Exit pollster Mitofsky: Too many assertive Democrats were polled

[nodictators]

In his appearance on the PBS NewsHour last Friday, Warren Mitofsky tried to explain the exit poll "errors" (emphasis added)(mods: this is a public domain article):

WARREN MITOFSKY: Well, we interviewed almost 150,000 people nationwide on Election Day. We interviewed in every state but Oregon, since they don't have any people at the polling places, and we also interviewed a national sample of polling places.

TERENCE SMITH: Why did the early numbers show Senator Kerry ahead?

WARREN MITOFSKY: Well, Kerry was ahead in a number of the -- in a number of the states by margins that looked unreasonable to us. And we suspect that the reason, the main reason, was that the Kerry voters were more anxious to participate in our exit polls than the Bush voters. That wasn't the case in every state. We had a few states that overstated the Republican margin. But for the most part, it was Democratic overstatement for the reason I just gave you.

TERENCE SMITH: So you're saying that some Bush voters would come out of the polling places and simply decline to participate; if so, why?

WARREN MITOFSKY: Well, in an exit poll, everybody doesn't agree to be interviewed. It's voluntary, and the people refuse usually at about the same rate, regardless of who they support. When you have a very energized electorate, which contributed to the big turnout, sometimes the supporters of one candidate refuse at a greater rate than the supporters of the other candidate.

TERENCE SMITH: Well, if you thought those numbers were suspiciously high for Senator Kerry, couldn't you correct the sample, as you say in your business?

WARREN MITOFSKY: Well, we recognized the overstatement in the exit polls in mid-afternoon, and we told the members of NEP about the suspicions we had, which they chose to ignore. The correction, in this case, is to wait for the vote returns in those same sample precincts and use that for projections. There were no mistakes in the projections. We were very cautious with them, and none were wrong, even though the exit polls did overstate Kerry in a number of states.
http://www.pbs.org/newshour/bb/politics/july-dec04/exitpolls_11-05.html

Are suspicion and recognition valid statistical methodologies? And what about "Democratic overstatment"? What the hell is that?

If too many Dems were being polled, wouldn't that be evidence that Dems were coming out in droves to vote? Mitofsky is denying reality!

Let's see: Repubs are really timid little creatures that are too scared and/or too ashamed to tell an exit pollster that they voted for Bush. OTOH, Democrats are assertive folks who will trumpet their votes for Kerry and demand to be polled.

Also, Mitofsky claims to be the godfather of exit polling. Really!

[NewYorkerfromMass]

1. the exit poll sample should be weighted to the precinct

and exit polls are weighted to poll "predictive" precincts only.

[trotsky]

2. Apparently Kerry voters were only enthusiastic

in the states where the polls were so wrong.

Elsewhere around the country, in those states that had verifiable voting, exit polls were gosh darn accurate.

[LisaL]

3. Yep. Kerry voters were enthusiastic.

On the other hand Bush voters were apparently ashamed to say they voted for him. Yet they came out in droves.
Get real!

[mike from ri]

17. exit polling was inaccurate in nh

where ther is an auditable paper trail.

[trotsky]

19. Great, let's audit it then!

I for one would like to know if it was all just exit poll error, WHY there was exit poll error.

[mike from ri]

22. yes, nh is being recounted

but the fact is the exit poll discrepancy has not been shown to correlate closely w bbv. nh is an excellent counter-example: had largest discrepancy of all but no bbv. bbv is outlawed there.

[nodictators]

27. And in Ohio the majority of counties use punchcards or op scan

Very few touchscreens there!

Of course, as Bev Harris has noted, all that a fraudster needs to do is get into the database and execute an action query which changes the vote data.

In Microsoft Access, frequently used in election systems, one can execute any query without ever saving it to the hard drive. After the action query runs and the vote data is changed, it can be deleted from memory and it'll vanish into cyberspace, never to be seen again.

[PatsFan2004]

29. Agree that the exit polls in NH were off. I'm from Kerry country (MA)

and the folks from NH are very independent and have voted down income taxes and sales taxes all the time. Those folks have moved away from MA to avoid our taxes and very active government.

[stickdog]

31. Any machine voting method that isn't randomly audited is BBV. (nt)

[petrock2004]

4. so how does that

explain the fact that the exit polls were only off in certain key places? is it only dems in ohio, fl, and nm that were loud about their voting preferences?

what a f**ing crock!!!!!!!!!!!


:mad:

[mike from ri]

24. exit polls were alos off in nh

which has no bbv.

[EST]

34. Link? Please!

[Pallas180]

33. In Palm Beach County there was no mistaking the exit polls

KERRY almost totally...

where did this guy come from - Mitofsky

[enki23]

5. bullshit.

complete, unmitigated bullshit. this is some of the most ridiculous rationalizing i've *ever* seen. it may be that the exit polls were wrong, but this attempted explanation is fucking absurd.

[stickdog]

6. The sound you just heard was George Orwell choking.

Translation:

The exit polls showed Kerry was going to win, so they had to be "flawed." Blaming the anxious, energized Dems -- who turned out in such great numbers that the exit polls couldn't hide the fact that nearly 10% more Kerry supporters vainly cast their votes -- for this "mistake" adds a delicious veneer of anti-utopian irony.

[NJ_Lib]

7. Ya, OK... Do we really appear that...

dumb?

[joeunderdog]

8. "... but entusiasm was down in Paper Trail states." n/t

[robbedvoter]

9. Karl Rove knew the "Unassertive ones"

http://news.bbc.co.uk/2/hi/americas/3987237.stm

Early exit polls quoted by media seemed to give Mr Kerry the edge, but colleagues said Mr Rove indicated right away that they did not tally with his information.

He used his own data to put Ohio and Florida in the Bush column - bringing cheers from the president and his family when he went into the Roosevelt Room and told them"

Ya know, they also said they overpolled women (remember security moms?)
kerry was leading amongst men and women in OH - so they underpolled the hermaphrodites.

They also said Bushies lied (afraid of police, challenges, intimidation, eh?)

me thinks they protest too much....Obviously they are aware of what we know. It's only kerry. DNC &comp that don't show any sign of comprehension

[stickdog]

10. The "huge pool of undersampled fundies" refused to identify themselves

because they thought the exit pollsters were gay. Karl Rove told me, so it must be true.

[bullimiami]

11. bull freaking hockey.

lies and the lying liars that tell them.

[dbonds]

12. Hmmm

If people were ashamed to admit they voted for * and refused to answer, wouldn't that show up in the polls as lots of refusals. And I'm not a statistician but it seems to me if the counties were weighted democrat and lots of dems were voting for bush that but avoiding the poll takers, that would throw the polls to a higher republican turnout and show more percentage to him.

[Straight Shooter]

13. They said Kerry had only lukewarm support; now they say otherwise?

Remember all the stories about how Kerry was merely an ABB candidate? The bush campaign was saying we didn't support him wholeheartedly. They tried to tear Kerry down, they tried to tear us down and demoralize us.

Now they're trying to say we were so enthusiastic, and to such a degree of the voting population, that it skewed the exit polls?

These people act like we don't pay attention.

I'm telling ya, it's all Damage Control. They're getting everyone out there to slam the exit polls. I don't watch television, but I bet they all speak with a calm, reasoned voice that lulls people into accepting what they say.

edited to remove unnecessary commentary

[robbedvoter]

14. They also created the "security moms" - then - oops! we polled too

many women!

And keep in mind: the Moral values malarkey is based ....on exit polls. The ones showing Kerry winning. Go figure!

[Straight Shooter]

15. And the war on Iraq and terrorism number 1 concern based on exit polls

Don't forget that one, either. They're trying to have their cake and eat it, too.

[Gregorian]

16. Hard to believe it happened only in the non papertrail voter areas

And all of that enthusiasm should have been exhibited everywhere.

[mike from ri]

20. nh is paper trail and there was a huge exit poll discrepancy;

the corelation beteen exit poll discrepancy and unauditable machines is not being accurately stated in some quarters.

don't get me wrong: i hate bbv. but the overall pattern has not been documented yet.

[onenote]

18. then why were the senate race exit polls more accurate?

sorry, but Mitofsky's rationalization doesn't hold water when you consider the senate races.

onenote

[AmyCrat]

21. And why were they more accurate in places with paper ballots? n/t

[Eloriel]

23. Balderdash

See this, in particular the graphs:

Important re Exit Polls
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x33103

[artv28]

25. Too ashamed

I know several repubs that are bigots but they would never admit that to certain people like a pollster. Rather than lie or speak how they really feel they would rather avoid discussing it. This does not explain why exit polls have been historically accurate until this election.

[kayell]

30. they were off in FL in 2000 too.

[milkyway]

26. Contradicts what repugs have been saying about how energized their

voters were. Now he says they weren't? And if you're going to use "actual" vote totals instead of your own exit polls, what the fuck is the purpose of the exit polls in the first place? To just confirm that "actual" totals were correct when the two agree?

[JohnnyCougar]

28. what a farce

Rove's Psy-Ops at work...discredit the exit polls.

[Mike W.]

32. Oh, yes, Republicans are so reluctant to talk...

Total Bullshit. No Nazi has EVER been reluctant to voice their opinion. Complete horse hockey.

[Robert Oak]

35. this is fudge

complete BS. The sample rate is 40% of the total population leaving the polls...

the sample is taken by counting, person 1 walks out skip, person 2,
ask and so on.

This is probably where the MIT prof is getting such a low probability the exit polls were wrong...because it is such a large sample size, assuming all precincts were covered by pollsters.

Now, if they screwed up and didn't distribute their exit polls or something, that's another story ...
but exit polling is a fairly established technique, although in 3rd world countries they sample 90% versus 40% (as an additional audit
trail) which since the US is now becoming a 3rd world country...guess they should here also!

They cannot claim that most Republican will not take exit polls and most
democrats will because within their own sampling, they have an additional metric..

plus that conclusion is spurious at best.

I don't believe that Republicans won't answer exit polls and democrats will. They would have to prove that one in another study.

[cthrumatrix]

38. repugs are "church mice"...thats the new excuse....pay no attention

[PatsFan2004]

39. My repuke acquaintances hate the mass media and

would never help them out. Whether that would be reflected in their attitude toward exit polls, I can only guess.

[SoCalDemocrat]

36. This is like the 5th excuse I've heard

The most obnoxius was on Alan Colmes where the pundit claimed Kerry supporters were tipped off as to the locations of the exit polling, and somehow a massive conspiracy was undertaken to ensure that democracts in those precincts were alerted and made sure to fill out exit poll results to spike the counts. The theory goes on to suggest this was done to dissuade Republican voters from turning out in swing states.

Other excuses:

CNN on Nov 3rd suggested they had oversampled "black voters" and that was the reason they adjusted the exit polling data.

Sometime later the story changed to state they oversampled young women voters.

It changed again to say that young females turned out in higher numbers earlier in the day, which biased the exit polls.

The latest spin is as I mentioned earlier, that Kerry folks were tipped off and intentionally spiked polls in swing states.

All of these assertions are patently false and unsupportable by any analysis of the facts.

[Karmadillo]

37. The exit poll of Reps who refused to talk to exit pollers confirms this

explanation is true.