Off the Kuff Rotating Header Image

The Blum poll

I don’t know what to make of this.

Ill will toward President Barack Obama and a voter belief that Republican Gov. Rick Perry has helped save the Texas economy are giving Perry a re-election lead over Democrat and former Houston Mayor Bill White, according to a new poll done for the Houston Chronicle and four other state newspapers.

Perry leads with 46 percent support to 39 percent for White, with Libertarian candidate Kathie Glass trailing at 4 percent among likely voters; 11 percent were undecided.

[…]

The enthusiasm of Republican voters is dramatic. Among registered voters who answered the survey, Republicans held an advantage of 9 percentage points — roughly typical of the vote in recent elections.

Among those who said they are likely to vote, the Republican advantage jumped to 18 percentage points over Democrats.

The telephone survey was of a random sample of 1,443 Texas adults, including 1,072 registered voters and 629 likely voters, conducted Sept. 15-22.

The margin of error for the likely voters is plus or minus 3.9 percentage points. The survey was conducted by Blum & Weprin for the Houston Chronicle, the San Antonio Express-News, the Austin American-Statesman and The Dallas Morning News.

Perry and White are in almost a dead heat in Houston and San Antonio. Perry is drubbing White in the Dallas-Fort Worth area, with a lead of more than 30 percentage points in those cities. White holds a 7-point lead in Austin.

The two are even in urban areas, but Perry has an almost 11-percentage-point lead in non-urban areas.

I could believe White and Perry are running even in Harris and Bexar counties, which is what I presume they mean by “Houston” and “San Antonio”. I expect White will do better than that, but it’s at least a plausible result to me.

White is leading in Travis County (again, my interpretation of “Austin”) by seven points? Chris Bell, in his four-way race, won Travis by 19 points. David Van Os, who lost by 22 overall in 2006, carried Travis by almost nine points. White leading by seven? I’ll take the over on that.

Perry leading in “the Dallas-Fort Worth area” by 30 points? I’ve already shown that these are Tony Sanchez numbers. Once again, color me skeptical.

On the flip side, how is Perry only leading by 11 in non-urban areas, which are the Republican strongholds, especially in a year where the electorate is supposedly nine points more Republican than expected? (Note that this is not a universal finding.) If Perry is doing better in the places that White would be expected to be strong, how is he doing worse in the places where he’s supposed to clean up? Putting it another way, if White is underperforming in the urban counties, and the electorate is extra Republican this year, how is it that Perry isn’t leading by double digits? Something doesn’t add up.

Related Posts:

2 Comments

  1. blank says:

    I agree with you Kuff. These statements seem way off the mark.

    Perry is drubbing White in the Dallas-Fort Worth area, with a lead of more than 30 percentage points in those cities.

    Dallas County
    John McCain/ Sarah Palin REP 310,000 41.97%
    Barack Obama/ Joe Biden DEM 422,989 57.27%

    Tarrant County
    John McCain/ Sarah Palin REP 348,420 55.43%
    Barack Obama/ Joe Biden DEM 274,880 43.73%

    Combined
    John McCain/ Sarah Palin REP 658,420
    Barack Obama/ Joe Biden DEM 697,869

    The two counties combined maybe swingy, but there is no way that White is losing a region by 30 points that Obama won by almost 40K votes.

    White holds a 7-point lead in Austin.

    Travis
    John McCain/ Sarah Palin REP 136,981 34.44%
    Barack Obama/ Joe Biden DEM 254,017 63.86%

    I’d be shocked if White doesn’t win Travis by 20.

    The two are even in urban areas, but Perry has an almost 11-percentage-point lead in non-urban areas.

    Only if they way over sampled Hidalgo.

  2. […] The Express News version of this story says the poll was of “1,073 registered voters, including 673 likely to vote”. The question I have, given that this poll was conducted between Days 5 and 9 of early voting, is why wouldn’t you ask people if they have already voted? Seems to me those are a decent chunk of your truly likely voters by now. At the very least, you can see how things differ between those who have voted and those who have not (yet) voted. I have no idea why the pollster didn’t take this approach. You can add that to the problems I had with their previous poll. […]

Bookmark and Share