect, it is available for states thought to be close. In states where there is no or little polling data available, it is not very difficult to predict a winner.
Starting with the Republicans, Gov. Romney is comfortably ahead in 23 states with 191 electoral votes, with the biggest states being Texas and Georgia. The only states where Obama might have a chance is in Georgia, where Polster.com has him just three points back in early September, and in Missouri where the meltdown over Republican senate candidate Akin’s “legitimate rape” remark might sink the national ticket. However, Obama was eight points back in early September. The rest of those 23 states seem to be reliably for Romney.
Obama is ahead in 20 states and the District of Columbia that total to 247 electoral votes, the biggest of these being California and New York.
This 247 votes is not quite as secure as Romney’s 193. It includes (unlike Romney’s total) several states that were battlegrounds in previous elections. This includes both states, such as Pennsylvania and Minnesota, which did vote Democrat and Nevada that voted Republican in previous elections. However, Obama is at least seven points ahead currently in Pennsylvania and Minnesota and five in Nevada.
Battleground states of 2012
We are left with seven states totaling 100 electoral votes that look to be the place where the 2012 election will be decided: Florida, Ohio, North Carolina, Virginia, Wisconsin, Colorado and Iowa.
Polster.com has Obama with very narrow leads in five of the seven. Despite the national narrative of the campaign moving in Obama’s direction over the past two weeks, these states have shown no consistent movement toward bigger leads for Obama.
The mathematics of the race do favor Obama. Win Ohio and he needs only one of the other five to get past 270. Furthermore, Obama did win all seven of these states in 2008, but in some cases (North Carolina, Ohio and Florida) by slim margins.
Predicting
It was Neils Bohr, perhaps, and not Yogi Berra, who first said, “Prediction is difficult, especially about the future,” but the election seems closer, and more difficult to predict, than one might expect.
On the one hand, it is entirely possible that a 2 or 3 percent shift nationally will translate into several of those battleground states being solidly for Obama and thus make his re-election sure. On the other, it is also possible that the new wave of rioting in the Middle East will be spun by the media as a failure by Obama and thus effect a 1 or 2 percent shift away from him, significantly damaging his chances.
Another significant unknown in early September is just how well Romney “wears” with the public as they have more and more encounters with him. Most on the Democratic side are hoping that familiarity with Romney and Ryan will lower their favorability.
Stay tuned.