exit polling - would voters lie about voting for Trump?

4 views
Skip to first unread message

Doug Cragoe

unread,
Jan 12, 2017, 4:35:09 PM1/12/17
to Election Integrity
I've heard that voters may have been embarrassed to tell a pollster that they voted for Trump.  That seems possible, if the voter was speaking to somebody in person or thought his identity along with his presidential preference would be known.  But I've read that exit polls are more like voting again with a paper ballot, which seems like it would keep the identity of the voter secret.  

I hope somebody can educate me on how those exit polls were taken.  Another thing - what about exit polling for other elected offices or ballot issues?  Do those show the same "errors" as seen in the presidential race?
Thanks.

Doug Cragoe
Los Angeles  

Verified...@aol.com

unread,
Jan 12, 2017, 5:48:16 PM1/12/17
to Election...@googlegroups.com
Very briefly, yes: exit polling is by confidential written questionnaire, not verbal interview.  This does not, however, rule out response (or selection) bias effects.  It is not hard to imagine a lower level of cooperation as you move towards the right of the political spectrum--it is, after all, consistent with worldview.  But the posted exit polls are weighted, in part to counter such effects.  If the exit poll sample (seen from the crosstabs) corresponds with a reasonable demographic breakdown of the electorate, then unlikely that red shift is rooted in response or selection bias.
 
For the most part, competitive down-ballot races exhibited a red shift as well.
 

Doug Cragoe

unread,
Jan 17, 2017, 10:54:44 AM1/17/17
to Election Integrity
Thank your Jonathan Simon for your reply.   When did exit polling start?   One of the things I've read is that exit polls at one time were pretty accurate - they matched the official results most off the time.  But then they started to exhibit the red shift as election machinery changed.  True?  If that is true, then we have a strange situation where conservative voters at one time were just as willing to participate in exit polls as liberals, but then after several decades they perhaps changed and less of them were cooperative.  Seems unlikely to me.


On Thursday, January 12, 2017 at 2:48:16 PM UTC-8, Jonathan Simon wrote:
Very briefly, yes: exit polling is by confidential written questionnaire, not verbal interview.  This does not, however, rule out response (or selection) bias effects.  It is not hard to imagine a lower level of cooperation as you move towards the right of the political spectrum--it is, after all, consistent with worldview.  But the posted exit polls are weighted, in part to counter such effects.  If the exit poll sample (seen from the crosstabs) corresponds with a reasonable demographic breakdown of the electorate, then unlikely that red shift is rooted in response or selection bias.
 
For the most part, competitive down-ballot races exhibited a red shift as well.
 

Verified...@aol.com

unread,
Jan 17, 2017, 12:15:03 PM1/17/17
to Election...@googlegroups.com
From: Verified...@aol.com
To: Election...@googlegroups.comns
Sent: 1/17/2017 9:04:59 A.M. Pacific Standard Time
Subj: Re: [ei] exit polling - would voters lie about voting for Trump?
 
The history of exit polling in the US is, unfortunately, rather murky. The person who could best answer our questions, Warren Mitofsky, died in 2006 (not that he was all that forthcoming when alive). The premise that, before computerized vote counting (CVC), EPs matched well (and thus were "accurate") is difficult to verify. Yes, the archived EPs do match well in most elections from the pre-CVC era but quite frankly we were not screencapturing EP postings prior to 2004, so we don't know whether or to what degree the EP results we are comparing with official returns were adjusted to match those returns.
 
It could be that little or no adjustment was necessary because there was little or no red shift. It is also possible, however, that the adjustments of the pre-CVC era were just as hefty as those we routinely see in the CVC era. We just don't have any reliable way of knowing. 
 
Exit polling, in one form or another, started in the 1970s (I think), developed by Mitofsky for CBS (again, I think) and then spread to other networks, who initially each conducted its own proprietary poll before they were all yoked together in a multi-network consortium that has since gone by several names.
 
Perhaps David Moore would have more comprehensive (and accurate) information on this.
 
To post, send email to Election...@googlegroups.com. Please review the "Posting Guidelines" page.
 
Please forward EI messages widely and invite members to join the group at http://groups.google.com/group/ElectionIntegrity/members_invite.
 
If you're not a member and would like to join, go to http://groups.google.com/group/ElectionIntegrity and click on the "join" link at right. For delivery and suspension options, use the "Edit my membership" link.
---
You received this message because you are subscribed to the Google Groups "Election Integrity" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ElectionIntegr...@googlegroups.com.

David Moore

unread,
Jan 17, 2017, 1:44:22 PM1/17/17
to Election...@googlegroups.com

To respond to Jonathan Simon:

 

I wrote a brief history of the exit polls in How To Steal An Election, the story of what happened to the exit polls on Election Night 2000, when all the networks followed FOX in projecting George W. Bush the winner in Florida (and thus the next president of the United States) at about 2:15 in the morning, only to rescind the call two hours later.

 

That history includes the following information:

·       Mitofsky was hired by CBS in the summer of 1967 to come up with a better system to project the winners on Election Night than what the network had been using.

·       The first try at an exit poll was conducted by Mitofsky in 1967 in Kentucky.

·       In 1968, he used exit polls in six state presidential primaries and in 21 states during the general election. In all cases, the exit polls were used only in conjunction with actual vote count – where the polls and the vote count agreed, he was confident in calling the winners.

·       In 1970, he lengthened the ballot to include demographic and policy questions, so now he had a useful analytic tool for the networks to discuss what factors influenced the race.

·       By 1973, NBC had begun its own exit poll. By 1980 ABC had followed suit, and all three networks had their own national  exit polls. NBC also augmented the national exit poll with exit polls in select competitive states (the other networks did not do that in 1980). Also, NBC used exit polls alone to project winners, even before the vote count had come in.

·       In 1982, 1984, 1986 and 1988, the networks continued their own exit polls.

·       In 1990, they formed a consortium, called VRS (Voter Research Service), under Mitofsky’s direction, that served all the networks – it had become too costly for each network to conduct its own exit polls.

·       VRS had some technical problems in 1990, but got through the election without too much outside notice of the VRS problems. 1992 worked well as well, VRS providing the information to each network, while each network airing the results as though it alone had conducted the exit polls and made the analysis.

·       By 1994, VRS had been expanded to include the NES – the network consortium that reported the vote count – and was now called VNS (Voter News Service). Mitofsky was out (he formed his own exit poll operation), but his colleague, Murray Edelman, became the head of the exit poll operation.

·       ABC didn’t like being subservient to the CBS cabal, so in 1994, it used its own Decision Team, headed by John Blydenburgh, to project election winners before Edelman did so for VNS. Thus, ABC beat everyone to projecting the winners in several states.

·       In 1996, CBS, NBC, and CNN formed their own Decision Teams to compete with ABC. That’s when Mitofsky came back into the fold as the joint Decision Team Leader for BOTH CNN and CBS (though officially Joe Lenski headed CNN, and Mitofsky headed CBS, they were both located in the same room. I was with them for that election and for the presidential elections in 2000 and 2004).

·       After the 2000 election, VNS was blamed for the bad call at 2:15 in the morning for George W. Bush, though it was the competition among the Decision Teams that caused the problem. Fox called Florida for Bush first, because Jeb Bush’s cousin, John Ellis, was at Fox and was persuaded by Jeb Bush that Florida would go for George Bush. Then NBC followed, and Mitofsky/Lenski followed seconds later for CBS and CNN, and finally ABC reluctantly went along, for fear of being left out. AP never made the call, which was rescinded two hours later by all the networks.

·       In 2004, Mitofsky and Edison Research (where Lenski is Co-Founder and Executive Vice President) took over the exit poll operation consortium.

·       Mitofsky died in September, 2006, and Joe Lenski has remained the overall chief of the consortium, now called National Election Pool (NEP).

I don’t know how accurate the early exit polls were, but I do know that in 1981, Mitofsky projected the Democratic candidate to win, based solely on exit polls. His polls skewed heavily toward Democrats, but the Democratic nominee actually lost by 49.46% (Republican) to 49.38% (Democrat) – roughly 1700 votes out of 2.3 million cast.

 

It’s important to note that the purpose of the exit polls was 1) to help project winners as soon as possible, and 2) to provide data to analyze where support for the candidates came from. So, all the weighting to have the exit polls reflect the vote count was never part of any conspiracy to cover up election fraud, as some people on this website have claimed.

 

Cheers.

David

--
To post, send email to Election...@googlegroups.com. Please review the "Posting Guidelines" page.
 
Please forward EI messages widely and invite members to join the group at http://groups.google.com/group/ElectionIntegrity/members_invite.
 
If you're not a member and would like to join, go to http://groups.google.com/group/ElectionIntegrity and click on the "join" link at right. For delivery and suspension options, use the "Edit my membership" link.
---
You received this message because you are subscribed to the Google Groups "Election Integrity" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ElectionIntegr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

George Klees

unread,
Jan 17, 2017, 4:48:00 PM1/17/17
to Election Integrity
From what I've heard, the first major exit poll discrepancy was in the 2000 Florida election. Gore was ahead by 6.6% in the poll, leading VNS to initially call the state for him, yet the official results ended up so close that the recount debacle ensued. Given all the election shenanigans that went down (hundreds of thousands of uncounted votes, and electronic tampering in Volusia, Brevard, and possibly other counties), it's reasonable to believe that this exit poll was a truer indicator of Gore's vote count than the official results.

Next was in 2002, when the exit polls were so far off that the media didn't even release them. This year had several suspect elections, like the Georgia races for Senate and governor, the Alabama race for governor, and the Nebraska race for Senate. VNS disbanded not long after, and NEP took its place.

Then 2004 was when the serious exit poll analysis began.

Since there was never (to my knowledge) a major controversy over missed calls on the basis of exit polls before 2000, and (as David Moore says) the media began using exit polls to project winners around the 1980s, it seems that the exit polls were relatively accurate until 2000.

However, I've also heard the Votescam authors (Jim and Ken Collier) assert that the exit polls were fabrications. Not sure what people here think of this, since so much of our election forensics is based on exit polls.

On Tuesday, January 17, 2017 at 12:15:03 PM UTC-5, Jonathan Simon wrote:
From: Verified...@aol.com
To: ElectionIntegrity@googlegroups.comns
Reply all
Reply to author
Forward
0 new messages