Skip to comments.Polling: The Good, the Bad and the Ugly - Final: Don't Trust Exit Polls and Notes on Tracking
Posted on 11/01/2012 8:57:34 AM PDT by GOPFlack
Polling: The Good, the Bad, and the Ugly Final A Warning on Exit Polls and Notes on Pre Election Weekend Tracking
For years I have been on a personal crusade against exit polls. It started obviously in 2000. First, back then they were conducted by the now infamous conglomerate VNS - a pooled resource of various networks and news outlets. In those days we were used to multiple leaks of exits starting with the early morning data running in roughly two - hour increments until late in the day. The 2000 election day may have been the most nerve wracking of my life. First, they bounced all over the place. Early in the day, Bush was ahead in PA by a couple but vastly underperforming in Florida and some other states. While I was just looking at them and worrying, the Democrats were doing something far more proactive with the data. In the early afternoon, I heard Jesse Jackson and others working with the Gore campaign had a conference call to further dial up the vote in Philadelphia, St. Louis and other urban areas where they needed a bump. Furthermore, they were using the exits to pump up their ground game in general by showing that Gore could in fact win despite the perception going into the weekend.
In essence what they did worked. Bush lost PA, faulty sampling led to an early call in Florida that cost us thousands of votes, and we lost multiple competitive Senate races all because of the ability to target their ground operations (in many cases like Philadelphia and St. Louis to commit outright fraud). Some of this was the Bush campaigns fault as they went to sleep over the weekend while Gore barnstormed the country and ramped up the massive urban GOTV machine that Clinton had started putting in place in 1998 to save himself from impeachment. We were totally flat footed. I remember a worried supporter asking Don Evans about ballot security the night before the election and the answer was not terribly reassuring as to how seriously they were taking what was going on in the urban areas.
I recount this story because this was the first time exit polls were used as a weapon, ie. the early calls in states that were actually Bush states, etc. by a media that couldnt resist helping Gore. But on balance, the numbers themselves werent too far off. In the 2002 midterms something changed. The 2000 exit polls made the Democrats and their buddies in the media realize that exits were a potent election day weapon. In 2002, the exit polls of course began leaking early in the day in three critical Senate races NC, NH, and CO. To say they were horrendous was an understatement. They showed Liddy Dole and Wayne Allard getting trounced. It turned out they were absurdly wrong by miles. In fact, I went to dinner absolutely certain that Allard and Dole had lost and Sununu was most likely a loser as well. At about 9 or 10, a friend made a phone call to Allards folks in Colorado. He hung up and said its over, they are having dinner. And then he said, they won, big. Same with the other races though Sununus margin was smaller. At that point folks knew something was seriously wrong with the exit polls. What I suspect was happening and then happened in 2004 (and Dick Morris proffered this theory) is that someone inside VNS was leaking the locations of the exit locations before election day and the Dems had folks in place to salt the exit polls at those locations. Its pretty simple, walk in the polling place, loiter, walk out and then find an exit pollster and fill out the sheet.
In 2004, the same thing happened again. The exits were vastly inaccurate so much so Bob Shrum called Kerrey Mr. President at 4:00 that afternoon and Matthew Dowd of the Bush campaign was forced to hold an emergency call that afternoon to claim the exit data was wrong so as to avoid depressing Bushs turnout in Florida and Ohio in particular. Heck, they even showed Bush losing or barely winning Virginia where he won in a blowout. We all know how that one played out. While VNS was ultimately disbanded and a new system put in place, I am still highly suspicious of what goes on with these polls as they are still run and paid for by the liberal media.
Now, the data is embargoed until 5:00 and supposedly anyone working in the exit polling center is not allowed any communication with the outside world that day (surrendering cel phones, etc.). But I still am not sure how they choose the sampling locations and whether the Democrats still have those locations on election day (I suspect they still do). So, based on past experience and all they have invested in saving Obama, please do not believe any leaked exit polls until we count the votes Tuesday night. And another word of warning the number of states with exit polling have been dramatically curtailed this year. The states that were eliminated were almost all deep red states so you can be sure that they wont be calling a bunch of Romney states early. There was absolutely no justification for this move other than to assist Obama on election night. So sit tight, let them count the votes and above all dont let any of this garbage depress you. Thats what they want to drive people from the polls who would vote for Romney or another GOP candidate.
Finally, a few words on weekend tracks before election day. The party and campaign committees that run them usually come back Monday with very accurate results. They definitely pick up on ground momentum and any changes that might be taking place over the weekend. The use a tight likely screen and usually a sample of 200 or more per state, per night. A departed friend who was one of the top Republican aides in the Senate told me a great story about 1980 (when people actually answered pollsters and all had landlines). He said he got a call from the Senatorial Committee on Monday morning before the Reagan election and his friend said you have got to get over here and see this. He went down to the Committee and said he couldnt believe his eyes on the Senate races. Ones that werent even considered competitive Alabama, Georgia etc. were shaping up as wins. Frank Church and Birch Bayh were going to lose as well. It was stunning and turned out to be dead on. The polls caught the wave that no one saw.
In 1998, a year we thought we were going to see some big gains and actually went backwards, I saw the Monday numbers and it was totally depressing. Even Carol Mosely Braun was closing in on what was thought to be a hopeless re-elect. Coverdell was in trouble (though he won). Faircloth was toast, etc. They were spot on as usual. In 1994, I saw the same numbers when we took the Senate the Monday before Santorum, Grams, and a bunch of others were leading in States we didnt originally expect to win.
I guess this is a round about way of saying to see who is loose and smiling Monday and see who is dour. That will give you a good idea of what is happening. Thanks for indulging me and I hope this has been helpful. Dont fret, work hard this weekend if you are volunteering and go vote to save what is left of our once Federal Republic.
Thanks for the heads up.
Yes, helpful. Thank you.
Put the sound on mute.
Ignore any numbers on the TV screen.
Ignore the reactions of any Democratic pollsters, pundits or office-holders.
Look in the eyes of the MSM talking heads on election night.
If they look like they just came back from a family funeral you will know what happened.
Thank you for taking the time to write these items.
Have you observed/are you aware of the phenomena of polls tightening in the week before the election, only to open up again the day or so before?
Do you think this storm will have an impact?
Q1 - Yes, usually over the weekend. But I’ve also seen them tighten on ground game and momentum.
Q2 - Don’t know. Not sure it makes much difference. But, it could impede the Philly ground game. They got a late start and it takes buses and fuel. Something in short supply in that area right now. Just a hunch though.
Thanks for the ping and for doing this series.
This has been a great education. Thank you.
This is an incredible series of posts you’ve given us. Thank you!
Another thing: we’re seeing an avalanche of polling from Democrat-aligned pollsters trying to tell us that Dems are leading or tightening all over the place. And yet Rasmussen, from what I’ve seen at least, has shown good numbers for our side in these last few days. Do you think this is a coordinated effort by the Dems to excite their base, while trying to depress Republicans into staying home?
Yes. If they don’t do that and even more people stay home, they could start losing down ticket races that weren’t even considered in play. A defeat turns into a rout. (I hope).
The best way to not get down over exit polls is to be too busy working at the polls, or on GOTV, to even know what the exit polls say. I worked from before the polls opened, until they closed, in 2004. I never had to stress over the exit polls, because I never heard about them.