Texas - The pollsters should just admit it's all guesswork now

Following up my last election post (Hillary's last stand ... or back to trench warfare?) ... here is somewhat different roundup of the major Texas polls, one day before the primary, from RealClearPolitics:

Poll	Clinton	Obama	Spread
RCP Average	46.8	46.5	Clinton +0.3
Rasmussen	47	48	Obama +1.0
InsiderAdvantage	49	44	Clinton +5.0
PPP (D)	50	44	Clinton +6.0
SurveyUSA	48	49	Obama +1.0
Reuters/CSpan/Zogby	44	47	Obama +3.0
WFAA/Belo Tracking	46	46	Tie
M-D/Star-Telegram	45	46	Obama +1.0
FOX News	45	48	Obama +3.0

Averaged together, they come to an almost perfect tie. But they can't all be right.

=> Most of these polls show Obama ahead--just barely, but since February 5 he has usually out-performed his polling figures.

=> Two of them show Clinton ahead and widening her lead. But even the people who did the InsiderAdvantage/Majority Opinion Research poll, which shows Clinton ahead by 5 points, clearly find it difficult to feel confident about their results:

A new InsiderAdvantage/Majority Opinion Research poll of Texas runs contrary to most surveys of the state’s Democratic presidential contest. We show Sen. Hillary Clinton with the lead. Only PPP, a Democratic polling firm based in North Carolina, also shows Clinton leading. All other polls over the race’s final days have Obama leading.

Our survey of 609 likely Democratic primary voters was conducted March 2. It has been weighted for age, race, gender and political affiliation. It has a margin of error of +/- 3.8%. The results were:

Clinton: 49%
Obama: 44%
Undecided: 7%

InsiderAdvantage CEO Matt Towery: “This is a lonely and difficult spot for a pollster. However, we have polled thousands of voters in Texas and have yet to show Sen. Obama leading. Our weighting is designed to discount Hispanic turnout in areas where historically their turnout has been lighter, and to maximize potential African-American turnout. Even with this model, which we believe to be accurate, though obviously not a perfect situation for Clinton, she leads. [....]

“It is entirely possible that we have missed the mark here, given that so many pollsters have shown Sen. Obama with a growing lead. But I would note that the most recent surveys all indicate a tightening of the race; and also that the breakdown by demographics in our survey is amazingly similar to the PPP numbers. If nothing else, it appears that Texas truly is up for grabs,” Towery said. (click here for crosstabs)

In short, anything could happen tomorrow. (And these figures predict only the popular vote in the primary election. After the voting in the primary is over, that will still leave the caucuses and the peculiar delegate-allocation system to deal with.) Stay tuned ...

--Jeff Weintraub