Below is the Parlor Political Index by county for Alabama. The goal is to show some indication of a county’s underlying partisan preference.
A full explanation is below the chart.
In short, the percentages of Democrat votes for the 2002 Governor’s race and the 2006 Lt. Governor’s race are averaged for each county, and the same is done for Republican votes. Those percentages are compared to how the state voted as a whole in those two elections.
Example: In the two races of the 2002 Governor’s race and 2006 Lt. Governor’s race, Houston County voted Republican 14 percentage points higher (on average) than the state as a whole (on average).
| PPI | |
| Autauga | R+12 |
| Baldwin | R+20 |
| Barbour | D+11 |
| Bibb | D+2 |
| Blount | R+7 |
| Bullock | D+25 |
| Butler | D+5 |
| Calhoun | D+2 |
| Chambers | D+6 |
| Cherokee | D+9 |
| Chilton | R+7 |
| Choctaw | D+12 |
| Clarke | D+3 |
| Clay | R+4 |
| Cleburne | R+3 |
| Coffee | R+5 |
| Colbert | D+9 |
| Conecuh | D+9 |
| Coosa | D+11 |
| Covington | R+3 |
| Crenshaw | D+2 |
| Cullman | R+1 |
| Dale | R+9 |
| Dallas | D+18 |
| DeKalb | R+2 |
| Elmore | R+13 |
| Escambia | R+3 |
| Etowah | D+6 |
| Fayette | D+7 |
| Franklin | D+9 |
| Geneva | R+11 |
| Greene | D+33 |
| Hale | D+19 |
| Henry | D+1 |
| PPI | |
| Houston | R+14 |
| Jackson | D+8 |
| Jefferson | D+4 |
| Lamar | D+5 |
| Lauderdale | D+2 |
| Lawrence | D+14 |
| Lee | R+5 |
| Limestone | R+3 |
| Lowndes | D+26 |
| Macon | D+36 |
| Madison | R+1 |
| Marengo | D+14 |
| Marion | D+5 |
| Marshall | R+3 |
| Mobile | R+3 |
| Monroe | D+1 |
| Montgomery | D+5 |
| Morgan | R+4 |
| Perry | D+25 |
| Pickens | D+8 |
| Pike | D+1 |
| Randolph | R+1 |
| Russell | D+15 |
| St Clair | R+15 |
| Shelby | R+16 |
| Sumter | D+26 |
| Talladega | D+7 |
| Tallapoosa | R+2 |
| Tuscaloosa | D+3 |
| Walker | D+8 |
| Washington | D+7 |
| Wilcox | D+26 |
| Winston | R+5 |
- The goal is to show, relative to the whole state, how strong a county’s underlying partisan preference is by comparing how the county voted (Democrat or Republican) to how the state voted as a whole in 2002 and 2006 state elections.
- The initial idea was to look at how the state voted in the 2002 and 2006 elections for governor, and compare that to how each county voted in those elections.
- Someone (in GOP circles, fwiw) suggested that I not use the 2006 governor’s race in the measure because Riley in 2006 pulled a lot of votes (for example, from the African-American community) that do not typically go to Republicans. This person suggested that voting results from the 2006 Lt. Governor’s race might be more informative about how voters tend
to break by party. I offer the observation that this suggestion fits with the widely held notion that the Governor doesn’t have coattails; some people are voting for the Governor but are not voting for other Republicans down the ticket.
Another friend to the Parlor noted approvingly that by using the 2002 Governor’s race and the 2006 Lt. Governor’s race, four candidates are involved. This could help downplay any skew that we might get from geography. If we used both elections for governor in this measure, then perhaps Riley’s home of Clay County would skew Republican a little more than it otherwise would. The goal here is to look at the underlying partisan preference.
- Introducing… the Parlor Political Index. Here’s how it is figured.
The results from the 2002 governor’s race and the 2006 lt. governor’s race are averaged by party. The same is done for each county. If a county is scored “D+3” in the index, that means the county voted Democratic 3 percentage points higher than the state did as a whole in those two elections. “R+12” would mean that the county voted Republican 12 points higher than the state did as a whole in those two elections.
That’s it.
You may download an Excel spreadsheet that shows the calculations.
Danny – absolutely wonderful information.
Yes, very good work.
pretty sophisticated for an “internet tabloid” – seriously thanks for the info
Good job and very informative, thanks for putting in the work. The counties of the 2nd sure have an interesting total…
[...] Baldwin R+20 Introducing the Parlor Political Index [...]
Good job, Danny; however, I think those high-profile races that you chose to index have many factors other than party affiliation, such as personality, and the effectiveness of their advertising campaigns (who can forget Folsom’s “I never played tennis at the Mountainbrook Country Club”? )
When I’m interested in gauging the political party leanings of a county or a precinct, I use one of the more obscure state wide races, non-incumbent, male vs male or female vs female, where neither are hometown folks. I think those races give the best indication because in the absence of other information about the candidates, party preference moves to the front as the voters deciding factor.
Good observations, Mike. I started with the idea of using the Governor’s race, well… I guess because the Cook PVI (that prompted all this) used the Presidential race. When I wanted to work this out on the state level, I figured I would also use the top race in the executive branch (and explained above how that evolved a bit).
This may well continue to evolve… I have told some folks that I liken it to batting average in baseball in that it is one measure that is a useful tool but is not the end-all. It does have the benefit of being simple and easy to understand so that folks can decide for themselves how much validity they want to put on it. I hope some people will find it as interesting as I did…
Would be interesting to see something similar by Senate or House districts…