« More on the Berkeley Report | Main | FAQ: Questions about Exit Polls »

November 24, 2004

Stones Cry Out on Exit Polls

A blogger and San Diego State graduate student named Rick Brady posts his own exploration of the exit poll discrepancy.  Brady plows a lot of ground that will be familiar to regular readers and (at risk of ruining the surprise ending) ultimately finds MP's take on the issue.  Nonetheless, he reports some interesting interesting information obtained about who has access to what and when we might expect a full release of raw data. 

First, he picked up a bit of color and information from Liz Doyle of Edison Media Research (the company that ran the exit poll along with Mitofsky International):

Ms. Doyle shared that she has been inundated with calls and e-mails from professors and bloggers demanding data because they are convinced there is some kind of conspiracy...I could tell Ms. Doyle had had it with these conspiracy theories. The bottom line was that everyone wants access to their unweighted data and methods for independent review....

Ms. Doyle politely informed me that the [raw] data I was requesting would be available via the Roper Center and due to the unprecedented demand for the data, the NEP was working as quickly as possible to prepare the data for public use. However, these unweighted data couldn’t be expected for at least three more months

Brady then spoke with Richard Morin, the polling director at The Washington Post and got the following response via email:

The Post was one of the subscribers to the exit polls, like the New York Times, WS Journal, USA Today. We got the results of the final national poll and a few states when it was completed but before it was weighted, so I know that the poll had Kerry up by 3. We do not have all of the states, however, and don't know who, if anybody, saved them. They came to us as PDFs, not as data sets, so we cannot analyze them using SPSS or SAS. I saved the national but I do not believe I saved the four states we bought though my assistant did print them out and we have those copies. My recollection is that all of the states were off by a bit, all had a Democratic bias.


I am working on some more on the ongoing exit poll controversy and -- as time allows (he says, knowing that his family reads the blog too) -- will post a bit over the holiday weekend.

Related Entries - Exit Polls

Posted by Mark Blumenthal on November 24, 2004 at 05:17 PM in Exit Polls | Permalink


How could any media outlet not save the PDFs? It's not like disk space is expensive these days.

Paint me skeptical. I'm skeptical of the premise that the exit polls indicate fraud. But I'm also skeptical of the reasons for not publishing at least the full information that was released to the subscribers.

What gives?

Posted by: Observer | Nov 24, 2004 7:23:58 PM

Ok, so are they holding the data for three months because they think their paying customers should get to write the first articles based on the data? (Reasonable.)

Or are they claiming they cannot produce the "raw" data in less than three months? (Load of crap.)

Posted by: Greg D | Nov 24, 2004 11:05:40 PM

The subscribers paid for pdf's?! As the Brits say "pull the other one, it's got bells on."

America deserves to have answers now, not in three months and IMHO the 3 week lead subscribers have gotten so far is plenty.

Can anyone explain why it should take three months to "prepare the data for public use" which is what Ms. Doyle claims they are doing?

The 4 point national difference claimed above. And average 1.9 NEP has claimed need to be explained.
Poll error, spoilage, mistabulation and fraud probably all play a part. This site's tendency to put it all down to poll error makes me highly suspicious.

Posted by: John Kesich | Nov 25, 2004 12:20:53 AM


Someone had to pay for the exit polls to take place. Those someones are clearly entitled to a first shot at the data.

Like I said, if they want to say "our subscribers get sole access to the data for three months, then everyone can see it", then I'll have no problem with that. It's their data, they can do what they want.

But if they're going to claim they CAN'T produce the data for another 3 months, then I'm going to cry "bullshit!"

Posted by: Greg D | Nov 25, 2004 7:08:45 PM

Actually folks, Liz told me the data could be available from 3-6 months! And that was supposed to be the rushed data because of the hype. This has nothing to do with client access from what I gathered from our conversation. She said it has everything to do with archiving. An interesting question for the NEP and the Roper Center who has the 2004 Primary Exit poll data: When was the 2004 Primary exit poll data made available for public review?

Posted by: Rick Brady | Nov 25, 2004 10:20:18 PM

I might certainly be missing something, and if so, please let me know, but one can copy and paste from PDFs and manipulate the resulting mess in a program like BBEdit, no? And if that's not an option, I know I've hired temps to do data entry, and no project has ever taken three months. What gives?

Posted by: segmentis | Nov 26, 2004 2:08:42 AM

Okay, so I'm not a very good reporter (blogging is my hobby). I should have pressed her on this more. Liz kept referring to someone, a "she", in the singular tense, as if they only have one person keying the data for the Roper Center release.

Posted by: Rick Brady | Nov 26, 2004 2:31:01 AM

Perhaps NEP would need three months if they have only very small number of people for a complete re-analysis and also to try to re-weigh their numbers based upon preferences of early voters vs. late voters according to their polling data and actual number of people who voted early or late. By the way, do polling stations keep track of when people voted let alone pollsters? Otherwise, this type of re-weighing would have to be done with fair bit of guesswork.

Below is a relevant excerpt (slightly rearranged) of Rick Brady's post (a.k.a. MP's take)).

Occam's Razor tells us that the simplest explanation is most likely the correct one. The explanation given by Dr. Parker seems to me to the most likely. [Dr. Richard Parker's] explanation for the discrepancy?

It’s in the method. Early voters vs. late…The assumption has always been that they (the NEP) had a handle on the differences between early and late voters. I have heard that they really did not think that there were many differences (in voting preference) at all. Apparently, Bush voters turned out later than Kerry voters. That surprised the pollsters from what I understand and that is what they are likely now assessing. (bracketed clarifiers [by Richard Paker])

Posted by: Ho-Yon | Nov 26, 2004 12:42:26 PM

If the problem is that it will take time to re-weight the data, then they should release everything now. And then release re-weighted data as they get to it.

Hell, you could end up with people like Jay Cost re-weighting the data for free.

Posted by: Greg D | Nov 27, 2004 4:16:54 AM

Ho-Yon, Rick Brady:

Dr. Parker's feeling that:
"It’s in the method. Early voters vs. late…The assumption has always been that they (the NEP) had a handle on the differences between early and late voters."

This has nothing to do with the FINAL exit polls, which are a complete sample released at the close of precincts. This is a non-explanation to the final exit poll discrepancy. It is an honest mistake on Dr. Parker's part, I imagine.

BTW, I think the simplest explanation to date is GOP voter non-response.

Posted by: Alex in Los Angeles | Nov 27, 2004 6:12:55 PM


I believe the last samples were taken around 2pm EST.

Posted by: Rick Brady | Nov 27, 2004 7:54:00 PM

The comments to this entry are closed.