Dave Wasserman last night on Twitter interpreted Ohio early vote totals by county to make the argument that early vote turnout is higher in McCain counties in 2012 while turnout in Obama counties is flat or lower than 2008, and this indicates an enthusiasm gap favoring the GOP in the state. Though we will know in a few hours whether he is right or wrong, I thought I would write a short diary on my experience with this kind of prediction. In 2010, I tried to do a similar analysis on Texas early voting trends and make the argument that there was no Democratic enthusiasm gap. I, of course, was way off.
In my analysis, I tried to use publicly available turnout information which is provided by the Texas Secretary of State to make trendlines comparing 2010 turnout by county to 2006 turnout by county. My finding was that in 2010 there was no drop off from 2006 in Obama counties compared to McCain counties, and 2006 being a good year for Democrats, that was evidence against a large pro-GOP enthusiasm gap.
This did not work.
Here you can see a comparison of the increase in early vote turnout by county compared to Obama performance in the county:
Even though turnout after early voting had closed was 156% of 2006 turnout in Obama counties and 147% in McCain counties, this obviously had no predictive power.
You can see the data I used here:
Texas EV (Google spreadsheet)
I would caution Dave Wasserman and others from inferring too much from county trends, though this election will soon give us another set of data on how useful they are.