11 January, 2009

Voting Stuff...


In the theme of recent posts, here's a graph of Obama's vote share vs. turnout and a quadratic fit to the data with a 90% confidence interval. What I thought was interesting was that the peak of turnout was not in the states that were split evenly, which is what you might expect if voters with the best shot of affecting the election are more likely to vote. But the maximum predicted turnout actually occurs at 55.5%.

2 comments:

Elliot said...

It looks like what matters is the contestedness of the states by the campaigns, rather than the end result, which would make a lot of sense. In the middle there you have WI, MN, IA, OH, PA, CO, NV, NM and I think VA and FL but I'm not sure cause its scrunched. Not all of those were close, but they were the major targets of the campaigns' resources.

spencer said...

Also NH is up there. Although PA, NV, and NM actually have pretty low turnout relative to the rest of the states, so I'm not sure how that pans out. Do you have data on per capita spending by the campaign?