Some shots at R-J pollster hit, others miss the mark

On television on election night, while offering his "expert" commentary, a quisling Republican who backed Harry Reid in exchange for support of his choo-choo train to Victorville said the Review-Journal should fire its pollster.

On Wednesday, Democratic Senate Majority Leader Reid himself called polls meaningless because his get-out-the-vote machine could deliver more warm bodies to the polls than the undisciplined enthusiasm for Republican Sharron Angle.

"I've been wanting to say this for some time," Reid told reporters. "We've got to do something about these misleading polls. They are all over the country, they are so unfair, and you just gobble them up -- no matter where they come from. You just run with them as if they are the finest piece of pastry in the world. They are false and misleading, and people pay for those polls, so you use them."

Dutifully and on cue, on Thursday the front page of the Reid-sycophantic Las Vegas Sun gleefully declared in a bold-faced, all-caps headline: "POLLING HASN'T CHANGED WITH THE ELECTORATE." This was followed by a red-tinged subhed: "REPUBLICAN BIAS?" and "Newspaper's last polls gave edge to Sharron Angle, who lost, and Joe Heck, who barely won."

The final Mason-Dixon Polling & Research telephone survey showed Angle leading Reid by 4 points, within the 4-point margin of error. It also showed Joe Heck leading incumbent Democratic Rep. Dina Titus by 10 points.

When the final numbers were posted near midnight Tuesday, Reid prevailed by 5 points and Heck by 1 point.

Reid was right. And he was wrong.

He was wrong about doing away with polls, because that would be like playing football without yardage stripes. All a poll does is tell everyone what yard line the ball is on at any given time in the game.

He was right, because on Tuesday and in early voting, the Democrats called a quarterback sneak and used their 350-pound linemen -- unions, casinos, power-hungry Republicans for Reid -- to shove the ball over the line by busing thousands of unionized casino workers to the polls, offering rides to polls, telephoning and pestering those leaning Democratic and having casino executives confront by name workers who had not yet voted. It was a tactic completely unmatched by the Republicans.

The Wall Street Journal noted that hundreds of Democratic congressional aides, operatives and lobbyists -- including 30 aides for Sen. Max Baucus -- descended on Nevada for the get-out-the vote effort. The Republicans sent none.

It is one thing to ask people six days before the election whether they will vote and for whom. It is another to be handed a pre-marked ballot, marched onto a bus and taken to the polls by your shop steward.

Brad Coker, managing director of Mason-Dixon, admits he failed to anticipate the lopsided ground game that came in the days after the poll was taken. What may have been correct six days before the election did not withstand the all-out push. Never mind that every other race and ballot question was accurately forecast, most within the margin of error -- from governor to secretary of state to sheriff.

Both Reid and the Sun -- which hasn't spent a dime on polling of its own, but frequently quotes the Review-Journal's -- made much of the fact that random-dialing polling does not access that one-quarter of U.S. households that have no land line.

But the assumption that failing to call cell phones undersamples Democrats is a red herring.

Mason-Dixon matches its polling sample to the jurisdiction's demographics of likely voters -- gender, age, race, political affiliation, urban-rural. There is zero evidence a 24-year-old, cell-only female Democrat votes differently from one with a land line.

The latest poll oversampled Republicans, as was clearly stated, because that was how the enthusiasm level was breaking nationwide.

Coker notes he and every other major polling firm in the country were seeing a consistently higher turnout among Republicans than Democrats. "In all the races we polled, other than Nevada, this turnout pattern held up on Election Day and is evident in the massive Republican wins all across the rest of the country," he says. "This polling was done in other states far more 'blue' than Nevada, such as Michigan, New York, Pennsylvania, Illinois and Wisconsin, and it proved to be very accurate based on this basic assumption."

Considering Reid's 56 percent unfavorable rating and Nevada's highest-in-the-nation joblessness and foreclosures, he did not anticipate the turnout would be different in Nevada from elsewhere.

Lesson learned.

Thomas Mitchell is editor of the Review-Journal and writes about the role of the press and access to public information. He may be contacted at 383-0261 or via e-mail at tmitchell@reviewjournal.com. Read his blog at lvrj.com/blogs/mitchell.