Fox News’ “Impeachment Poll”

I was challenged by a friend when I brought up the weighted aspect by this: “Have you taken any graduate level courses on polling Sean? I have.” So I guess you have to take graduate level courses in statistics to be fooled? I don’t know… I don’t put ANYTHING past these #NeverTrumpers any-longer. But I never say, “have you read over 2,000 books cover-to-cover, have a library of over 5,000 books as well as 3,000 more digitally?” — to make a point become true.

BIZPIC has this about the Fox News Poll:

The problem is that the poll was heavily biased because it over-sampled Democrats, thereby leading to skewed results.

According to analysis by the New York Post, a poll weighted for party affiliation would’ve concluded that 44.9% of voters favor impeachment, while 44.4% oppose it.

In other words, a less-biased poll would’ve shown that the majority of voters (55.1%) oppose impeachment.

Braun Research conducted the Fox News poll by sampling a pool comprised of:

  • 48% Democrats.
  • 40% Republicans.
  • 12% Independents.

In reality, registered American voters are:

  • 31% Democrat.
  • 29% Republican.
  • 38% independent.

Any poll that oversamples Democrats will lead to a skewed result. This is exactly what Fox News host Greg Gutfeld said this week when he dismissed the Fox News poll as bogus.

“We got to point out that it’s weighted toward Democrats,” Gutfeld said. “It’s 48% Democrats, 40% Republicans, 12% other. Also, it’s being taken at the fever pitch of media coverage about a single topic, so what do you expect?”

Gutfeld underscored: “We have to remind ourselves how many Democrats and how many Republicans are in this poll, and that these polls have been known to be wrong.”

AMERICAN THINKER continues the breakdown and shows how Rasmussen attempts to correct for such things:

How did the Fox polling unit come up with this number?

The Fox news polling companies interviewed 1,003 registered voters, ostensibly throughout the length and breadth of the United States.  Many polling companies use either all adult Americans (254 million) or registered voters (158 million in 2016) as their universe for polling.  Obviously, the greater the number of potential people to contact and question, the easier a poll is to complete and to skew a result.  In reality, what matters is who votes in an election.  In 2016, 86% (or 136.6 million) of registered voters cast a vote.  A poll of likely voters would inherently be more reliable but more difficult to achieve.  Currently, only Rasmussen among national polls uses exclusively likely voters and they are among the most reliable.

As the issue of impeachment is overtly political, the political make-up of the respondents in any poll is critical.  In this recent Fox poll, 48% of those polled claimed to be Democrats, 40% Republican, and 12% independent.  However, as Gallup points out in its most recent research, 31% of all Americans identify as Democrats, 29% as Republican and 38% as independent.

Ideally, all polls, as does Rasmussen, should strive to reflect that political affiliation dichotomy or as close as possible, considering the difficulty in finding people willing to be polled and be honest in their responses.

Therefore, the Fox poll, with its political make-up of respondents, is manipulated to come up with a desired result.  The issue isn’t the difference between the number of Democrat and Republican respondents, but the gross undercounting of independents and the massive overcounting of Democrats.

Over the past three months to date, in a variety of polls, an average of nearly 84% of all Democrats favored the impeachment of Donald Trump.  Therefore, when Fox uses 48% Democrat registered voters, the poll immediately, before taking into account any other group, will indicate 41% in favor of impeachment and removal.  Thus, to get to 51%, only 20% of those identifying as Republican or independents in this poll would have to be in favor of impeachment.

However, if Fox had used the actual political breakdown of 32% of Americans identifying as Democrats, then instead of an immediate impact of 41% in the result, it would have been 27%, or 14 percentage points less.

Further, over the past three months, polls have averaged 92% of Republicans and 56% of independents opposed to impeachment and removal.  If the Fox poll sample had been 29% Republican and 39% independent, using these average poll results, the final tabulation would have been 44% instead of 51%.

But there would have been no headlines and breathless anchors on CNN, MSNBC, CBS, NBC, and ABC, nor any banner headlines in the Washington Post and New York Times trying to convince the nation that the citizenry is turning on Donald Trump….