A Rigged Poll Deconstructed

DIGG THIS

Rigged “scientific” polls are the chief method used by the mainstream media to keep Ron Paul down. Nothing can be clearer than a recent South Carolina poll issued by a group known as SurveyUSA. Let's take a look at this poll.

GOP Results

Rudy Giuliani 26% Mitt Romney 20% Fred Thompson 18% John McCain 14% Mike Huckabee 12% Other 6% Undecided 4%

The poll then explains how they picked their “likely voters."

“Filtering: 2,200 South Carolina adults were interviewed 11/09/07 through 11/11/07. Of them 1,895 were registered to vote. Of the 1,895, a total of 1,483 were identified by SurveyUSA as likely to vote in the South Carolina Presidential Primary. Of them, 201 were disqualified by SurveyUSA because they could not commit to voting “no matter what date the primary is held,” and of the remaining 1,282 likely voters, 257 were disqualified by SurveyUSA because they said they had not decided yet which Primary they were going to vote in. The net yield is 577 likely Republican Primary voters and 447 likely Democratic Primary voters. All of these are voters who say they are certain to vote no matter what day the primary is held, and are certain today which one of the two primaries they will vote in. Caveat: Measuring African American turnout in South Carolina is a challenge in any election, but uniquely so in a Primary where an African American is a candidate for President. It is unknowable at this hour how many infrequent black voters, if any, Obama may turnout in a Primary, and, separately, how many infrequent young voters an Obama candidacy may attract in a Primary. These survey results should be evaluated with these caveats in mind.”

This is a rigged poll. Here is my analysis.

“Other”

According to this poll, Ron Paul does not exist. But since Tom Tancredo and Duncan Hunter usually poll at about 1%, we can safely assume that 4% of this is Ron Paul. It might have been more, had his name been mentioned. Also, 2200 adults were called and none of them heard Ron Paul's name so the poll itself is promoting the other so-called first tier candidates.

“Undecided 4%”

This poll makes the preposterous claim that only 4% of the voters are undecided and implies that 96% are decided. That is ludicrous. If 1/3 of the stated support for each of the five named candidates is really undecided leaners, then the undecided vote jumps to 34%, and Giuliani's support drops to 17%. That means Ron Paul can beat Giuliani by winning the support of undecided voters. I wrote about this in more detail here.

"Leaners" and Polling Bias

“2,200 South Carolina adults were interviewed 11/09/07 through 11/11/07. Of them 1,895 were registered to vote.”

This is a poll of only registered voters. But in South Carolina, you have until December 19 to register. So there is no reason to eliminate unregistered voters if they say they will register and vote.

“Of the 1,895, a total of 1,483 were identified by SurveyUSA as likely to vote in the South Carolina Presidential Primary.”

The poll does not say exactly how they were identified, but if voting in the 2004 GOP primary was a criteria, then they may be missing Ron Paul voters. In the 2004 GOP primary almost everybody voted for George Bush.

“257 were disqualified by SurveyUSA because they said they had not decided yet which Primary they were going to vote in.”

So the poll has eliminated 257 out of 1282 likely voters, about 20% of them. These eliminated voters are crossover voters and it's widely known that Ron Paul has crossover appeal because of his stance on the Iraq war. Let's assume that half of these voters vote republican, and of those who vote republican, vote for Giuliani because they perceive him as tough on terror, vote for Thompson because he was a TV star, vote for McCain because they like the McCain-Feingold law, and vote for Ron Paul because of his stance on the war. If so, that would add 32 votes to Ron Paul out of 577 + 32 = 609 total votes. Ron Paul probably loses 5% in this poll from just this one action.

“Caveat: Measuring African American turnout in South Carolina is a challenge in any election, but uniquely so in a Primary where an African American is a candidate for President. It is unknowable at this hour how many infrequent black voters, if any, Obama may turnout in a Primary”

This is true, but the same logic can be applied to Ron Paul. It is likely that many infrequent and disgruntled voters will vote for Ron Paul.

“and, separately, how many infrequent young voters an Obama candidacy may attract in a Primary.”

Again, this logic can be applied to Ron Paul since Ron Paul has shown special appeal to America's youth.

“If the election were held today, would you vote for … “

This is the standard way poll questions are asked for elections including this poll, but it is actually biased, since it does not measure true candidate support. Ed Thompson (Tommy Thompson's brother) did a poll that measured true candidate support in a 2002 Rasmussen poll for Wisconsin governor, and his support rose from 11% to 23%.

STATEWIDE OPINION SURVEY OF 1,000 WISCONSON VOTERS WHO PARTICIPATED IN THE 2002 ELECTION

Cell Phones

This poll makes no mention that they are not calling cell phones. There is reason to believe that Ron Paul voters are more likely than the average to not have a landline. Although pollsters are not allowed to call cell phones, the effect of not calling them should be analyzed.

Get Out The Vote (GOTV)

Polls do not measure GOTV strength. Ron Paul will have the strongest GOTV strength because a much higher proportion of his supporters are putting up yard signs, voting in straw polls, and donating money. In other words, Ron Paul supporters are motivated. Anyone can answer the phone and claim they will vote for McCain. It is quite another thing to show up at the polls and vote.

Monkeying With Demographics

This poll does not tell us how demographic adjustments were made. There is every reason to not trust this poll, so I have zero confidence that the demographic adjustments were done fairly and scientifically.

Timing of the Poll

This poll was conducted on November 9, 10, and 11, right after Ron Paul visited South Carolina and made his big splash by raising $4 million dollars in one day. This poll was designed to undercut Ron Paul's support in South Carolina.

Approval Ratings

This poll does not list the approval and disapproval percentages for the candidates. Typically, front running candidates will have higher disapproval percentages than candidates with lower poll numbers. Also, even candidates polling at 1% often have approval percentages of 10% or 15%. Not listing approval percentages is a good way to make people think that nobody likes the low-polling candidates.

Selective Release of Polls

Scott Rasmussen does a national poll every day and posts it on his website. You can bet that the large media corporations like NBC and FOX are doing polls every day as well. I don't know who SurveyUSA is or who owns it, but I suspect they are not releasing every poll they do. Of course, when you don't release every poll, you can withhold those polls that show higher poll numbers for Ron Paul.

The Horserace Effect

Over in the democratic primary, this poll says that Barack Obama is polling at 33%, only 14% behind Hillary Clinton. The 33% is out of line with other polls from South Carolina. SurveyUSA is trying to portray Obama as having a chance to beat Clinton. If people believe that, they are less likely to cross over and vote for Ron Paul. And SurveyUSA wants us to believe that Ron Paul has no chance to win, even though most polls show that Paul is as close to Giuliani as Obama is to Clinton.

What Ron Paul Needs To Do

In ancient Greece, demagogues often corrupted democracy. What we have today is much worse; rigged polls. Ron Paul is running a brilliant campaign, but he is going to lose the election if he doesn't hire a fulltime pollster and challenge every single poll that contains scientific bias. He should be putting out a press release every day on the subject and commissioning his own polls. In 2002, Ed Thompson lost his bid for Wisconsin governor because of rigged polls.

ED THOMPSON DEFEATED BY THE POLLS

December 7, 2007

Political Theatre

LRC Blog

LRC Podcasts