Two-way partisan registration by county, 2014-2015

[Adam Griffin]

I'm nearly "done" with this. This examines voter registration statistics by state and county, breaking down the totals and comparing only Democratic & Republican registration totals against each other.

In some states (such as KY, where unaffiliated registered voters are less than 5% of the voting pool) there isn't much loss in reducing to the two-way model. In other states (like MA, where 53% of registered voters are unaffiliated), there is a dramatic loss of voters in the analysis.

The unfinished New England states all have data, but they are broken down by town instead of by city (ugh). I'll fix those sooner or later. The other light-grayed-out states either have temporary issues with SoS (SD) or do not have this data available online at all (UT, AR). The dark gray states do not have partisan registration.

You can definitely see old trends - on both sides - throughout various states. What surprises you the most?

Full-size image

[Adam Griffin]

I have completed everything except the 5 trouble-making states (New England + UT, along with SD which is currently unavailable).

What year and from where did you pull that data? I've noticed some differences between yours and mine. I used January 2015 data for OK.

[Adam Griffin]

What surprises you the most?

Iowa and to a lesser extent Florida.

Iowa shocked me a lot, too, if only because of what I saw initially before making the map: Republicans only have a 3,000-voter advantage statewide among active voters (what was used in this particular instance); Democrats have a ~15,000-voter advantage among active and inactive. To be fair, once you start going county-by-county, there are a lot of really close R majorities. Iowa is also insanely rural throughout most of the state, so probably not that surprising.