Is the Mitt Romney campaign beyond the point of being able to win? That is the narrative that is being repeated in the media right now. We decided to independently assess his chances by using publicly available polling data. The reason we are using poll data is that is the best metric (before any votes are counted on Election Night) to use.

More specifically, we think it is a reasonable basis of comparison to use the average of poll data from the last week of the 2008 campaign and compare it against the last 15 days of polling data as of the time of this article. Why are we using 15 days? At this point in the game, people are just starting to tune in, so events are not moving as fast as they will be once the debates are underway (the first Presidential debate is on October 3).

Furthermore, this is an election which, realistically, is only competitive in a handful of states. Therefore, we are limiting our discussion to those states we think will not be securely for Romney or Obama for the duration of the campaign. If you look at the 2008 and 2012 data, you will notice the following:

2008 vs 2012 comparison

(1)    From the graphic, it is apparent (and ironic) that the states Romney most needs to win are the states where the “Obama plunge” is minimal. This is likely due to heavy advertising by the Obama campaign in those states, plus the fact that there are few “swing” voters in these states, given the massive amounts of advertising each party has done in those states;

(2)    The correct way to interpret this graph is that, as the race stands now, President Obama’s plunge in his poll standings relative to 2008 would only cost him Iowa and New Hampshire. That would enable him to emerge from the Electoral College with a 322-216 victory over Mitt Romney (only 270 electoral votes are required to win) (Note: we are assuming right now that Mitt Romney has 206 electoral votes based on our last analysis) ;

(3)    If the Romney campaign can “move the needle” (i.e., drop the Obama poll percentage) by just 2 percentage points, he can carry Colorado, Florida, and Ohio. These states would give Romney 56 more electoral votes and a 272-266 Electoral College victory;

(4)    If the Romney campaign can “move the needle” (i.e., drop the Obama poll percentage) by 3 percentage points, he can pick up Pennsylvania and Virginia. Those states would give Romney 33 more electoral votes and a 305-223 Electoral College victory;

The bottom line: while the media is ready to write the Romney campaign off, this in fact is still a close race. In fact, the national Gallup tracking poll for the past 7 days shows a 47-46% Obama lead. Hardly the numbers that justify calling the race “over.” We would like to close by noting that our use of polling data to make these predictions have the following limitations:

(1)    The media is assuming that the extraordinary levels of Democratic intensity in 2008 will occur again, and are developing their polling models on this assumption. The limited absentee voting data we have so far does not suggest that this is the case. However, the first “swing state” to begin its in person early voting will be Iowa, on September 27. At that point (since there will be far more early voters than absentee/mail in voters), we can get a better assessment of voter enthusiasm;

(2)    In 2008, Barack Obama was the challenger, while this year, he is an incumbent with a record that leaves few voters with neutral feelings about him. And given that he is the incumbent, we believe that the undecided vote (at least as measured by the last week of polling in 2008) will not split 50-50 for him like it did in 2008.