Response From Warren Mitofsky Re: Release Exit Poll Data

[davidlynch]

Friends, I sent a memo to Warren Mitofsky which was copied from another DU posting:

Dear Mr. Mitofsky:

I urge you to abide by the polling standards promulgated by the
American Association for Public Opinion Research by publicly:

-- Releasing data from polls for review by polling researchers to ensure the calibrated results can be replicated by others.
-- Discussing possible sources of error.
-- Describing any special scoring, editing, data adjustment or
indexing procedures (or calibration) used.
-- Describing your sample design.
-- Describing your sample selection procedure.
-- Delivering a full accounting of the final outcome of all "sample
cases."
-- Describing how non respondents differ from respondents.

The data no longer have proprietary value to your subscribers.
Election Night is long over, they hired you to help them promptly
report results. Now it's time to make a transition to public service,
Mr. Mitofsky.

Sincerely,
Dave


Here is his response:


I do abide by the polling standards of AAPOR. See www.exit-poll.net for the methods statements that meets these standards. There is no standard that even suggests I release my clients data. We will issue a report some time in January with our investigation of the NEP data. The raw data from the exit polls and the questionnaires will be archived at two universities, as all previous exit polls have been. I started that trend in 1972. This should satisfy all your requests and those of all the other people who sent almost identical letters.
warren mitofsky

I sent him a follow up question:

Thank you for your timely response. One follow up question: are the university archives available to the public? The raw exit poll data is apparently key to determining whether fraud was committed on election day. There are many folks that would be relieved to get their hands on the unweighted data.

I'm currently waiting and hoping for a response.

[pleiku52cab]

1. Anyone contacted Zogby about his data

He was the one that was so sure Kerry was going to win.

[troubleinwinter]

3. But I don't believe Zogby has exit poll data?

[RaulVB]

5. Zogby says "my polling was very good..."

Why doesn't he gives us the details?

[mattclearing]

17. Republicans should be terrified that Zogby had people at Conyers' hearing

eom

[Clark2008]

25. They did? Good.

I didn't know this.

But, then again, how COULD I have known this. There was no MSM coverage of the event.

:shrug:

[troubleinwinter]

2. Why January?

Why not release the raw data NOW?

What is needed is obviously a subpoena. I do wonder if that is included in Arnebeck's suit (discovery)???

[TexasChick]

6. Yep! My exact question and exact thoughts! They are dragging this out

as long as they have to to make sure that the Chimp is inaugurated! Man, I hope Arnebeck's suit will give us all the info we need from many of these people!

[RaulVB]

4. Send him one more, please

IS NOT HIS CLIENTS DATA, ANY LONGER.

It becomes public domain, since he is sending the data to universities and the "owners of the data" don't seem to object that.

Mitofsky really is very unprofessional, IMO, or would he prefer to testify under oath in a criminal court?

[davidlynch]

13. Mitofsky Replied

I asked:

Thank you for your timely response. One follow up question: are the university archives available to the public? The raw exit poll data is apparently key to determining whether fraud was committed on election day. There are many folks that would be relieved to get their hands on the unweighted data.

And Mitofsky replied:

It is available to anyone paying a small fee to the Roper Center.

January may be too late for the actions planned by Arnebeck, but I'm just relieved that we'll get the data at some point.

[troubleinwinter]

7. My understanding is that the data IS available...

if you PAY for it. Any clue what the cost is? If the info is available for sale, it is seemingly not proprietary to the clients, but rather, for sale.

This Mitofsky stuff just pisses me off no end, considering that his data is supposedly what the 'legitimacy' of the election is measured by, and we cannot get it for analysis!!!

[RaulVB]

8. Myself and others have started threads on this topic

The reports are not that expensive but I did not detect much interest in contributing to the effort of buying them.

[troubleinwinter]

9. How expensive? I haven't seen a $ number.

[RaulVB]

10. "Roper Center" (UConn)

I contacted the "Roper Center" one of the institutions where Mitofsky will send the data and they were talking of different ranges for different reports.

Starting from 150 dollars up to 350 for "non academic" buyers. That's what THEY told me.

[troubleinwinter]

11. Oh. I had thought that one could purchase the data...

straight from Mitofsky now, before his placing it in archive next year. Looks like I was wrong on that?

I still don't 'get' why it is being held back at this time if the costs were paid. If it is to become public later, why not now.

[Igel]

12. There's a demand now.

Therefore someone might buy it and he might make some more $. This is his income. (Did *you* ever turn down a bonus?)

[shiina]

14. Subscribers to Mitofsky Info. - possible source of data?

From the website http://www.exit-poll.net/

Subscribers for Election 2004 Exit Polls: Edison Media Research and Mitofsky

Newsweek
The New York Times
Los Angeles Times
USA Today
The Wall Street Journal
The Washington Post
Chicago Tribune
National Public Radio
Atlanta Journal Constitution
Baltimore Sun
Boston Globe
Cleveland Plain Dealer
Dallas Morning News
KCRG-TV Cedar Rapids
Milwaukee Journal-Sentinel
Newsday
New York Post
NHK - Japan National TV
NY1-TV New York 1 News
Raleigh Observer
RAI - Italian National TV
Sacramento Bee
St. Louis Dispatch
St. Paul Pioneer Press
State News - Columbia SC
WCBS-TV New York
WCVB-TV Boston
WHDH -TV Boston
WNBC-TV New York
WNYW-TV New York
Detroit Free Press

[senegal1]

16. hey there didn't the Baltimore Sun run the first MSM article using

the word fraud? Why not ask them to buy it and look at it?

[newyawker99]

22. Hi shiina!!

Welcome to DU!! :toast:

[althecat]

19. $750 per state... is what Edison Mitofsky quoted me or $10k for the lot.

They also said it would be cheaper once the universities had it.

The key to getting this released as suggested below is to get the NEP members to release it these are ABC. NBC. CBS. Fox. AP and CNN..... the subscribers don't own it these six organisations do.

[Niche]

15. Ques.: earlier post? Said he will be at ? University Mon.

Talking about another election... i'm burnt out can't remember i've read so much today. but is he releasing data there?

[Paligal]

18. Why January? Does he want to answer that?

Why is he waiting until January to release the information? This would suggest partisanship, as that would be after Bush's inauguration, and he knows it. I see no reason why he needs to wait. He has the data now. What possible motive does he have for holding on to it? Answering our questions after the fact is lame. It's like asking someone if there is poison in your food and them saying, "I'll tell you after you eat it".

Aargh. F them!!! All of them, Blackwell, Mitofsky, Rove, the whole stinking, coy, game-playing with democracy lot of them.

[nickdw]

20. a post I made a while back about poll transparency

Thu, 7 Feb 2002 00:38:32 -0800 (PST)

Subject: Americans coming to their senses

How can we be sure Bush's popularity was 92% to begin with?

I worked for a telecenter that did political surveys. The numbers we dialed were random but from databases of phone number AREA CODES and PREFIXES used again and again, so I guess there it depended on your definition of random. On some jobs the database was uploaded by the client so the survey was only minimally handled by the third party (telecenter).

So what about demographics? In this article I see lots of percentage points and 1,545 Americans. Counties? Military personnel voting? Let's see if I'm a Republican think-tank and I want to influence the survey I'm putting out wouldn't it make sense to instruct the telecenter to only dial conservative counties where 80% or more voted for Bush to get that 92%-80%? This is exactly what happened at the telecenter I worked at. Who says 1,545 was the running total? Perhaps it was 3,000+ originally and surveys were tossed to favor a certain outcome before going public?

We need double-blind surveys conducted by independent third party telecenters with proportionate dialing of demographics, including ALL results, with no surveys getting tossed on the backend. Transparency in surveying is crucial to accurate polls.

This way you might get 60%-50% favoring Bush, as I suspect the real
numbers reveal or will soon enough.

BTW, I don't know what "weighting" is but the 2000 Census has a margin of error by a few million people.
http://www.internetional.se/toft/missingmillions.htm

This numbers shifting thing is everywhere and not just in auditing and
accounting.

"How the Poll Was Conducted

The Times Poll contacted 1,545 Americans nationwide by telephone
Jan. 31-Feb. 3. Telephone numbers were chosen from a list of all exchanges in the nation. Random-digit dialing techniques were used so that listed and unlisted numbers could be contacted. The entire sample was weighted slightly to conform with census figures for sex, race, age, education and region. . ."

[tritsofme]

21. And every polling outfit was in on it?

But come election time they start putting out their "real" polls.

Even when he was at 90% approval, he reelect numbers were never anywhere near that high, and that's what I always found signifigant about that bounce.

[newyawker99]

23. Hi nickdw!!

Welcome to DU!! :toast:

[k8conant]

24. If those raw data are the clients' data, let's get it from the clients!

Maybe at least one of those papers on the list will release themt. Apparently Mitofsky is claiming he won't because the data are the clients'.