Articles and Analysis


Re: SurveyUSA Texas

Topics: 2008 , ABC/Washington Post , Gallup , IVR Polls , Likely Voters , Pollsters , PPP , Rasmussen , SurveyUSA

A few comments on our post of the new SurveyUSA Texas poll raised two questions worthy of further discussion.

First, reader s.b. notes:

[W]ith an automated survey, if its in English, they aren't sampling spanish only or mostly spanish speakers. I think it skews these results.

Some pollsters (such as Gallup) offer voters the opportunity to complete the survey in Spanish when they encounter Spanish speaking respondents. Most pollsters, however, will simply end the interview in these instances. I asked SurveyUSA's Jay Leve about their procedure in Texas and he notes that while they do have the facility to offer respondents the option to complete a survey in either English or Spanish (and have done so in mayoral elections in New York and Los Angeles and some congressional districts), they did not offer a Spanish interview for their Texas poll.

However, before leaping to conclusions about the SurveyUSA results, keep in mind that none only one of the other Texas pollsters report using bilingual interviewing for any of their surveys [Correction: interviews for the Washington Post/ABC News poll "were conducted in English and Spanish"]. Three of the other pollsters -- Rasmussen Reports, PPP and IVR polls -- also interview with an automated methodology rather than live interviewers.

And before leaping to conclusions about all the Texas polls, we might want to know just how many Latino voters in Texas speak only Spanish. I have not done survey work in Texas, but my memory from conversations with pollsters that do is that the percentage that will actually complete an interview in Spanish when offered is typically in the low single digits.

Second, several commenters have speculated about the small changes in the demographic composition of the last two SurveyUSA Texas polls. For example, "Mike in CA" points out:

Hispanic turnout at 28% sounds just about right. The last SUSA survey had it at 32% which was way too high. It seems SUSA has scaled back their Hispanic estimates, so they must have a reason. Additionally, the boosted AA to 23%, from 18%. Seems reasonable considering the extraordinary increases in early voting turnout from Houston and Dallas [emphasis added].

That's not quite right. Keep in mind that SurveyUSA's approach to likely voter modeling is comparable to that used by Iowa's Ann Selzer, in that they do not make arbitrary assumptions about the demographic composition of the likely electorate. As SurveyUSA's Jay Leve explains, they "weight the overall universe of Texas adults to U.S. census" demographic estimates, then they select "likely voters" based on screen questions and allow their demographics to "fall where they may." So some of the demographic variation from survey to survey is random, but large and statistically statistically significant variation should reflect real changes in the relative enthusiasm of voters. Leve goes into more detail in the email that I have reproduced after the jump, which also includes the full text of the questions they use to select likely voters.

Comments emailed by SurveyUSA's Jay Leve:

SurveyUSA makes no “forced” assumptions about the size or composition of any likely voter demographic group. We weight the overall universe of Texas adults to U.S. census, and let the sub-groups fall where they may.

What that means is: if, using the identical methodology from one tracking poll to another, the composition of Hispanics drops from 32% to 28%, it did so occurring naturally. In the same way, if the composition of seniors drops, it did so occurring naturally. If the number of likely voters increases 31% of adults on 2/19/08 to 35% of adults on 2/25/08, it did so occurring naturally. (same exact instrument used for both surveys).

Any one of these changes may indeed have happened by chance alone – because of random sampling – but when a couple of groups that could be aligned with support for one candidate appear to shrink simultaneously, it is possible to infer some diminished enthusiasm for that candidate. (Composition of Hispanics declined from 32% to 28%, and composition of voters age 50+ declined from 46% to 42%, from 2/18/08 SurveyUSA TX poll to 2/25/08 SurveyUSA TX poll).

In the same way, though on a separate point: SurveyUSA found in 2/25/08 survey that of 2,000 adults in Texas, 704 were likely to vote in the Democratic Primary, 484 were likely to vote in the Republican Primary. If you do ballpark math, you might (incorrectly) infer that Democrats are 45% more plentiful in TX than Republicans. They are not. But Democrats are more likely to tell SurveyUSA they have already voted, or are likely to vote, in the Texas Primary. That enthusiasm, or lack of enthusiasm, among certain groups, filters through to the final primary poll composition numbers.

The same exact survey instrument that produced 704 likely Democratic Primary voters and 484 likely Republican Primary voters, was also used to produce the following general election results (released today 2/26/08), which are based on registered TX voters, not likely primary voters. The number of registered Republicans and number of Registered democrats is approximately equal. (39% Republican, 38% Democrat).

In summary: in almost every case, by design, the water in a SurveyUSA poll seeks its own level, and as enthusiasm for a particular cause or candidate manifests itself, it flows through to our results (comment continues below image).

02-26 SUSA graphic.png

The exact questions used to determine likely primary voters in TX were:
Copyright 2008 SurveyUSA.

1) Are you registered to vote in the state of Texas?
Yes, press 1.
No, 2. [branch to demographics]
Not sure? 3. [branch to demographics]

2) On March 4 … [On Tuesday …] [Tomorrow …] Texas will hold a presidential primary election. Early voting is possible. Have you already voted?
Yes, press 1.
No, 2. [branch to Q4]

3) Did you vote in the Republican primary? Or the Democratic primary?

If you ...
Voted in the Republican primary, press 1. [branch to GOP voting question]
Voted in the Democratic primary, 2. [branch to Democratic voting question]
Can't remember? 3. [branch to demos]

4) On a scale of 1 to 10, how likely are you to vote…where 10 means you are certain to vote … and 1 means you are certain NOT to vote?

[highest-interest voters continue; others are branched to demographics]

5) If you are …
Certain to vote in the...Republican Primary, press 1. [branch to GOP voting question]
Certain to vote in the Democratic Primary, 2. [branch to Dem voting question]
Not yet certain which Primary you will vote in, 3. [branch to demos]



Wow! That's great information. I think it's much better to use census data and let the screen questions determine the % etc. However it is then also subject to margin of error.

I do think however that those who don't speak English fluently are probably more likely to be elderly and identify and vote more with their ethnic group. This means those people would be Clinton voters.

Certainly, where I'm from, politicians love people who don't speak English well or fluently because they are easier to target through their cultural leaders and groups and they do vote as a block, overwhelmingly and they come out to vote, even for delegate selection meetings that others don't.

But great questions answered by SUSA. I love reading their cross tabs.



I might add, it is clear from numerous polls, that Texas is about the % of hispanic turnout, and to a lesser extent women and seniors.

What I find interesting in the polls from Texas is that Obama is only getting about 75% of the blcak vote. Sure that sounds like a lot and it is, but its not 90%, which could also make a difference.



I have to give to you guys, you mean you haven't drunk the cool-aid on Obama. wow, great. It seems everyone else did. I'm so sick of this guy. Hillary was not my choice, but after the way Tim and those guys treated a lady, i'm voting for Hillary now.



Trying to drill down on the question of under-sampling of spanish speakers - Nationally, about 80% of U.S. hispanics prefer to communicate in english. Because the border areas of Texas have a very high immigration rate, it seems likely that this percentage is a little lower in these areas (as opposed to say, New Mexico).

However, I am originally from South Texas (McAllen), and I can say with some certainty that nearly all of the legal population (potential voters) is bilingual. You cannot become a citizen without knowing some english, and people in South Texas generally (both 'white' and hispanic, although the area is 90%+ hispanic) interchange english and spanish regularly. They don't call it Tex-Mex for nothing.

I don't there are many hispanics unable to answer a simple english phone poll, but there is probably a small percentage who are less likely to bother going through it in English. Looking at it this way, this issue starts to resemble many similar issues inherent in phone polls (cell-only voters, no-phone voters, caller-id screening voters, etc.). As such, I don't think there will be any noticeable variation specifically because of this issue.




Let me get this straight. You're going to base your vote for President of the United States on what Tim Russert said during a presidential debate. That's kinda like deciding who to vote for based on a poll. You're sad. I hope to God we as Americans can hold up a standard and expect that someone qualified will be able to assume the post. Good luck with that vote.



Great post. I wanted to note - but am too lazy to look up - the Census publishes statistics on "linguistic isolation", that is, the percent of speakers who do not speak English (there are various categories). The data is also broken down by race/hispanic origin.




Stop drinking Obama cool aid. It is Senator Clinton who is clearly the most qualified candidate. As historian Sean Wilentz has pointed out, it is the Obama camp that has played the "race" card by falsely accusing the Clintons of playing it.
Apparently you believe all of the misogony out there, so you slime the Clintons at every opportunity.



I support limiting the comments on this site to ones discussing poll issues in particular, not discussions of why one supports or opposes a particular candidate. There are many blogs and comment sections for the latter.



For what it's worth, I work in Dallas, as liaison to our customer survey vendor. Our vendor tells us that about half of the 11,000 survey calls they do for us in Dallas County require Spanish. Admittedly that doesn't say much about likely voters (we're in the health business), but I wouldn't want anyone to take away from the previous comments that it's rare that an adult in north Texas (over 400 miles from Mexico) is primarily a Spanish speaker. Linguistic isolation is a rapidly growing problem in the Dallas area.



Really a nice survey. It reviles the truth.
thomas maximus
texas drug rehab


Post a comment

Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.