Loading ...
HOW SOON WE FORGET
In the course of our daily scooping, we came across the new Mason-Dixon presidential primary poll which showed actor Fred Thompson (R-Tenn.) and Sen. Barack Obama (D-Ill.) leading their respective parties in South Carolina. We were at first quite shocked by the numbers, until we recalled the company’s sordid past in the Palmetto State.
After a little research, we found Lee Bandy’s 1998 article in The State Mason-Dixon’s release of bogus poll numbers during the heat of the gubernatorial election. The stunt got them fired by most of their client media outlets in the state.
Mason-Dixon Political/Media Research Inc. of Columbia, Md., released two different versions Thursday, trying to smoke out how some media outlets that do not pay for its polls have received the results before its paying members could release them.
The two-day survey showed Beasley ahead by 5 percentage points, whereas the three-day poll had him leading by 9 percentage points.
The State, The Associated Press, WIS-TV in Columbia and WYFF-TV in Greenville reported the two- day results. The governor had 44 percent of the vote to 39 percent for Hodges. Seventeen percent were undecided. Given the 4 percent margin of error, Hodges could be ahead. It was based on two days of interviews with 650 registered voters.
WCBD-TV in Charleston and The (Charleston) Post and Courier reported the three-day findings, which had Beasley leading 46 percent to 37 percent. Some 800 voters were surveyed. The margin of error was 3.5 percentage points.
The polling firm explained what happened.
“We decided we’d be a little mischievous,” Del Ali, Mason-Dixon’s senior vice president, said Friday, explaining why some outlets got the two-day poll and others the three-day version. “We’ll see if the snake comes out of the grass and takes the bait.” ["Pollsters' 'Stunt' Lifts Beasley's Numbers" - LEE BANDY, The State (10/03/1998)]
And we were about ready to post this, when ANOTHER poll, this one from Cook Political Report and RT Strategies, was leaked across the TPS tip-line. But, unlike the Mason-Dixon poll, this one shows far different numbers. In the Cook poll, Rudy Giuliani edges out McCain for first – 20 percent to 16 percent – and Hillary Clinton leads Obama by 10 points.
So what’s the difference? Which poll are we to believe?
Well the difference is that one is from a reputable, proven, bona fide pollster (Cook), and one has a history of “cooking” its numbers to fit an agenda. As far as which to believe – only you can decide.
[...] BJ and Adam (McCain’s SC communications duo) try to discredit the poll for their guy. [...]
Relying on any single poll is frought with folly. However, if you take the average of the polls, you are likely to get a pretty accurate result. Look back at the polls from last fall if you have any doubt.
Who are bj and adam and why are the shot people so obsessed with attacking The Scoop? Its like you guys cant post anything without them bitching about something. Guess they think the best way to rise is attack the top but you must get annoyed.
I guess the next round of polling will tell the answer, but would you really be surprised? Thompson is much more of a natural for the state than any of the other contenders, not to mention he’s been rising in the polls but in some faster than others. I’m also not sure about the attack on Mason-Dixon. This was a 3-day poll. Are you implying maybe Thompson is ahead by even MORE? Please, please — you’re setting expectations too high. Let’s all take a deep breath now.
You are comparing a state poll to a national poll. That’s comparing apples to oranges. Come on BJ, you’ve delt with polls long enough to know the difference.
[...] the company that released the poll, Mason Dixon polling, has a history of skewed results. In 1998, the company released two separate polls for the Governor’s race in the state within [...]
I am pretty impressed! Fred has his minions manipulating polls all across the land! His organization must be even better than we all thought! Be afraid, be very afraid…
[...] Repbulican voters versus the 600 by the American Research Group. Additionally, Mason-Dixon has faced scrutiny in the past for issuing polling results which don’t seem to add [...]
Your misreporting the facts surrounding the “Two Polls” story. Indeed, two sets of numbers were released — one based on the first 625 interviews that were completed and the second based on the final total of 800 interviews.
The reason this was done was because a reporter at one of the television stations was consistently leaking our poll reports to The State — which was not a subscribing media client. Note that The State published the first set of numbers even though at the time they did not subscribe to the Mason-Dixon poll.
The State’s primary newspaper competitor, The Post & Courier, was a paying subscriber and they were tired of seeing the poll they were paying for on a print exclusive basis showing up in The State on the same day. To prove that the poll was being leaked illegally, the Post & Courier and their TV partner (WCBD) paid us to conduct a larger sample (by staying in the field one extra night) to ensure that they had some level of exclusivity.
The difference between the two polls was negligible — within the margin for error. But when the Post & Courier numbers differed from what The State was reporting, the conspirator who was violating the terms of their contract by giving those numbers to The State went nuts because they were caught red-handed.
The version of the “scandal” story you are reporting is by Lee Bandy — a columnist for The State. As The State was the beneficiary of the breach of contract, this was their attempt to save face by painting us as having violated some ethical standard. The fact is there is nothing wrong with providing results from two overlapping surveys — one a two-day track and the other a three-day track. The Post & Courier paid for the three-day track, while The State paid us nothing. What ethical standard was violated? Is it ethical to steal the work product that your competition is paying for (and you’re not) and then complain that it ws unethical that you didn’t get the same work product?
As for “losing clients”, the only client we lost was the television station that was leaking our polls to The State — a clear violation of their contract and our copyright. That relationship was over the minute it became clear what they were doing. We signed up one of their cross-town competitors within days.
To be clear about one thing, we are talking about the leaking of our entire poll report. Certainly overall numbers can be obtained off of television broadcasts, but in this case the level of detail (including the lifting of direct quotes from our written analysis) was far more than anything that was on TV. This whole issue was about copyright infringement and it had been going on for a long time. If we had not done this, we would have lost the Post & Courier as a client. This incident was the culmination of a long simmering problem that had to be dealt with.
Final point — this incident has nothing to do with the accuracy or independence of our work. It is about media competition, content theft and the protection of a business.
Thanks for posting my response. Please let me clarify a couple of other points.
1) This incident was almost 9 years ago.
2) The polls were not “bogus.” They were two different sets of real data.
3) Cook Report poll is not a South Carolina poll, it is a national poll.
4) Mason-Dixon has polled for McClatchy Newspapers since 2004 — McClatchy owns The State. This issue was settled.
This is pretty ancient history. It just gets dug up and reinvented by every desperate candidate in South Carolina that is doing poorly in our polls. Unfortunately, McCain and Romney’s people didn’t do their homework, they’re plowing an old field.
[...] from Mason-Dixon. There’s some evidence that Mason-Dixon polls lean Republican (See here and here). Besides, polls are, at best, a snapshot of the population at a given point in time and you [...]