Dem in the City, Revisited

Kuff breaks down the city council district vote results in the 2008 elections. It's a more fine-tuned version of the numbers I broke down to view City vs Rest of County results here and here. Add to this, my own ancient mapping/calculating of District F after the 2002 election.

That in mind, there's a few items I think are worth tack onto Kuff's research:

» Kuff points out that the precinct boundaries don't always match perfectly to city boundaries. This is an especially acute problem when the city boundary snakes through a road to capture retail tax revenue basis, but zero homes. In other words, the precinct is listed as a partial precinct but there are zero votes ever cast from the precinct. More common in other districts are precincts along the COH boundary that are part-in and part-out, with a range of in/out allotment. This is generally less of an issue in the math since the biases tend to average out.

The snaked precinct problem is particularly strong in Dist. A and Dist. F, and I think may skew those numbers quite a bit more than the rest. District F snakes down Highway 6, while goes all over FM1960 (and parts of Highway 6). District E goes through the Ship Channel, but I'm not speculating as to the effect this has on the numbers. In the case of the other two, it has the affect of adding to the Dem numbers and could skew it beyond the point or two that you'd expect from the traditional issues of split precincts.

That said, there were only a handful of precincts in District F that voted for McCain. All were in Sharpstown and Briar Meadow. This is obviously the district I'm most familiar with, but it also seems to be the most dramatic in terms of who turns out in an even-numbered year and who turns out in an odd-numbered year.

» This should be an obvious point and is left out of Kuff's analysis. Presidential year electorates aren't the same as off-year electorates. Wealthier, whiter homeowners are likelier to vote in 2009 than the poorer, minority, apartment-dwellers that showed up in droves in 2008. That reality - to me, at least - explains Kuff's point about Republican overperformance in district elections. I don't know that I'd say they overperform. I would say that there's a disconnect that the more strident candidates seem to not have much of a solution for. On the one hand, the makeup of the electorate levels the playing field for moderate-conservative candidates. But the issues attached to local elections don't fit into the nice, neat cubbyholes of Republican talking points. I believe this is the lesson that should be learned from Orlando Sanchez, who hewed to those points rather well in 2003, only to watch his support wane from 2001 levels despite having an open seat contest.

That said, there is an evident shift in several wealthy, Anglo precincts. A good case in point here is District C. The numbers Kuff posts seem in line with my belief that inner-urban wealthy residents are gaining in acceptance of Democratic candidates. One precinct really sums this up: 0222. Obama beat McCain 1040 to 797 in this very-upscale precinct. Closer to home: Skelly over Culberson 1068 to 723; and Ellen Cohen over Joe Agris 127 to 617. How that translates into local politics is tough to gauge. It's not a given that a candidate could waltz in with a credibly funded campaign peddling standard Democratic talking points and expect to win. This is ground zero for modern swing voters and the issues they move on aren't quite cut along a perfect ideological outline.

On the whole, I think it's important to realize the difference in Houston & non-Houston numbers as exhibiting a reasonable difference given the methodologies used. I've been tossing around 2% as the difference when including 0-voter and partial precincts since before I ever did my own work and it turns out that that's about the degree of difference over the entire city.

Neither approach really addresses the turnout issue, but there does exist a means to measure things like Apartment vs Homeowner turnout in different elections. Whoever's got the most free time can tackle that issue next.