TX-23: SUSA Poll Gives Edge to Bonilla

Thanks to X-Stryker in the diaries, we get word of this SUSA poll (likely voters, no trendlines):

Rodriguez (D): 46
Bonilla (R-inc.): 53
Undecided: 1
(MoE: ±4.6%)

Perhaps the most salient thing to note is that 59% of likely voters in this sample were white, and just 36% Hispanic. (Hispanics favored Democrats heavily this year, so a high Hispanic turnout is key for Rodriguez.) Meanwhile, census figures show that district is 55% Hispanic and 41% white. The sample isn’t implausible, though. Latino voting-age population (VAP) tends to be lower than the white VAP, and the registered voting-age population (RVAP) lower still.

Moreover, according to the census, Latinos comprised almost 35% of the state’s population in 2004. However, exit polls this year in the Texas gubernatorial and senate races say that Latinos made up only 15% of those who went to the voting booth on election day.

If the exit polls are accurate, then only some 43% of Hispanics statewide turned out to vote. If the SUSA sample is right, then Hispanics are showing up at a 65% clip. That’s quite an improvement – but it looks like it may not be enough. Don’t forget that despite the big changes redistricting wrought here, the new 23rd CD went for Bush by a hefty 57-43 margin in 2004 (it had been 64-35 – though I should note that Al Gore only lost by about 7 points).

So I really don’t know if Ciro can do much better among Latinos than he’s already doing. However, he could still pull it off if he can increase his margin among independents (who favor him by just three points).

And if you’re outside of the district and want to help, the best thing you can do is sign up for online phonebanking.

Update (James L.): As noted in the comments by blank, the 55% figure for Hispanic population in this district does not reflect the district’s make-up after the 2006 redistricting process.  The current Hispanic population of TX-23, therefore, is 65%, meaning that the Latino turnout rate on November 7th in this district was around 55%.  We’ll have to see if that can hold for the special election.

7 thoughts on “TX-23: SUSA Poll Gives Edge to Bonilla”

  1. Having looked at election returns from throughout the country and accompany population growth projections, I’m of the mind that it will probably be two generations before Latinos are a serious force in American politics, and I suspect the sample for this demographically divided district is accurately depicting the “likely voter model”.  If current voting and population trends continue, my calculations suggest that it will not be until at least 2040 that there will be more Latino voters than African-American voters, even though Latinos already outnumber African-Americans in the general population.

    Long-term (and I mean VERY long-term), places like TX-23 will be favorable turf for Democrats barring a political realignment in which Latino voters become Republican.  Right now though, I’m afraid I’m more inclined to buy the premise that 59% of likely voters in TX-23 are white West Texas cowboys than a premise in which the census’ reported 55% Latino population in the district will be the least bit representative of who votes in next week’s special election.  With that said, Rodriguez is hanging tough in that he’s even within striking distance considering how Republican the whites in this district are.

  3. if SUSA conducts interviews in Spanish as well as English?

    With 1/3 of the CD claiming Hispanic heritage, the language of the interview and the response rate are very important.

    The language issue should be obvious, and the response rate can go a long way to telling us whether large numbers of people didn’t answer the survey because the call started in English.  In either case you could get a skew.  Hispanics seem to be falling into the same immigrant group modality that marked the Irish, Italians, and Poles before the fully assimilated.  But it isn’t that simple, because I’ve seen surveys before that indicate first generation migrants, people who’ve themselves are immigrants are substantially more conservative than their children. 

    But I have to wonder once we get out to the 3-4 generation whether the group effect wears off as people begin to speak English exclusively, and move out of ethnic enclaves into more diverse neighborhoods. All of this has important implications for the future.

    As hispanics assimilate, and as many begin to attend protestant churches will the group effect wear off?

    I don’t know.

  4. If Ciro can get more hispanics, primarily from Southside Bexar, than he can win this. He needs a superb ground game for this.

  5. Meanwhile, census figures show that district is 55% Hispanic and 41% white.

    That was the old Tom DeLay map.  The latest map has 65%.

    On that note, I just sent the following email to the Editor of SurveyUSA:

    Dear Survey USA Editor,

    I just read the results of Survey USA Election Poll #11357 from the following web page:

    http://www.surveyusa…

    The summary says, “If Hispanics, who are 55% of the population in TX 23, make up more than 36% of those who vote in the Runoff, the contest will be closer than SurveyUSA’s numbers here show.”

    My concern is that Hispanics do NOT make up only 55% of the population in TX 23 as stated.  Hispanics made up only 55.1% of the population in the 109th PLAN 01374C.  However, this is an outdated map.  They in fact make up 65.1% of the population in current map PLAN 01438C.

    So my question is: Since the summary gives population information from PLAN 01374C, is it possible that TX 23 from PLAN 01374C was sampled instead of the one from PLAN 01438C?

    Judging from the other numbers, I suspect that the answer is “No.”  After all, only 8% of respondents indicated that they did not vote on November 7, which sounds way too low if the wrong district was sampled.  However, I just want to have this point clarified.  In any case, Survey USA may want to update its summary to include the correct percentage of Hispanics in TX 23.

    Sincerely,

    XXXXX

  6. I don’t like the looks of this poll, for the ciro to win under this hispanic/anglo about 70/30 opposite splits both ways( a few points better for ciro among hispanics) he’d need almost half the voters to be hispanic. Even with a generous 45% of voters being hispanic he’d lose like at this distibution by nearly 2 %. He’d need about 47.62% of the vote to be hispanic to win.

    Also how’d you get

    “The current Hispanic population of TX-23, therefore, is 65%, meaning that the Latino turnout rate on November 7th in this district was around 55%.  We’ll have to see if that can hold for the special election.”

    That seems way too high.

Comments are closed.