At this point in the Presidential race, there appears to be an inconsistency between what the national polls and the state by state polls are saying. The weighted average of polls taken over the last 10 days shows President Obama with a narrow 46-45% lead over Mitt Romney, while the average of state by state polls as of last Friday shows the President with the 270 electoral votes he needs to be re-elected.
To explain why/how this can happen, we looked at the 2008 popular vote, as well as available statewide polling for the last 30 days. In doing so, we made three assumptions:
(1) Voter turnout similar to 2008 (not a bullet proof assumption, given that a previous analysis we did of voter registration trends shows lessened Democratic enthusiasm this year);
(2) All of those who voted for John McCain will vote for Mitt Romney;
(3) The third party vote will be about the same as it was in 2008.
The next piece of information needed for our analysis was a calculation of the amount of the “Obama plunge.” In doing so, we compared the average of the last 30 days polling for each state for President Obama against the percentage of the vote he received in 2008.
In performing this analysis, we have noticed that, relative to 2008, there has been a fairly constant “Obama plunge”, which we have calculated as follows:
7% “plunge” in “solid Obama” states (as of 7/26/2012)
8% “plunge” in “lean Obama” states (as of 7/26/2012)
6% “plunge” in “tossup” states (as of 7/26/2012)
6% “plunge” in Republican or Republican leaning states (as of 7/26/2012)
When we apply these “plunges” on a state by state basis against the 2008 vote, we come up with a popular vote total of 46% for President Obama – exactly what he has been getting on average in recent polling.
So why does it appear that he is leading in individual state polling? It all has to do with where you believe the undecided vote will break. The news media and political pundits tend to focus only on the spread between Obama and Romney, and from that perspective, it looks like the President would be re-elected.
We have a different belief. It’s important to understand the context of the Presidential race right now. President Obama, as an incumbent President, has a record to be evaluated, while Mitt Romney is relatively unknown. Nor are voters in a hurry to decide: it’s summer time, and you have the additional distraction (from a political standpoint) of the Summer Olympics. Therefore, the fact that 9% of voters are undecided more than 13 weeks away from the November election is not particularly surprising.
Therefore, ever since we started our weekly analyses in May, we have reiterated our belief that the undecided vote is really a hidden Romney vote, since a President as polarizing as President Obama leaves few voters “on the fence.” Assuming that this is the case (i.e., 100% of the undecided vote goes to Romney), Mitt Romney would win 52-46% in the popular vote and 331-207 in the electoral vote. Here’s what the electoral map would look like under that scenario:
We are not necessarily comfortable with the idea that all of the undecided vote will go to the challenger, because you have to assume that some of the undecided consists of late breaking Democrats who will “return home” to their party by November. Therefore, we recalculated the state by state totals assuming that 75% of the undecided vote goes to Mitt Romney, and 25% stays with Obama. In that situation, Mitt Romney would win 51-48% in the popular vote and 305-233 in the electoral vote. Here’s what the electoral map would look like under that alternative scenario:
There are three things to keep in mind under either of the two scenarios we have laid out: (1) because of GOTV efforts at the Presidential, state, and local level (and because of the presence of local issues), the “Obama plunge” will not be uniform from state to state, which could result in states surprising us, like (from the Republicans viewpoint) Indiana and North Carolina did in 2008; (2) even with these projections, we found seven states (totaling between 58-68 electoral votes) in the 49-51% range, so it wouldn’t take much for these states to “swing” to either candidate on Election Day; (3) the presence of third parties (like Ralph Nader in 2000 and Libertarian Gary Johnson this year) could also enable one of the candidates to “carry” a state with less than 50% of the vote.