Is New Math the Future of the Census?

» Washington Post: Census chief Robert Groves: We’ve got to stop counting like this

Robert Groves, on his way out the door at the Census …

“Because of the constitution, the country will always have a census,” he said in an interview Friday at his office in the bureau’s Suitland headquarters, already stripped of his personal belongings. “But how we do the census and surveys will have to change.”

Cost is a big reason. Even though it came in $1.9 billion under budget, the last census cost $13 billion, about $42 a head. The pricetag has doubled every decade since 1970.

So there’s talk of relying more and more on private databases. At the basic level, there’s something to be said for that. But once you get beyond household utilities and tax records, I’m curious where that trend leads to.

A Quickie Early Vote Aggrepost

Since I’m up early and a few links seem worth a mention …

» New Republic: Achilles Tar Heel
North Carolina and Virginia are definitely going to be some of the most interesting states to see how the Presidential race moves (or doesn’t).

» Chron: Galveston dropping resistance to public housing
» GC Daily News: Isle officials briefed on GLO’s housing plan

I’m not one for denigrating the wisdom of voters, even if my preferred candidate loses. But Galveston voters aren’t helping much with their choice of mayor in the last election.

» Wash. Post: Baltimore puts out welcome mat for immigrants, hoping to stop population decline
There are actually a number of tenents to Baltimore’s efforts. Not harrassing undocumented residents is one of them. If I were the scorecard-carrying type, I might bemoan this sort of coverage as “biased.” Instead, it’s just incomplete and a little lazy. There is a quote from a Baltimore elected who does begin to introduce some of the additional context, the reporter just ends the story there for all intents and purposes. The city of Baltimore has apparently been focusing on population for quite a while and the latest Census info didn’t exactly do any wonders. Here’s the Baltimore Sun’s topic page for “population decline” for more context.

» NY Review of Books: Getting Away With It (Paul Krugman & Robin Wells)
Set aside for reading on Wednesday or any insomnia attack before then. I downloaded a preview of the Scheiber and Edsall books. Edsall’s is a definite for continuing on with and I may opt for Chris Hayes’ latest instead of Scheiber’s. For his review, I’m merely hoping that Krugman is better suited for elaboration in a long-form take rather than his usual spotty analysis in his blog or column. That said, I wasn’t much of a fan of his book, “The Conscience of a Liberal.”

» Foreign Affairs: Confucius and the Ballot Box [$]
This looks interesting for a little dose of continuing education in Asian politics. So it goes into the Wednesday reading pile. I’m assuming this will relate more to the impact that a lack of democracy has to American foreign policy. But I’ll be reading for some extra hints on any meaning for Asian voters (and more importantly, non-voters) here at home.

And on a non-political, non-current-affairs sort of note, I’ll offer a sermon of choice …

» Mars Hill Church: A Church That Believes in “We” and “Opt-In”
With a little bit of gratuitous back-patting over catching up on some Mars Hill podcasts, I happened to spin a few old sermons in order of their delivery. So now I’m on a kick to read through the book of Acts (and possibly restart this old habit), which this sermon is part of a series on. Sadly, I realized that my spotty record of downloading sermons finds me 7 downloads short of the 24-part series on Acts. So a small sacrifice of remuneration to the altar of archived podcasts must be given.

ADD-ON: One late entry via facebook discovery …

» Texas Monthly: Why Johnny Can’t Learn
The book under review definitely seems like a worthwhile read if I ever remember to pick up Ravitch’s last one to read as a preview.

The Over/Under Extremes for the State House

Below are the most over-populated and under-populated State Rep districts throughout Texas. Not surprisingly, they track with some of the known overall growth patterns: high growth in Collin County, north of Plano; high growth in the Katy area and northerwestern (Cypress) area of Harris County. The slow-growth area still being in the inner-city cores.

What’s interesting here is that, among the redistricting arguments you will here is that future growth patterns are an allowable reason for population variation. Districts 70 and 132 started the decade pretty close to the maximum deviation for underpopulation, which really puts an exclamation point behind them being the top two most over-populated districts.

What’s interesting, though, is that some of the slow-growth areas also started the decade underpopulated. In the case of District 22, that may have been because expanding the district out would have made it harder to maintain over 50% African-American population. As-is, that district cuts into Orange County, which is a pretty unusual county slice. But Anchia’s District 103 nearly maxes out the under-represenation limit and would not have had any problem adding population while still staying close to the 66% VAP Hispanic population it started off with. In fact, only one of the listed under-populated districts started the decade off as over-populated.

On a sidenote, 28 districts are within the allowed +/-5% range for state population.

Most Overpopulated
HD70 – Ken Paxton (McKinney, Collin County)
2010 population: 300,801 – overpopulated by 133,164
2000 population: 132,671 – underpopulated by 6,341 (4.56%)

HD132 – Bill Callegari (West Harris County)
2010 population: 264,426 – overpopulated by 96,789
2000 population: 133,149 – underpopulated by 5,863 (4.22%)

HD28 – John Zerwas (Fort Bend, Waller, Wharton)
2010 population: 263,682 – overpopulated by 96,045
2000 population: 139,748 – overpopulated by 736 (0.53%)

HD89 – Jodie Laubenberg (Collin, Rockwall)
2010 population: 253,976 – overpopulated by 86,339
2000 population: 134,550 – underpopulated by 4,462 (3.21%)

HD130 – Allen Fletcher (Northwest Harris County)
2010 population: 252,386 – overpopulated by 84,749
2000 population: 134,227 – underpopulated by 4,785 (3.44%)

Most Underpopulated
HD103 – Rafael Anchia (Dallas)
2010 population: 117,346 – underpopulated by 50,291
2000 population: 132,316 – underpopulated by 6,696 (4.82%)

HD22 – Joe Deshotel (Jefferson, Orange)
2010 population: 126,184 – underpopulated by 41,453
2000 population: 133,159 – underpopulated by 5,853 (4.21%)

HD114 – Will Hartnett (Dallas)
2010 population: 126,576 – underpopulated by 41,061
2000 population: 141,298 – overpopulated by 2,286 (1.64%)

HD143 – Ana Hernandez (Houston)
2010 population: 127,381 – underpopulated by 40,256
2000 population: 133,869 – underpopulated by 5,143 (3.70%)

Census View of Bexar County State Reps

Yet another one to add to the “to-be-mapped” list. Bexar County’s magic number is 10.23, so they have the unusual task of keeping the same number of seats (Travis County will also maintain their current level). And with that, there’s still some pretty wild population swings that will have to be adjusted for. As I’m sure anyone who’s familiar with the county knows, the north/northwestern part of the county grew like wildfire.

So here’s what jumps out to me:

– Two majority-Anglo districts. The surprising part, though, is that both represent fairly slim majorities, yet each district has an outsized share of GOP vote within them. Again … sometimes demographics alone don’t tell the full story.

– The two most over-populated districts are right by each other. It could be interesting to see if that means there might be an opening for former State Rep. David Liebowitz to return in 2012. And, not surprisingly, the two most under-populated districts are side-by-side. One of those members is the Vice-Chair of the Redistricting Committee. It’ll be interesting to see if that works in Villareal’s favor, or if there’s a clear divide within the committee like there was in 2001. I wouldn’t expect it to be an excuse to pair the two incumbents, but you never know.

Usual caveat: as noted in the Harris and Dallas County version of this post, +/- is based on the countywide average for number of seats while the overall state map will require a +/- 5% in terms of population variation statewide.

   DISTRICT       +/-   Population  Hisp%     Asian%   Afram%    Anglo%
116-Fischer     -28,533   142,944   69.14%    4.03%    4.74%    20.49%
117-Garza        48,883   220,360   63.04%    2.35%    5.31%    27.34%
118-Farias      -18,668   152,809   68.18%    1.02%    3.03%    26.48%
119-Gutierrez   -14,371   157,106   71.12%    0.79%    7.15%    19.61%
120-McClendon    -8,290   163,187   50.85%    1.82%   25.62%    19.39%
121-Straus      -12,604   158,873   35.40%    2.45%    6.75%    53.23%
122-Larson       75,369   246,846   32.35%    4.29%    3.67%    57.59%
123-Villarreal  -39,035   132,442   74.05%    0.85%    3.15%    20.86%
124-Menéndez      6,567   178,044   69.88%    1.76%    6.51%    19.81%
125-Castro       -9,315   162,162   68.91%    2.52%    4.17%    22.84%


Census View of Dallas County State Reps

Another one that I’ll get around to mapping out later in the weekend. Dallas County’s magic number is 14.1, so there’s no path that I see to avoid them losing two seats. And with the exception of HD109, every district essentially has to grow a little to add population. Dallas should be a pretty clear cut case of watching the GOP devour one of their own, barring retirement by any of the members.

A couple of things jump out from the demographics:

– Linda Harper-Brown’s current district is now much more evident as to why it’s become so competitive. That 17% Asian population isn’t going un-noticed, but I’m curious what the voter rolls suggest in the district. It’s close to DFW airport and there’s a sizable stewardessflight attendant population in the apartments nearby. How many of them are registered voters is a bigger question than anything else.

– Like Harris County, Dallas County has a “black-to-brown” problem with some over-extended African-American districts carrying a higher share of Hispanic voters now and with African-American populations under 40%.

– Comparing the demographics of HD102 and HD107 strikes me as interesting since the early voting results in 2010 had 102’s Carol Kent being the lowest-running incumbent Dem and I figured that 107’s Allen Vaught might hang on. In the end, Kent performed a mere point less than Vaught. But the demographics are a little more uphill in Vaught’s 107. That may or may not be an indicator that he had a little better crossover support, it might not mean much of anything. But it’s at least a interesting datapoint. Certainly worth watching to see what district (and what kind of district) Vaught falls in for 2012.

As noted in the Harris County version of this post, +/- is based on the countywide average for number of seats while the overall state map will require a +/- 5% in terms of population variation statewide.

   DISTRICT       +/-   Population  Hisp%     Asian%   Afram%    Anglo%
100-Johnson    -20,119   149,033    46.51%    1.41%   37.61%    13.31%
101-Burkett     -5,551   163,601    32.70%    3.43%   21.42%    40.30%
102-Carter     -37,825   131,327    35.58%    5.67%   19.08%    37.85%
103-Anchia     -51,806   117,346    74.54%    2.17%    7.15%    15.23%
104-Alonzo     -37,252   131,900    81.40%    0.66%    7.54%     9.61%
105-Harper-Brn  -4,914   164,238    35.67%   16.83%   13.87%    31.10%

106-Anderson    -9,436   159,716    49.93%    4.91%   13.22%    30.02%
107-Sheets     -28,695   140,457    36.77%    2.35%   10.60%    48.67%
108-Branch     -25,621   143,531    23.61%    4.04%    6.67%    63.87%
109-Giddings     6,103   175,255    17.69%    0.78%   63.66%    16.19%
110-Caraway    -18,449   150,703    47.52%    0.19%   38.17%    13.02%

111-Davis       -5,778   163,374    35.51%    1.47%   46.72%    14.98%
112-Button     -20,241   148,911    23.09%   13.69%   11.00%    49.83%
113-Driver      -7,849   161,303    30.02%    7.23%   15.86%    44.76%
114-Harnett    -42,576   126,576    28.76%    4.36%   14.20%    50.80%
115-Jackson    -28,284   140,868    26.89%    9.24%    7.50%    54.27%

Census View of Harris County State Reps

I’ll get around to mapping this out for a visual overview of the situation. But, for now, here’s the math-heavy view of things. The demographics are interesting enough, but if you want to see whether districts need to shrink or expand, the +/- number is the key. I calculated the county population divided by 24 to get an average of 170,519.

A couple of caveats … first, there’s no telling if there might be some magic trick to allow Harris County to get 25 seats. I have a hard time seeing it, so I’m operating under the assumption that we get 24 for the sake of planning ahead. Second, the entire state map has to contain districts that are +/- 5% in terms of population variation which will be calculated against the statewide average – not the county average. That number is 167,637. So, at maximum, a district can contain 176,018 people and a minimum of 159,255. All that to say: I use the countywide average as a simple baseline, but it’s not the magical number that the Justice Department will be looking at.

That aside, some of the highlights that jump out at me:
– Not one district in Harris County presently has a majority of African-American population. The decade started with four districts that had a bare majority and two that were between 45-50%. In fact, Harold Dutton and Garnet Coleman’s districts are under 40% African-American. For all the critics of the VRA, I can’t help but notice a positive in there. Of course, there’s a point to be made that the share of African-American protected seats is inflated. Harris County is 18.4% African-American. Out of 24 seats, that would warrant 4.4 seats. Yet we have 6 African-American seats. Will the GOP lege look at that as an opportunity to put one of those seats on the cutting-room floor? There’s a lot of interesting math at work in many of these districts, so I’ll probably get around to a full blog post devoted solely to this topic.

– My district – HD137 – now has almost 64% Hispanic population. It was originally drawn to at 55% Hispanic. That almost evenly matches the decline in Anglo population. Throughout the county, 10 seats are majority-Hispanic. 5 others are plurality-Hispanic.

– Six districts now have over 10% Asian population in Harris County. More growth in the Asian population happened outside of city limits than within. Welcome to the new world of minority population diffusion!

Disitrict         +/-   Population  Hisp%     Asian%   Afram%    Anglo%
126-Harless     1,755   172,274    29.58%    9.23%    20.88%    38.33%
127-Huberty    16,583   187,102    23.74%    2.09%    11.00%    61.33%
128-Smith     -21,702   148,817    37.53%    1.45%    10.00%    49.47%
129-Davis     -19,721   150,798    20.26%   10.62%     7.72%    58.90%
130-Fletcher   81,867   252,386    22.10%    7.07%     9.16%    59.54%

131-Allen     -17,630   152,889    41.48%    3.78%    45.36%     8.17%
132-Callegari  93,907   264,426    37.46%    7.10%    13.72%    39.53%
133-Murphy    -15,223   155,296    36.61%   11.78%    23.23%    26.34%
134-Davis     -23,373   147,146    12.73%   10.78%     4.61%    69.69%
135-Elkins     -3,582   166,937    39.91%   11.41%    12.51%    34.31%

136-Woolley   -23,665   146,854    23.63%    8.82%     6.66%    58.88%
137-Hochberg  -32,643   137,876    63.75%   10.57%    12.76%    11.55%
138-Bohac     -33,638   136,881    51.41%    4.24%     9.22%    33.80%
139-Turner    -19,600   150,919    44.22%    4.02%    44.27%     6.42%
140-Walle     -31,244   139,275    83.98%    0.56%     6.71%     8.24%

141-Thompson   14,201   184,720    43.56%    1.58%    40.12%    13.33%
142-Dutton    -15,725   154,794    45.21%    1.60%    39.22%    12.79%
143-Hernandez -43,138   127,381    85.66%    0.28%     2.26%    11.22%
144-Legler       -804   169,715    56.04%    4.80%     7.23%    30.68%
145-Alvarado  -37,789   132,730    85.89%    1.75%     3.18%     8.61%

146-Miles     -27,399   143,120    22.62%    7.60%    46.81%    21.39%
147-Coleman   -23,662   146,857    38.50%    3.67%    37.45%    18.81%
148-Farrar    -29,573   140,946    59.29%    1.79%     8.42%    29.38%
149-Vo           -683   169,836    33.12%   18.31%    23.22%    23.26%
150-Riddle     41,965   212,484    25.99%    4.71%    16.49%    50.46%


First Pass: Harris/Fort Bend Census Numbers

The 2010 State of Houston-area Demographics …

             Harris             Ft. Bend       
Total     4,092,459            585,375               
Anglo     1,349,646  32.98%    211,680  36.16% 
Afr-Am      754,258  18.43%    123,267  21.06%
Asian       249,853   6.11%     98,762  16.87% 
Hispanic  1,671,540  40.84%    138,967  23.74%

             Houston            Sugar Land
Total     2,099,451            78,817
Anglo       537,901  25.62%    35,014  44.42%
Afr-Am      485,956  23.15%     5,744   7.29%
Asian       124,859   5.95%    27,672  35.11%
Hispanic    919,668  43.81%     8,324  10.56%


The fact that Houston clocks in at 2.099M instead of 2.1M warrants some attention. I’d think we might go ahead and draw 11 districts anyways, but the fact that there’s a differing datapoint may very well mean that any party who doesn’t like the end product map will have something to argue in front of a judge regardless of which way we go. I’m spending more time in front of numbers than I am news … hopefully there’s a quick statement from Mayor Parker on which route we go.

As for Harris County, I’ll be curious if the VAP numbers track with Hispanics being the plurality in the county. If so, I think it makes it very difficult to argue for three safe Anglo GOP seats and zero Hispanic opportunity seats. Either of the two proposals I drafted strike me as being worth some consideration.

UPDATE: Via Council Member Ed Gonzalez, the city is proceeding with plans to add two new council seats despite the Census showing us a hair short of the 2.1M population count. Good to see that!

Early Numbers from the Census

The Census puts out some prelim numbers while I await the raw data. As expected, growth tracks with the suburbs. Of more immediate political importance is the fact that the Harris County number for State Representative seats clocks in at 24.4126. I knew I should’ve bet Doc Murray on that! Unfortunately, it means we’re now likely to lose a State House seat in Harris County. Dallas County’s number is 14.1266 – as expected, they’ll lose a seat. Tarrant County’s number is 10.79 – as expected, they’ll gain a seat.

Texas in 2010: Majority Minority

» SA Express-News: 2010 Census: State increasingly urban, Latino

The big-picture numbers are out …

The iconic image of Texas is fading further into history, as new data from the 2010 Census shows that the state is increasingly urban and Latino.

According to the Associated Press, Latino’s accounted for two-thirds of the state’s growth over the last decade.

Latinos now make up 38 percent of the population, up from 32 percent in 2000, according to the AP.

The Anglo population has now dropped to 45.3 percent, down from 52.4 percent in 2000. About 11.5 percent of the population is black, unchanged from 10 years ago.

Almost all of the state’s population growth is expected to have occurred in just four areas: Houston-Galveston, Dallas-Fort Worth, the Austin-San Antonio corridor and the lower Rio Grande Valley.

I’ll be in the weeds of those particular details all weekend long.

Before & After: 2000 I35 Region Population Density

One final map product before the 2010 numbers land in my lap. This is the “before” picture of the I35 Corridor, which is expected to have substantial population gains over the last decade. Zoom out for the bigger picture or scroll around for the other counties in the area.

As I get numbers together, I’ll ultimately be looking at where population in each section grew (plus, I’m hoping to do a South Texas map completed over the weekend) and what demographic groups accounted for that change. Because, ya know … that’s how I roll.

– If you prefer a bigger map, here ya go.
– If you want to look at the file in Google Earth, here’s that.

dark green – 5,000 people per square mile
green – 2,500 people per square mile
light green – 500 people per square mile

Light Blogging Ahead

Census data may become available as early as tomorrow. Even if it’s not, I’m still backlogged with a lot of pre-work that goes into getting ready for the analysis work that will follow from having the new numbers. So I’ll basically be in Excel/mySQL/PHP/Google API hell through Monday.

Should some portion of the world blow up between then and now, I trust you’ll get the news elsewhere. But when you want to see income differences broken up by Census block groups, you know where to find me. Expect some population density and demographic maps to be posted as soon as possible. But the more “in-the-weeds” type of data will wait until I’ve gone through things a bit more comprehensively.

Before & After: 2000 Greater DFW Population Density

As a second act to the Houston population density map from the 2000 Census, the DFW Metroplex version is below.

As far as what to anticipate, I’m obviously not as deep into the weeds here as I am with the greater Houston area. But what should make this region more interesting is that the population in Dallas County is nowhere near keeping up with the average growth in surrounding counties. So while the current map has a lot of the high-density areas in Dallas, it’ll be interesting to see where some of the mid-range density census tracts start to move up.

One easy target to look at is anywhere around McKinney and Frisco in Collin County. Already, tract 30503 (western McKinney) has grown from a total population of 11,868 in 2000 to a count of 52,441 as of the 2005-09 ACS data. Massive development projects have a way of doing that. Needless to say, expect that number to be higher when the 2010 data is out.

– If you prefer a bigger map, here ya go.
– If you want to look at the file in Google Earth, here’s that.

dark green – 5,000 people per square mile
green – 2,500 people per square mile
light green – 500 people per square mile