Virginia Polls: Romney Leads, But It May Not Matter if Obama Wins Ohio


Last Week, I wrote two articles which were published on PolicyMic; one about the Virgina polls and one about the Ohio polls. In those articles I pointed out that Virginia was all but locked up by Romney, and Ohio was a coin-toss which favored Romney slightly at the time but would favor President Obama as the election neared. I decided to add polls to the analysis I made for both of those states; in doing so I added four Ohio polls (The Ohio Poll, PPP, CNN/ORC, and Columbus Dispatch) and one Virginia poll (PPP). The results changed, which was expected, but the relative impact of the addition of the new polls didn't alter the basic premise of every election that party affiliated turnout differential is what wins elections, Although one could say that winning the crossover and/or independent voter battle is crucial, it is not always necessary.

In the newly added Ohio polls there is a significant adjustment in independent voters going towards Obama, while Romney's numbers stay relatively the same. President Obama only moved from slight favorite at near optimal Democratic turnout levels of D+6 or better to slight favorite, or better at D+4 turnout or higher. Romney lost his edge in the D+4 and D+3 turnout models, but Romney was more of a favorite four days ago than Obama is now at D+4  models. The expected small swing towards Obama is what I said would happen and would make him a slight favorite in the Ohio race. Romney can still win with D+5 models although those chances are slimmer than he had before, and the same goes for Obama at D+2 models which didn't exist before for Obama. This is why both campaigns are still optimistic about Ohio as both camps lean towards their own optimal turnout models to their own peril.

Despite the change in model computations the PPP Virginia poll provides, it doesn't move the needle enough to change the expected outcome of the Virginia race. It shifts the needle by a half a point in President Obama's direction, but is that enough? No, not unless turnout differential is near optimal for President Obama. It just lessens the probability of Romney winning. Capturing a significant portion of crossovers is a sign of bipartisan reach, but that's not evident in this election and why capturing independent voters is crucial it doesn't ensure victory.