Swing County Project

Swing State Project is a place for advanced election returns watchers, so here’s a tool for those who want to delve a little more into the nuts and bolts of what’s happening at the county level as swing state returns come trickling in, rather than just relying on the statewide numbers. The focus is on states with competitive senate or governor’s races, but there are also a few presidential swing states thrown in as well. Bookmark it or print it for easy reference tomorrow!

What I’ve done is start with the 2004 presidential numbers and look at all the counties that comprised 1% or more of the state’s votes (so that you aren’t spending your time worrying about 50/50 counties that turn out to only have a few thousand people in them). I’ve added (or subtracted) from those numbers to find where we need to be in each county to get over 50% on a statewide basis (regardless of whether you’re looking at these states in terms of the presidential race or another close statewide race). The assumption is that each states’ counties form a pretty consistent left-to-right spectrum no matter what the race is.

A few caveats: I didn’t want to spend weeks on this project, so these numbers presume essentially the same model as 2004 and push the percentages in exactly the same way in each county. Obviously, there are going to be some differences from that. Take Colorado, for instance. We can probably bank on increased African-American turnout in Denver, and increased young voter turnout in Boulder and Ft. Collins (Larimer County), which may express itself both as a greater Dem percentage gain than other counties and as a higher percentage of the total state tally. We also should factor in that disproportionately more growth in the last four years in Colorado has occurred in red counties (especially booming Douglas and Weld Counties), meaning those counties may also be a higher percentage of the total state… but we should also account for the fact that as suburban ripples spread out into exurban counties, they also tend to get bluer. Unfortunately, you’ll need to put on your political geography expert hats and make any of those mental adjustments yourselves.

Also, for statewide (governor or senator) races, assume that there might be regional concentrations in favor of particular candidates (i.e. the part of the state they’re from and where they know everybody… for example, look for Kay Hagan to overperform in Guilford County or Gordon Smith to overperform in Umatilla County). So what happens tomorrow won’t necessarily mirror my numbers (especially since in many places we’ll be easily exceeding 50%); they aren’t projections as much as just benchmarks on where we need to be at a minimum in order to win.

Colorado

Florida