Should we demand exit polls that are meant to detect fraud

[Goldeneye]

and are sponsored by a nonpartison commision for future elections? The data should not be proprietary and anyone should have access to it. Is this possible?

[tridim]

1. They already are "meant to detect fraud"

Unless of course the fraud is committed by BushCo.

[Goldeneye]

3. I've never read the Edison/Mitofsky mission statement

but I don't think it says anything about fraud...correct me if I'm wrong here.

I do think that the exit poll data we have shows fraud, but if we had exit polls specifically meant to detect fraud (that were non proprietary), that would be a big step up. This would eliminate the nay sayers next time exit polls are off...they'd have no argument to fall back on.

[Time for change]

10. Ditto

Best of all, the MSM would then be forced to cover it, and it would have legitimacy with the bulk of the American electorate. That should prevent something like what was perpetrated in this past election from ever happening again.

[euler]

17. exactly

[autorank]

9. tridim is RIGHT!

Let's see, they detected fraud in the Ukraine. Aren't ours at least as good as there's this year. Of course they are.

'Forensic statistical analysis' is the sleeping giant of this fraud election. This is being done by academics. There will be a consensus and that will be known at some point in the next few months. CM (corporate media) will ignore it but we won't nor will the internet community.

Can't wait. There is more than enough data for this task and we know the outcome based on the reluctance to release source data and the apparent changes in exit polls election night.

Drip...drip...drip

[Goldeneye]

11. I hope you're right.

Like I said, I believe the exit poll data we have shows fraud. But if we had done our own exit polls we would have the raw data right now, or at least people like freeman, baiman, and simon would have it.

[rdmccur]

12. Concur

[euler]

19. No he isn't

The exit poll in Ukraine did indicate fraud. It was designed for that purpose. In the US, exit polls are not designed in a way that is suited to fraud detection.

Goldeneye's query is absolutely on the right track. The next exit poll in the US must be conducted in such a way that guarantees the respondents are a random sample of voters. We don't have that guarantee with the 2004 presidential election exit poll.

View this link for a analysis of the 1999 Ukrainian Presidential election exit poll.

http://qev.com/reports.international.ukraine99exitpoll.htm

Here is a list outcome priorities that are designed into the Ukrainian methodology:

Exit polls are an important resource both in understanding and promoting the political development of a country. They serve three purposes: 1) exit polls provide an independent measure of the integrity of the voting process; 2) they make the balloting process more transparent and give news organizations something to report on election night when public interest is highest; 3) they provide a profile of the electorate which is of substantial political and sociological value.

#1 and the first half of #2 simply do not figure into the design of US exit polls. This is because the entities that pay for the exit poll here in the US, want something different from our exit polls than the Ukrainians from their exit poll. The media pays for exit polls in the US, so they want polls that are 'media driven'

Consider the reasons the exit polls are conducted in the US and compare that to the reasons they are conducted in the Ukraine.

In the US, the exit polls serve at least three functions: (a) They help give producers and reporters a head start in preparing their election night broadcasts and stories, (b) they assist the networks in "calling" winners and (c) they provide a resource to help reporters and the general public interpret the results of the election.

Source: http://www.mysterypollster.com/main/2004/11/the_difference_.html

Two different sets of requirements gives two different exit polls with different designs and methodologies.

For some suggestions on how we need to modify the exit poll design in the US if we want to be able to use the exit poll for election verification is here:

http://election04.ssrc.org/research/InterimReport122204.pdf

some analysts inappropriately attempt to use current exit poll results to investigate whether the results in a locale (state or country) are accurate or whether fraud might be involved in an election. A certain form of exit poll could be used for this purpose, but again the designs would have to be different. To validate results in specific precincts or from particular machines, the designs would have to incorporate larger numbers of interviews with voters leaving the polls for precision. And the stratification strategy would also need to be different, focusing on a combination of machine types and geography, for example, including a larger number of precincts at the first stage. There is little likelihood that the member organizations in the NEP would be willing to support the costs of such a design.

A lot more here:

http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=284055&mesg_id=284055&page=

[seaclyr]

26. OK, so we're told that the U.S. and Ukrainian exit poll systems

have different goals. Does anyone know what the specific differences are (in methodology, that is), what (at least in theory) makes the Ukranian system any better at detecting fraud, and what evidence there is in practice that the Ukranian system is better in this respect? Otherwise we're simply dealing with words and people's opinions.

[euler]

30. I have only been able to find...

...documents that state the purpose for the exit poll and not the specific design of it.

For a thorough discussion of the differences beween European and US exit poll methodology, see the link below.

http://www.mysterypollster.com/main/2004/12/what_about_thos.html

The German exit poll is well suited for election verification. If we would start doing exit polls the way Germany does, I think we could use the exit poll for election verification too. But, it will cost a lot more money.

[euler]

16. No they aren't

[xray s]

2. I would suggest this

First, adopt a universal vote tallying system that has a paper trail.

Second, have manditory recounts of a certain amount of precincts, say 5% of them, randomly selected.

I would think that would do the trick.

[Goldeneye]

4. I agree with both those things.

I still think we should have exit polls done on a non proprietary basis though.

[euler]

20. This too is a real solution.

But we are years away from universal anything. The states can't be forced to go to a common method. But, the states couldn't be forced into changing the speed limit to 55 in the late 70's. The solution was to withhold federal highway money from any state that did not comply with the new speed limit. We will probably have to do arm twisting like this to get common voting methods in all states.

[johnaries]

5. YES!

The current polls are set up to "predict the winner" for the Networks.

[Goldeneye]

6. I don't understand how they can

predict a winner without detecting wide spread fraud if it occurs, but if we have an exit poll that is set up specifically for ensuring a fair elecion, we would be making progress. Of course...when those exit polls turned out to be "wrong"...it would be the same old argument...exit polls just aren't very accurate...*sigh*...

[euler]

21. Browse this CNN report

http://archives.cnn.com/2001/ALLPOLITICS/stories/02/02/cnn.report/cnn.pdf

Predicting a winner is not a high priority among MSM anymore.

One quote of many:

Exit polling is extremely valuable as a source of post-election information about the electorate. But it has lost much of the value it had for projecting election results in close elections... Cease the use of exit polling to project or call winners of states.

I agree, it is counter intuitive that a exit poll can be a "extremely valuable as a source of post-election information about the electorate" but not be accurate enough to predict winners, or verify the fairness of the election. However, a couple quotes might help you understand it better. They helped me.

The first, from this source:
http://election04.ssrc.org/research/InterimReport122204.pdf

The exit poll results, in conjunction with other elements of statistical models used by the National Election Pool (NEP) and the decision desks of their news organization members, are best suited for determining the difference between the two leading candidates

'Difference' is important for understanding this phenomenon. read this article in the Washington Post

http://www.washingtonpost.com/wp-dyn/articles/A64906-2004Nov20.html

Money Quote:

I learned early in my Washington Post career that exit polls were useful but imperfect mirrors of the electorate. On election night in 1988, we relied on the ABC News exit poll to characterize how demographic subgroups and political constituencies had voted. One problem: The exit poll found the race to be a dead heat, even though Democrat Michael Dukakis lost the popular vote by seven percentage points to Dubya's father. (The dirty little secret, known to pollsters, is that discrepancies in the overall horse race don't affect the subgroup analysis. Whether Dukakis got 46 percent or 50 percent didn't change the fact that nine of 10 blacks voted for him, while a majority of all men didn't. The exit poll may have under- or over-sampled either group, producing an incorrect national total, but the within-group voting patterns remain accurate.)

It's starting to make sense now, isn't it ? (I'm not going to dwell here about the fact that this reporter makes note of the fact that the 1988 presidential exit poll was not accurate. I think people are tired of hearing it.)

When you are talking about differences within a sub-population or differences between 2 sub-populations, the exit polls provide accurate information whether or not the total percentages are correct or not. It's all about demographics and demographics are all the MSM is interested in. They pay for the exit poll, so they decide what they want it to be able to tell them. Then, Mitofsky designs the poll from their criteria.

In 2008, we need a exit poll that also gives us what we want too (it can be designed into the same exit poll). Thanks for raising the question.

[Goldeneye]

25. I've seen these arguments before

and I don't think they explain everything. What is the difference between Ukraine's exit polls and our own? And how did the exit polls in Ohio which showed Kerry winning men and women get it so wrong?

Is it just a coincidence that almost every e voting irregularity recorded favored Bush? If it was no coincidence, and there was fraud as you say you believe there was, wouldn't it be reflected in the exit polls? I think it would be and it was.

[understandinglife]

7. We have exit polls that did detect fraud. We have a press that suppressed

...that fact.

The 'exit polls' functioned beautifully. That's why, since shortly after midnight 3 Nov 2004, so many efforts have had to be made to fudge and obscure them in order to cover for the folk who systematically skewed the performance of vapor-ballot devices, optical scanners, and central tabulating computers (along with a variety of other disenfranchisement techniques).

Those efforts to jiggle the exit poll numbers -- along with the oh so much 'ink' that's been spilled to obscure what the exit poll numbers we have mean -- more than adequately inform us that fraud was detected and, if we happened to be Ukrainians, we'd already have re-voted.

No matter what system we have for voting, if we do not have vigilant citizens dedicated to enforcing fair elections and a press that serves the Constitution and not the special interests of one party, we're screwed.

Democracy is not a spectator activity.

WE HAVE NO ELECTION.

Peace.

"Bush Owns It: Let's Purge It"

[euler]

22. see post #19 and post #21

[redpillflux]

8. Good Idea

This would kill the specious argument that Mitofsky's polls "were not designed to detect fraud" and therefore irrelevant for this purpose.

Additionally, the poll (and methodology) would need to be completely open, transparent, and monitored by all stakeholders (voters, candidates) so as to avoid the appearance of pollster bias.

It goes without saying that the financing of such a poll would have to be sponsored outside of the MSM. And "non-partisan" sponsor should be further refined to preclude conflicts of interest -- the last thing we need is something run by "non-partisan" self-serving entities such as the CFR!

Beyond that, however, in order for such an exercise to have any teeth, it seems that legislation would be needed in order to spell out the remediation available should, in fact, said poll call the tallied results into question.

Good idea... Tall order.

[PaganPreacher]

29. So, you want to make federal law...

that regulates the private activity of exit polling. You want that law to include violation of the privacy rights and copyrights of private individuals and companies that do polling (by requiring "transparency").

The techniques and formulae used by pollsters are confidential, because they engage in competition for a limited clientele'.

Is that how we do things in America? Piss all over the 5th Amendment?

"Voters" are not stakeholders in privately-financed polling, nor are candidates (unless a candidate pays for the poll results). As far as exit polls go, voters are merely data sources and observers.

The stakeholders are:
1. The people who pay for the poll results;
2. The people who conduct the polls.

As clients, the media and campaigns provide the check and balance for polling: they will not use an inaccurate polling service in the future, and the polling service may go out of business.

If you have a better way to conduct exit polls, then start a company and offer your services. If your statistical model is closer to the actual final results (not the other way around, as is argued in all of the "fraud" threads), then you will make money and your clients will come back for more. The benefit for observers (non-stakeholders) will be that they have a better source for information than exists at present.

The Pagan Preacher
I don't turn the other cheek.

[redpillflux]

32. Not at all...

The OP specifically asks whether we should demand an exit poll that was designed to detect fraud. I am suggesting that, in order to be trustworthy, such a poll would have to be public, not private. This does not preclude other unregulated, private, confidential, proprietary polls conducted for some other purpose.

In post #13, the OP asks whether there is a downside to conducting polls to detect fraud. Well, there would be if we permitted fraud detection to be privatized. Imagine if polling company X, owned by * "pioneers" were commissioned to validate the actual returns reported by vote recording and tabulation company Y, also owned by * "pioneers."

Voters and candidates are most certainly stakeholders in an election outcome, and therefore stakeholders in any effort undertaken to validate the election results.

[newyawker99]

38. Hi redpillflux!!

Welcome to DU!! :toast:

[Goldeneye]

13. Is there a downside to having exit polls done

specifically to detect fraud?

[jkd]

14. It's probably easier to corrupt exit polls than actual votes?

[rdmccur]

18. The converse is much more likely.

Read Freeman's analysis.

[euler]

24. This is true

[euler]

23. There is to the side that committed it

[euler]

15. Someone will have to pay the difference in cost...

...between the exit poll the media wants and the exit poll we want. So, we should focus the effort on persuading Movon.org, or a consortium of progressive groups to put up the money.

[johnaries]

27. That's a good idea. It would be better if we could get a non-partisan

group to do it, but that's not very likely. However, even if a PARTISAN group commisions one, at least we will have some actual numbers. The opposing group will question them, of course, but that could possibly FORCE a non-partisan poll.

I suggest we contact MOVEON and every PAC we can and make this suggestion.

I'm also going to nominate this for the fron page.

[slackmaster]

28. Nonpartisan commissions aren't nonpartisan

Inevitably the makeup of any body that is supposed to be nonpartisan becomes questionable once that body starts dealing with an inherently partisan issue like an election.

The only way to ensure the integrity of an election is to count every vote once and only once.

[Stevepol]

31. Maybe I'm dense, but what's the difference in exit polls?

An exit poll is an exit poll. You interview people who have voted. You find out how they voted and then you get whatever demographics you want. Then you weight the results by making sure your sample has the same percentages of income, race, etc. as the population that voted. So how would you design one to detect fraud? How would it be different from one that is not designed to detect fraud? I don't mean generalities; I means what specifically would have to be done to design an exit poll to detect fraud?

If you're doing a statistical sample to find out if smoking causes cancer, you just compare the statistics for cancer in two groups: those who smoke and those who don't. The results are either highly suggestive or not, depending on what the level of significance or confidence is, whether .5 or .1 or better.

So do you just label your statistics, "Fraud Detecting Exit Poll" as opposed to "Exit Poll"? It seems to me if the poll is conducted properly, it doesn't make a damned bit of difference what the poll is called. It's an exit poll and it gives a sample that lets you state with a certain degree of certainty what the results are in the target population. The exit polls that have been done, if they were done properly, are already fraud detecting exit polls, and as far as I'm concerned, they have detected it.

[seaclyr]

33. Good question, see #26

euler has a link in #30 to check out re the differences. Haven't had a chance to look at it yet.

[qwghlmian]

37. Ok, Stevepol - you asked

Here is a "fraud-detecting poll":

You put a pollster in every precinct, sample every 10th voter coming out, ask them whom they voted for, and do minimal weighting for tiny adjustments for demographics where appropriate.

Here is US "non-fraud-detecting poll":

You put a pollster in select (few) precincts, sample every 10th voter coming out, ask them whom they voted for plus all kinds of other questions, total the results, then, after all the results are collected, shift the results so that they match the actual election results. This would compensate for any clustering skewing effect caused by picking just a few precincts and make the data that you're after (such as how many X voted for Y, and what do X hold more important than Y etc) accurate.

Why would the second type of poll's raw results not reflect correctly the official election results? Because the variation between different precincts in the same state, even with a couple of miles distance between them, can be enormous, and if you look at the precinct election results in the US you will see that if you look at a precinct or a few, you would not be able to predict the results of the election in the state with any kind of consistency. The way MysteryPollster explains it is the "marbles" analogy. If you put 50% red and 50% green marbles in a jar, then shake them vigorously to distribute them evenly, then take 100 out, you would get pretty close to 50 red ones and 50 green ones (maybe 48 and 52, or 54 and 46). But if you put in 50 red ones, then 50 green ones, give it only couple of shakes, then take out 100, you can easily get 30-70 or 80-20 results.

Why would you run the second poll and not the first one? Because of the cost. The cost of the first type of poll is hugely more than of the second type. But no one is stopping anyone in the United States from ponying up that money, organizing a couple of hundred thousand pollsters on the election day (I think there are at least a hundred thousand precincts in the US) and doing the "fraud-detecting" poll.

[davidgmills]

34. absolutely

They are the best check on election fraud.

Like a quartz watch to keep track of the time the grandfather clock (actual tabulation) is gaining or losing.

[berniew1]

35. Seems we need lots of checks; 2000 & 2004 were not democratic elections

[Carolab]

36. Schools/universities can conduct the polls.

Students already do in some locations. The problem is the polling information is then uploaded and manipulated once it is digitized and the electronics are under control of groups like the NEP/AP which are out of the public's control.

Polling data needs to be collected at and publicly available from multiple hosting sites that are secure and not subject to manipulation, then independently consolidated and disseminated. The results should not be available until after the election is over and the true, final results are tallied.

[anaxarchos]

39. It's not that different...

If you increase the sample size, you increase the statistical resolution. As far as methodology goes, you are still looking for "key precincts", just like the old days (and still to this day). The mix (or 'weighting') is still going to be "black art" as well. Worst of all, you refine by comparison to actual tally. So you confirm the tally (how?), adjust the poll, get to the next election and if you are "wrong" then you are back to "how many angels dance on the head of a pin?" issues again.

In truth, it is all back to the same as today with apologists screaming because "something happened" this time (fill in the blank with the excuse of the day) and those who think there was fraud arguing that all deviations couldn't go in the same direction.

The debate is an academic "red-herring". Exit polls will never be accepted as prima facia evidence of fraud in a political setting like ours. Better verified voting....

This "issue" today is more about casting doubt on the exit polls in 2004 rather than doing something "better" in 2008 or 2012.