Greg's Opinion Greg's big blog of whatnot


Kinder Institute on Houston’s Diversity

» Chron: Houston region is now the most diverse in the U.S (Jeannie Kever)

More research on the local demographics ...

The Houston region is now the most ethnically diverse large metropolitan area in the country, surpassing New York City.

Two suburbs - Missouri City and Pearland - have become even more diverse than the city of Houston. Other suburbs aren't far behind.

These findings are from a report released Monday by Rice University researchers, based on an analysis of census data from 1990, 2000 and 2010.

"We are a little United Nations," Pearland Mayor Tom Reid said. "You go to one of our neighborhoods, and there will be a person from Nigeria living next to somebody from India, living next to somebody from Mexico and somebody from Louisiana."

The report also found that while residential segregation has dropped over the past 20 years, it remains highest within the city of Houston; most suburban neighborhoods are less racially segregated.

You can dig through the Kinder Institute report here or just take in the overview video it here:

I'm curious how some other locations track with this. I'd have to think that there would be some similarities in Los Angeles, at a minimum. Obviously, some of what leads to remnant segregation of Houstonians is that you have major parts of town fully developed and that have had multiple generations living in close proximity. So it seems logical that some areas of Third Ward and River Oaks would remain as they are. And as you build up undeveloped areas around those parts of town, the pool of buyers is generally going to be more diverse. That's why you see pockets of Hispanic neighborhoods near Katy and newer Asian home-buyers filling in the area between Alief and Sugar Land, as well as between Alief and Willowbrook Mall. The fact that there was plenty of open space to develop creates this, in other words.

How this compares to Chicago, New York, or other major cities where I'm less sure that you'll see something comparable would be something that warrants a bit more study. If only to satisfy my curiosity. That Harris County, as a whole, has gone through a rapid pace of diversification is something that obviously fits well within my wheelhouse. For a visual, there's always this time-series of demographic maps that I tend to rely on for making the argument easily understandable.

And as an additional reminder, there's this snapshot of SW Harris County done with my standard-issue demographic color-coding down to the Census block level:

This, in short, shows the pattern of multicultural blocks (yellow) outside the city, but the remnant homogeneous areas within the city (red/Anglo, brown/Hispanic, black/Afr-Am). As a point of emphasis, this demonstrates that it's not just the larger aggregates of population that are settling in different ways. Block level aggregates are generally as small as 8-12 houses. Seeing a mix of population with no distinct majority living that closely together is something we'll definitely be seeing more of in the future.


Interconnected, But Distinct

» NY Times: Asia Society Expands, East and West

Un-clogging the backlog of stuff I've been meaning to blog about during the past week or so, here's one that caught my eye. And not just because it references the new Asia Society building in Houston. There's a wonderfully valid point here about why such a thing is considered necessary ...

Some may question the need for an Asia Society in an increasingly globalized world. But the society’s executives say the institution has become more essential because it can serve as the link among various constituencies.

“The world is far more interconnected today,” Mr. Chan said. “Hence the need to understand each other is greater than ever before.”

At the day job, we've lugged various presentations to a number of groups to talk about and demonstrate why multicultural population growth requires some new thinking in terms of how people communicate with their target audiences. The solution is never to assume uniformity (even though there can certainly be overlapping similarities in different markets). But the challenge is how to differentiate effectively when you're confronted by diversity that's far more fragmented today than it was before.

Asian culture is a key in the learning experience for that, since there are so many distinct cultures and nationalities that often require such differentiation. But its not just Asian populations anymore. Anyways, just article is an interesting datapoint somewhere in the vicinity of the topic. Read it and check out Houston's facility when it opens.


Addendum from an 80-20 PAC Presentation

To those of you who might have witnessed a presentation given at the Houston 80-20 PAC dinner Sunday night, the links below should help send you deeper down the rabbit hole of maps and demographics. It's always a pleasure to hang with one of my favorite local political groups and the interest in demographics by several of the folks at 80-20 events is always a relief. Enjoy the further reading and if there are other areas of interest that you'd like me to look into, I'm usually game for learning a little more about my favorite part of the world.

- This post on Asian population by Houston-area neighborhood is the basis for the neighborhood-by-neighborhood number crunching. For some further reading on how Alief's population (Asian and otherwise) is expected to change over the years, this may be some interesting reading. For some further reading on a part of the state that ranks 5th in terms of Asian population share, here's an intro to Moore County, which is north of Amarillo.

- The big spaghetti-mess of an image that I talked about is the one below (click to enlarge). Its a block-level view of the maps done here (and the interactive/comparison views here) at the block group level. This update just takes things to a more granular level and reveals some interesting differences in southwest Houston and Alief from other parts of the county. I'll talk about this more tomorrow in a separate blog post. For now ... enjoy the sneak peak.

The color-coding is as follows: red represents an Anglo majority within the block; brown, a Hispanic majority; black, an African-American majority; green, an Asian majority; and yellow means that there is no majority. I've made a few tweaks to the code that generates this, so a few quirks are being ironed out and I may have something for more interactive viewing tomorrow.

- It occurs to me that I never bothered to break out the full Asian nationality breakdown here on the blog, so here's that:

Harris County

Total Asian Population - 256,862
Vietnamese      87,693  (34.1%)
Indian          48,184  (18.8%)
Chinese         42,244  (16.4%)
Filipino        17,045  ( 6.6%)
Pakistani       16,821  ( 6.5%)
Korean          12,829  ( 5.0%)
Cambodian        4,118  ( 1.6%)
Japanese         4,022  ( 1.6%)
Taiwanese        3,715  ( 1.4%)
Bangladeshi      3,038  ( 1.2%)
Thai             2,865  ( 1.1%)
Laotian          1,846  ( 0.7%)
Indonesian         597  ( 0.2%)
Malaysian          587  ( 0.2%)
Sri Lankan         500  ( 0.2%)
Other Asian     10,758  ( 4.2%)

Fort Bend County

Total Asian Population - 101,213
Indian          37,318 (36.9%)
Chinese         22,481 (22.2%)
Vietnamese      15,517 (15.3%)
Filipino         9,218 ( 9.1%)
Pakistani        7,219 ( 7.1%)
Korean           3,179 ( 3.1%)
Japanese           779 ( 0.8%)
Malaysian          722 ( 0.7%)
Taiwanese          676 ( 0.7%)
Thai               252 ( 0.2%)
Indonesian         243 ( 0.2%)
Bangladeshi        190 ( 0.2%)
Cambodian          111 ( 0.1%)
Other            3,308 ( 3.3%)


Honkification Elsewhere

» LA Times: Upscale culture and gang violence share a small space

Not just the Heights ...

The historically working-class Latino neighborhood, which covers just one square mile, has seen a burst of restaurants, galleries, bars and boutiques open in the last few years. In addition to the low-cost eateries from Latin America and South Asia, Mission denizens can find prosciutto ice cream, broccoli raab pizza, or lavender creme brulee served from a cart that roams the district (with a Twitter following, of course).

Young techies, many ferried by corporate buses to and from their jobs at Google and Facebook, have flooded into the Mission's Victorian flats and newer lofts — joining the artists and activists who arrived in the late-1980s and first gave the neighborhood its cool cachet.

The Mission's Latino population has fallen by more than 20% over the last decade as families were priced out of the housing market. What remains are stark contrasts. On a recent evening, tattooed youth with yoga mats pushed past lines of workers wiring their day's wages to relatives in Latin America.

What's the allure?

It is that diversity that draws most newcomers, Martin said, although he concedes that for plenty of others it's about "'I like cheap taquerias when I'm drunk at 2:30 a.m.'"

I don't drink enough to get hammered at 2:30 in the morning (or in the afternoon for that matter). But the appeal of cheap taquerias here in Houston cannot be minimized. And the entertainment value of the 2:30am drunks isn't half bad, either.

Seriously, though ... is there anyone that's written a half-way decently researched book on the phenomenon that isn't prone to the polemics of, say, a Richard Florida? Between the tech boom in the San Francisco area and the preservation ordinances that drive a similar process in Houston, I'd be curious to see what else drives this kind of demographic change. And it's not so much the "reverse white flight" toward downtown areas that I care about. It's the outlying areas that defy the broader demographic trend you see elsewhere that I'm looking for. And in no small part, it's due to the fact that it flies in the face of Bill Bishop's "Big Sort" theory.


Ye Olde Days, Continued

One highlight from the 1980 demographic map posted below. The graphic below focuses in on SW Houston to show the triangle south of Westheimer and north of Bissonnet, which was practically all majority-Anglo as of 1980, yet now has no tracts that are majority-Anglo. Click to enlarge, if you feel so compelled. The degree of change should jump out even if you don't.

The most notable area that does the reverse demographic shift, not surprisingly, is around the Washington Ave. corridor. It goes from a mix of some majority-Black and majority-Hispanic tracts to more majority-Anglo and one or two that remain as a multicultural mix.

Here's the 1980 map (again, click it to big it):

And here's the 2010 version:


CVAP Conversion by City

Another take on the CVAP Conversion Rate here. This time, I'm looking at entire cities and the counts provided straight off of the American Community Survey estimates. The chart below shows the 25 biggest cities in Texas (ranked by overall population). The chart shows Hispanic counts at the level of Total Population, 18+ Population, Total Citizen Population, and Citizen Voting Age Population. As a reminder, the conversion rate is just dividing CVAP by the 18+ counts for Hispanic population.

Interesting differences among the major cities, to be sure. What's even more interesting is applying a bit of algebra to the numbers provided, which show that Houston's <18 population is 87.3% citizen. The trend is replicated throughout the state, with the statewide number for all cities being 92% for <18 Hispanics. --------- Top 25 Texas Cities, as measured by 2005-09 American Community Survey population estimates. CVAP Conversion is a measurement of Citizen Voting-Age Population (CVAP) divided by the total 18+ Hispanic Population.

CityTotal Hisp. Pop.18+ Hisp. Pop.Hisp. Citizen Pop.CVAPCVAP Conversion
San Antonio807,195552,100711,235467,14084.6%
Fort Worth224,835140,125151,11075,21053.7%
El Paso482,185323,355399,425247,46076.5%
Corpus Christi164,590112,845154,485104,01092.2%
Grand Prairie65,19039,57549,07025,40564.2%


A point of comparison to other large cities on their CVAP conversion rate:

Los Angeles ... 48.6%
New York City ... 67.5%
Chicago ... 60.0%
Philadelphia ... 83.8%


Addressing the Rock

» Chron: Planting seeds: Houston's redistricting plan will require time to grow Hispanic representation. (Editorial)

At first glance, the draft redistricting plan for Houston City Council formulated by Mayor Annise Parker's administration seemingly contradicts its purpose. Although the creation of two new district seats results from a voting-rights suit settlement more than three decades ago to expand Hispanic representation as the city's population grew, the proposed new Districts J and K would almost certainly elect Anglo and African-American representatives in November.

As a result, a coalition of Houston Latino groups is demanding changes that would guarantee at least one more Hispanic-dominated district to send a representative to serve on council with Ed Gonzalez of District H and James Rodriguez of District I. They cite fresh census statistics showing that an increase in the number of Latinos accounts for most of the city's population growth over the last decade.

It's not that simple, however. Mayor Parker summarizes the demographic conundrum this way: "The problem is 44 percent of our population is Latino but 16 percent of the vote is Latino." According to the mayor, the proposed map doubles the number of districts with Hispanic voting-age majorities, but "the rock we run aground on is citizenship."

The biggest problem with this analysis - and the city's redistricting approach - is that it ignores the Census Bureau's American Community Survey data on Citizenship Voting Age Population. By ignoring it (it was called "murky" during the Wednesday council hearing on the proposed redistricting plan), the city isn't properly accounting for the very problem they note.

As noted yesterday, the Texas Legislative Council's David Hanna is on record suggesting the data be used. On page 18 of the presentation made to the NCSL last year, he offered two methods for dealing with the different geographic definitions used in both the 2010 Census and the earlier ACS data collection:

- count only those block groups wholly within a district
- assume uniformity of split block groups and assign a ratio of the geography covered within the district.

If I can do this as a one-man show, it's not clear to me why anyone else can't. Had this information been used, it would be more readily apparent that a map can be produced with at least three districts that are over 65% Hispanic, with two districts over 50% CVAP and one within reach this decade. By not using it, the decisions for where you look for the opportunity get ... well ... murky. And if the best result is two districts where there are no voters to be found (as has been suggested), then I'd suggest that not enough has been done to locate a true opportunity. If you don't believe me, then feel free to check the side-by-side view of Harris County in any combination of Total Population, Voting Age Population, or CVAP at either a Census Tract or Block Group level.

Spanish surname registered voters are a guide, but they are a far more conservative baseline for measurement. Furthermore, since you're not accounting for voter-eligible, non-registered citizens, the problem of "Hispanics not voting" is etched in stone and preserved for at least a decade. I don't have a quarrel with using as many datapoints as possible, but it was a mistake to not use CVAP at all.

On a strict CVAP count, Houston Hispanics would account for 2.56 seats on council, whereas the full population would warrant 4.7. Locating an opportunity to build on that .56 opportunity in order to more fully reflect the city's population is no doubt a challenge. I don't pretend to offer my suggested map as a conclusive solution that is guaranteed to elect an Hispanic. But I also know of no Hispanic leader afraid of the challenge to win in that environment.

Build it ... and let's see who gets elected.


COH Redistricting: The Hobby Report

Here's the Hobby Report referenced in Wednesday's public hearing on the City Council District map. The takeaway I see in it is that Hispanic-Anglo polarization is notably less than Black-Anglo or Black-Hispanic polarization. A lot of that is intuitive, but I'm curious to see how much of that holds true when the Hispanic candidate doesn't have conservative and/or GOP bonafides. In other words: Morales, Sanchez, and Saenz might have had good odds of crossing over. Ironically, It's the Morales v Brown race in 2005 that has the highest polarization: Anglos were likelier to go for Brown, while Hispanics likelier to vote for Morales. Read and see what you come away with.

Hobby Report - Polarized Voting Analysis


Defining “Opportunity”

So the argument "for" the proposed City Council district map is that there are 4 "opportunity" districts for Hispanics. As the Chronicle's report notes:

Much of the discussion of Hispanic influence centers on the proposed A and F districts, now held by Brenda Stardig and Al Hoang, respectively.

Parker said Hispanics simply must "find the votes" in those districts, and Rodriguez acknowledged conventional wisdom that Hispanics would have greater power if more came out to vote.

"I think it's incumbent upon all Latino officials to get our numbers up," Rodriguez said, but added that turnout alone cannot offset the insufficient numbers of Hispanics in the proposed A and F districts.

"They vote at the same rate that other people do of that same socioeconomic level," Wood said. "It is the citizenship that brings down the Hispanic participation in elections."

Indeed, it is citizenship that brings those numbers down. So the open question I see is exactly how much opportunity exists to convert that total population majority and voting age population majority into an electoral majority. Unfortunately, there's some bad guesses out there on this:

Rice University political science Professor Bob Stein theorized Hispanics' 51 percent voting-age majorities in A and F could drop 10 points when citizenship, voter registration and inclination to vote are considered.

I just got through doing a first-draft count and here's my math ...

     CVAP    Anglo    Hisp    Afr-Am    Asian
A   85,005  41,584   21,480   17,160    4,084
            48.92%   25.27%   20.19%    4.80%
F   84,465  24,295   19,569   26,550   13,185
            28.76%   23.17%   31.43%   15.61%
H   90,800  23,630   47,440   18,578      784
            26.02%   52.25%   20.46%    0.86%
I   77,235  13,985   45,914   14,970    1,813
            18.11%   59.45%   19.38%    2.35%

The sum total of this breakdown is that you have two Hispanic opportunity districts, but two that contain illusory "opportunities" for Hispanic voters. The overwhelming majority of them do not have the opportunity to register to vote. The overwhelming majority of them do not have the opportunity to cast a vote. Yet they remain in two different districts instead of making an effort to include them within a whole district where they would have greater voting strength.

Some caveats are in order. I've noted before that CVAP calculations aren't done in the same way that they've been done in the past. They're now based on American Community Survey reports that the Census Bureau produces, rather than from the long form of the Census. And there may be even more issues with the way the numbers are calculated. And even beyond that, there's a margin of error to build into any use of ACS data. A second major issue with the data is that it reflects data points from years past - so it does not fully reflect the 2010 settling of demographic change.

For argument's sake, feel free to call the CVAP Hispanic numbers in A & F at 30%. That's not opportunity. That's fracturing.


The Asian Population in SW Houston

As mentioned in the previous post on District G's redistricting town hall, here's the map of Asian population concentrations within southwest Houston. The color-coding is of Census block groups and is as follows:

dark green: >20% Asian
green: 12.5-20% Asian
light green: 8-12.5% Asian

It's important to keep in mind that one restriction the city places on itself for redistricting is in the use of County-assigned voting precincts. I show the block groups here so that we can get a fuller view of the population in better detail. But it shouldn't be forgotten that some of these population shares get obfuscated by the way precincts are put together. What follows is some of my own logic in spelling out the case for why the boundary line for a District F should be moved north, to Westheimer.

The area I noted as being "debated" at the town hall was that area north of the current F/G boundary, which is essentially the Westpark tollway and Westheimer. That's the area that I think makes some sense for considering where to add population to a reformatted F. Since population will have to be shed, I think a safe delineation for that is that everything southeast of the Southwest Freeway is expendable. Sharpstown is a classic example of two city redistricting criteria being at odds since one voting precinct of Sharpstown is on the other side of a major highway. In this case, I see the highway as a possible eastern boundary for a reformated F.

For the time being, I'm agnostic on how Sharpstown ends up being treated, whether it's entirely in a new F, split up, or put in a new district. I don't doubt that the fine folks within Sharpstown would prefer to be kept whole and I'm certainly sympathetic to the argument of keeping as many neighborhoods whole as possible. We're likely to hear from a few of them at the District F meeting at the end of the month. That said, there are a few precincts that would make a good fit with a District F that seeks to solidify the Asian population as much as possible (311, 426, and 297). So I suppose how I'd personally end up feeling about Sharpstown being districted would be dependent upon how much better the Asian population was folded into such a district elsewhere.

As Ben Crocker notes in the comment to the original post, one of the areas well represented last night came from residents of Briar Hills, which contains much of the dark green block groups north of Westheimer. In a perfect world, I'd probably love to argue that those neighborhoods should be included in a stronger-Asian District F, also. It would certainly make the demographics better within such a district. But there are two reasons why I don't offer that as a possible addition for a "new F." One is that at some point, the ideal population kicks in as a factor. That sets a hard cap on how much you can meander out to set boundary lines. The second is what goes into this assumption of southwest Houston's Asian community defining a "community of interest." I haven't been through Briar Hills very many times in recent years, but I don't know that it could be argued that there is a dedicated commercial sector on par with what you have along Bellaire Blvd. And I don't know that you'd be as likely to run into as many language-isolated parts of the Asian community north of Westheimer. So the multicultural "cohesiveness" that you see in southwest Houston becomes something quite different when you get north of Westheimer.

As for my own precinct - 430 - I can say that I'm perfectly ambivalent with how it ends up being districted. 430 is a bit of an orphan precinct - by no means Sharpstown, but it only took on population that it now has after Gulfton ran out of room for new apartments. I can see it fitting in a new F, a new G, or a new district based out of Fort Bend and Fondren SW, or a district that includes Gulfton with the Braes Bayou neighborhoods. I can't say that I have a preference between any of those. It's a very weird position to be in given the extent to which I follow redistricting news, but there ya go. Just be sure to include it somewhere, I suppose.

Fatal error: Call to undefined function xwp_footerx() in /nfs/c01/h12/mnt/4748/domains/ on line 18