Media Polls Are Unreliable, Manipulative; I Say McCain By 3 Points - Plus

One of the most overused and unreliable indications of how political candidates are doing on the campaign trail is the avalanche of polling data showing minute by minute what we, the voters, supposedly think or plan to do.

Actually, the polls commissioned by and then reported by the majority of American media outlets in this presidential race are so inaccurate as to be laughable. But we can't laugh because once the media gets the answers or numbers it is seeking - by questionable methodology - it then reports the results as if they are facts.

The truth about most of these polls is that it is so easy to expose their prearranged biases, or the flaws in their methodology, that any reporter or news outlet that uses them without explaining what lies behind them has to be either incompetent or deliberately attempting to manipulate the electorate.

I have written about this before and it has gotten worse not better. The media creates a poll by tailoring a series of questions that are asked to select groups of respondents, who through their voter registration and other factors can safely be expected to respond as the media wants. Then the pre-programmed responses become a story, that themselves push further polls that become more stories.

While these polls are supposedly scientific and accurate within two or three percentage points, they range so widely in outcomes that either some are laughably inaccurate, or, more likely the case, all are flawed. A week ago for instance, Fox News reported that the Real Clear Politics average of 15 polls was showing Barack Obama ahead of John McCain by a 7.4 percent margin.

But that margin was nowhere near as wide as the differences in the individual polls that made up the average.

They ranged from one percent, which actually is a tie and means either candidate could be ahead, reported by the AP and Investors Business Daily, to a 14 percent advantage for Obama. That result was reported by the Pew Research poll which was one point higher even than the absolutely ridiculous 13 point advantage for Obama reported by the New York Times/CBS News poll.

Why do I say it was ridiculous? Because of how it was done.

The New York Times/CBS poll resulted from a minuscule sampling, which itself was heavily weighted in favor of Democrat respondents over Republicans, and also included oddball questions.

In fact, media commentators have traced the false claim that Sarah Palin is a negative influence on John McCain's campaign back to a question in that that biased, skewed poll.

But the real problem with virtually all of these polls is the people who get asked the questions. The owners of the polls claim a certainly reliability based on whether they are asking "registered voters," "likely voters," or as was the case with several polls in the past few weeks including the New York Times, "adults." Adults can be anyone from convicted felons in prison who can't vote, to high schoolers who just turned 18 and probably won't vote.

A poll of "registered voters" is considered less reliable than "likely voters," since in some election cycles up to 50 percent of the electorate doesn't participate at all. So likely voters, people who have voted in at least three of the last four elections, give the pollster a more reliable pool of respondents, since they have shown by past actions that they probably will show up at the polls and vote.

But therein lies the rub. The way that the majority of the pollsters determine the voting patterns of the people they are interviewing is to ask them! How often do they get a truthful response?

Who knows? As often as you can roll snake eyes when you are shooting craps? Perhaps. I don't know, but what is worse, the pollsters don't either. Yet they have the audacity to report these polls as accurate within two or three points.

This could easily be rectified by obtaining lists of what are called 3/4 and 4/4 voters from the major political parties or registrars of voters in the communities where the sampling is being conducted. But the pollsters, especially the media pollsters, don't always do that.

Which could explain how the New York Times/CBS poll put Obama ahead by 13 points last week, when the Gallup Poll, done by a venerable polling firm, put him ahead by only 4 points during the same time frame.

After reading that Real Clear Politics average, and discerning that it was heavily weighted with media polls, that clearly were out of sync with the professional pollsters, I decided to refine the results a bit.

The media polls are obviously self-serving since the media has been horribly negative about John McCain and smarmingly positive about Obama, so I threw out all media polls. Six of them. Gone. I kept the Investors Business Daily since it was the only one that correctly called the Bush/Kerry race four years ago.

Then I dropped the lowest and highest point spreads from the remaining nine polls, just as in done in the Olympics leaving me with seven. I took the average of those. It came out to 3.75, nearly half of what Fox News had been reporting. That is a big, big drop.

Yet even that result is not necessarily accurate since the methodology for several of the remaining polls was just as questionable as the ones I dropped. But it did show that with only one media outlet reporting - the one that had the best track record - and an average that consisted of reports only from polling firms, the outcome changes drastically from what the media was reporting.

Not only are polls themselves inaccurate by widely varying degrees, but news commentators can manipulate their audiences merely by choosing which poll to include in their report. This manipulation is compounded if the commentators, reporters, and anchor people "forget" to tell their audience what went into gathering the figures they are using in their reports.

Since Barack Obama is being sold to the public as America's first black presidential candidate from a major party, the media is talking a lot about the so-called Bradley Affect wherein white people supposedly lie to the pollsters, saying they will vote for a black man, but when Election Day rolls around they don't. I'm sure there are some people in this category, but I believe most people, of all races, are going to vote for the person they consider the most qualified.

I'm sick of this racism charge. I am sick of hearing it, and seeing it reported. There are plenty of black people in public positions in America these days, put there by white votes as well as black votes. John McCain has not raised the race issue a single time in his campaign, but Barack Obama has made it a cornerstone of his.

Personally, I believe Obama has set race relations in America back four decades.

And look what can happen on Election Day and afterward if these polls have been wrong all along, the media knew it all along, and Obama doesn't get elected. Suddenly the media takes these polls that were deliberately manipulative during the campaign, wrong as hell, but manipulative, and points to them to justify its claim of a "Bradley Affect."

"See, we told you. White people are still keeping the black man down. They told us they would vote for Obama but in the voting booth they didn't."

What do you think would be the motivation for doing that? Trying to justify the false claims of their original polls? Setting the stage for the violence that has repeatedly been threatened by Democrat operatives if Obama loses?

Either way the media gets what it wants, and America gets the shaft.

Oh, one other thing. All the polls are suddenly "tightening up," indicating that McCain suddenly is making a last minute surge. It isn't last minute. It has been there all along.

The conventional wisdom is that the people who do the polls have to report something close to the truth a few days before the election so they can make sales pitches about their accuracy to new clients after the elections.

Frankly, any polling firm that has engaged in voter manipulation the way so many have this year will never be a reliable organization with which to do business.

As far as the election outcome, I predict it will be McCain, especially if there is a heavy turnout. I predict he will win by at least 3 points, quite possibly more. Sarah Palin will win too. I didn't get that from any polls, I got it from watching people, and listening to what they are saying.

We'll know on Wednesday whether I was right or not.