State of the Race – September 15 (And Where We Were This Time in 2008)

Last week I was bemoan-ing the lack of new polls in the states. That isn't the case this week. There have been 40 polls conducted in 20 different states, including in 8 of the 11 battleground states I identified last week. Of those battleground states I identified last week, the only three that we don’t have polling data on as of this week are IA, NV & WI. So with that new data, here's where we stand halfway through September:

	Safe	Likely	Lean	Total
Obama	158	83	85	326
Romney	144	47	15	206
Tossup				6

If you want to see a visual breakdown of where states fall, see the photo above or click here for a map. Obviously, the darker the shade of blue or red, then the safer that state is for that candidate. If you want to see how I arrived at these categories, I would encourage you to review the first “State of the Race” post I did on August 11 which can be found here.

Recently, the national polling has indicated a slight break towards President Obama and this is showing in the state-by-state polls as well. The most significant move in the past week was Florida moving from “Tossup” to “Lean Obama.” Additionally, Michigan and New Hampshire moved from “Lean Obama” to “Likely Obama.” Finally, Montana has been moved from “Safe Romney” to “Likely Romney.” This leaves Iowa as the only state still in the Tossup category and it is actually predicted at exactly a 50/50 proposition, although it should be noted that it is one of the states that has not had any state-by-state polling done this month. At this point the map continues to favor the President, but what does that mean going forward and how much movement can we expect?

Well, September 17, 2008 was the first time I did this analysis using the data from 270towin.com, so let's take a look at where things stood in mid-September 2008 versus where they stand in mid-September 2012. I reviewed the spreadsheet I had done on September 17, 2008 and here was the breakdown using the same categories and criteria that I've been using for my 2012 analysis.

	Safe	Likely	Lean	Total
Obama	126	70	59	255
McCain	106	110	31	247
Tossup				36

For a visual break-down, you can check out the photo to the right that shows the map as it stood in mid-September 2008. (Note: the total on the map reflects state electoral votes as they stand in 2012 where the chart above reflects where they were in 2008 before the 2010 census.) As I looked at the spreadsheet as it stood on that date, I find a few things remarkable. First, as I looked at McCain's states where he had at least a 90% likelihood of winning, those were all the states that he won. All the states below that threshold went to Obama – meaning that the demarcation line for the states was pretty well-established. In other words, states weren't leapfrogging over one another to move to the other side. What does this mean for 2012? It means that the likelihood that states are going to switch sides can be predicted by the other states on the continuum. For example, since Pennsylvania is my original home state, let's take a look at the likelihood it would switch and go for Governor Romney. The chances that it would go for Romney are directly related to all the states that are in Obama's column currently but are less likely to stay there. So if Pennsylvania goes for Governor Romney, we can also expect states like Florida, Michigan, Ohio and Virginia to go for Romney as well.

There was another thing I found remarkable as well and that is how spread out the states were across my spreadsheet. There were several states that were in each category and there appeared to be more of a likelihood of movement within the states. This was obviously seen over the next month and a half leading up to the election. The spreadsheets I've been doing for the 2012 election have never looked like that. There is certainly a cluster of states that are closer to the center, but the states on the edge have been numerous and have stayed fairly consistent. There's been a lot of talk in this election about polarization of politics in America and I think this is a reflection of that trend. This data also supports the commentators who believe that the majority of people have already made up their minds about this election and there isn't much that could happen that would change their minds at this point. Whether this is true remains to be seen.

This brings me to my final point and my final reflection on the 2008 data: there was quite a bit of movement in the 2008 election from mid-September until Election Day. Obviously, the financial crisis that hit in the second half September had a lot to do with that along with Senator McCain's reaction. Plus as I posted before, I believe the debates played a large role as well. All this means that a lot can happen between now and November 6^th and all those things could very well make the map change dramatically. This includes the debates that will be coming up and it is also making me follow the situation in the Middle East and how both President Obama and Governor Romney react to that evolving situation. In other words, no matter how rigid the electorate looks now, all of that could change in the blink of an eye – which is why I will keep monitoring the trends as we proceed through the campaign calendar.