Polling: Civitas vs. Reality

It's been interesting to see another Civitas Poll making the rounds.

To see if we should pay any attention at all to these numbers thrown out by Civitas and reprinted by the MSM, let's do a reality check on some of the last polling numbers pushed by Civitas a few days before the Nov 2010 election.

Civitas suggested Democratic Rep. Brisson was down by 20% to his Republican challenger.

Meanwhile, looking at voters who said they are most likely voting in the fall, Szoka’s lead jumps to a 57 percent-37 percent margin.

Of course, this was pure nonsense as Rep. Brisson won with more than 52% of the vote.

In the same press release, Civitas suggested that Rep. Rick Glazier was down by 10% to Jackie Warner mere days before the election. While this was the closest legislative race of the 2010 cycle, this poll -- or the press release "analysis" of it, was utterly divorced from reality.

Here's what Civitas released -- and missed by 10%:

Four-term Democratic incumbent Rep. Rick Glazier is facing a difficult reelection bid as he trails Republican opponent Jackie Warner by 10 percent, according to a new SurveyUSA poll released today by the Civitas Institute.

Yet another case of poor Civitas poll press releases is their complete failure on the NC House 44 race. From their press release of Oct 29, 2010:

Additional numbers reveal that Republican candidate Johnny Dawkins is maintaining his lead over Democratic Rep. Diane Parfitt in the state House District 44 race – comprised of Cumberland County.
(snip)
However, among those who are most likely to vote in 2010, Dawkin’s lead increases to 56 percent-37 percent.

As we all know, Rep. Parfitt kept her seat. Certainly nothing close to a 19% loss.

As you read more about Civitas polls, please keep their pitiful record in mind.

20% here, 10% there -- pretty soon you're talking about a different reality.