Harris County: Where Are the Swing Voters?

I'm adding some new micro-cluster definitions to the mix and throwing in a combination of results to see how each category voted. The additions from this post are as follows:

African-American
(Southside: 236, 237, 573; Acres Homes: 7, 401, 784; Fifth Ward: 169, 197, 327)

New Stuff - identifying where there is an increase in new home development over the past 4-6 years
(Northwest: 124, 143, 709; Spring: 245, 551; Southside: 336, 630; West: 463)

A little extra precaution is warranted in comparing 2004 precincts to 2008 precincts - particularly in the "New Stuff" precincts. I did my best to match them up, but I could probably stand to re-check them before I do any posting on the 2004-to-2008 electoral shifts.

Here's a handful of results for each group. The intent here is to show the high mark (Garcia), the Presidential numbers (guess!), and some of the losing candidates.

The hope here is to maybe uncover a few ideas as to why there's a disparity in the results of each candidate. On with the numbers ...

Anglo GOP

Obama ....... 6,624 - 21,068  (23.9%)

Garcia ...... 7,358 - 19,441  (27.5%)

Bradford .... 5,371 - 21,368  (20.1%)

Mahendru .... 4,828 - 21,183  (18.6%)

Murray ...... 4,903 - 21,135  (18.8%)

New Stuff

Obama ....... 13,130 - 17,343  (43.1%)

Garcia ...... 14,022 - 15,768  (47.1%)

Bradford .... 12,415 - 17,063  (42.1%)

Mahendru .... 11,685 - 17,358  (40.2%)

Murray ...... 12,042 - 17,030  (41.4%)

African-American

Obama ....... 9,458 - 191  (98.0%)

Garcia ...... 9,370 - 163  (98.3%)

Bradford .... 9,341 - 191  (98.0%)

Mahendru .... 9,181 - 240  (97.5%)

Murray ...... 9,268 - 182  (98.1%)

Hispanic

Obama ....... 7,613-3,905  (66.1%)

Garcia ...... 8,813-2,597  (77.2%)

Bradford .... 7,597-3,547  (68.2%)

Mahendru .... 7,397-3,533  (67.7%)

Murray ...... 7,563-3,409  (68.9%)

Apartments

Obama ....... 8,695-4,648  (65.2%)

Garcia ...... 8,763-4,197  (67.6%)

Bradford .... 8,164-4,739  (63.3%)

Mahendru .... 7,931-4,760  (62.5%)

Murray ...... 8,009-4,678  (63.1%)

And here's where we dissect some of the urban legends from reality ...

1. Can we please stop trashing Hispanic Democrats? They vote for black candidates too, ya know.

PDiddie speculated the following before diving into the numbers:

In fact I would submit (and for the record, I have not seen the data; I'm positing the following based on what little I do know) that African-American Democrats voted for Hispanics on the ballot, but Hispanic Democrats may not have returned the favor to black Dems. More telling, "what's in a name" had as much to do with who won and lost as party label, the hundreds of thousands of dollars spent on teevee advertising, or anything else. And that naturally would not be due strictly to just Democrats and their biases.

Not to single out PDid, it's just that it's a common assumption that I've heard countless times. Right up there with "Hispanics don't vote." UGH! Nevermind that the two assumptions (which do often go hand in hand) are more than a little contradictary.

Anywho, the math above indicates that Hispanic voters did, indeed, "return the favor." Excepting the outlier of support for Adrian Garcia (who won or tied every precinct in his City Council district, btw!). In fact, Mekisha Murray got a higher percentage of votes than C.O. Bradford, who got a higher percentage than Ashish Mahendru, who got a higher percentage than Barack Obama. Bear in mind that this view is merely from a percentage view ... there's some seriously complex calculus if you want to measure total votes and the dropoff effect for each individual race.

2. Can we now turn our attention to the Anglo GOP area? ... and instead of suggesting that "this is where the racist Democrats are" maybe think more in terms of swing voters.

There was an overall spread of nearly 9 points between Mahendru and Garcia. Say what you will about names with too many vowels from the second half of the alphabet not playing well in Memorial Villages, but the spread between, say Obama and Mahendru was a more reasonable 5.3%.

The challenge here is that there are few sets of circumstances where Democrats tend to think in terms of outreach to heavily GOP areas as part of a campaign strategy. Well ... guess where the biggest dropoff was for Democrats? Hardcore Republican areas where the 20-25% or so voters that are either Democrats or potential Democratic voters just trailed off in their support down the ballot. Clearly, this needs to change.

Taken together, here's the Mahendru-Obama spread for each cluster:

Anglo GOP ............ -5.3%

New Stuff ............ -2.9%

African-American ..... -0.5%

Hispanic ............. +1.6%

Apartments ........... -2.7%

Check your assumptions about identity politics at the door, folks ... the numbers are telling a very different story.