Blasting RFK, Jr in Salon (Farhad Manjoo hit piece, in continuing series)

<http://www.salon.com/news/feature/2006/06/03/kennedy/index_np.html>

Here's the guy who set himself up prior to 2004 by writing election protection pieces, so that he can bash the same after the election. He calls RFK's piece "filled with distortions and blatant omissions" (even though Kennedy's article is already quite long)

ER has previously discussed Manjoo's surprisingly strident attacks on Mark Crispin Miller's book here: <http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=401618>

Manjoo belabors "dozens of experts" that he doesn't name who did not conclude that 2004 was stolen. But that, at the very least ignores the context here.

The most important context is this: It has taken almost 2 years to get this information published in the MSM except for Hitchen's Vanity Fair and the work of Mark Crispin Miller in Harper's and elsewhere. Under this kind of resistance to press freedom (since freedom's only needed to publish things people will disagree with) we should expect that only a minority of whistleblowing citizens or reporters will "get it" if our perspective is what we read in public newspapers, and that many others will be afraid to look, and if they look they will be afraid to think, and even if they think it, they will be afraid to say it, and even if they say it, they won't want it printed publicly.

I'm asking all critics of RFK, Jr or anyone else to say what you want, but please defend (at least) the right and the need for this debate to be in the press, and don't allow charges of "distortions" and "omissions" any chance of being read by anyone in the public to mean that the debate should not occur, that the piece should not have been published, or that "false" arguments should not be made.

Totalitarians are the ones with the greatest commitment to truth, at least in their own mind. It justifies censorship because of the damage that "lies" do in the press. It has come to the point where it's necessary to make clear on what side of the freedom line we are all on.

It's easy: the old quote "I disagree with what you say, but I defend your right to say it."

I disagree with much of what Manjoo says, but I defend his right to say it, so long as he's not interfering with others' rights to start debates that really need to occur in the press. And there is NO FREEDOM whatsoever without the freedom to be "wrong" in many people's minds.

In general, Manjoo continues in the tradition of the other side's de facto rule for election protection: If there's any plausible reason, ignore the arguments that it's fraud. The enormous problem with this is that if someone IS going to attack the election process, and there's enormous motive and stakes to do so, then they are highly likely to choose a method for which there's a plausible innocent explanation, in order to avoid getting caught and make the election seem legitimate. After all, he riggers or hackers want, probably more than anybody, to create an impression that the false is the true. Given that lies must appear as truth, riggers or hackers MUST cover their tracks with plausibility and secrecy or some combination thereof.

Manjoo establishes a profound misunderstanding of elections when he cites what he thinks is a plausible explanation for an election irregularity, and then MOVES ON, as if the plausibility of the explanation proves that the event is not fraud. It most decidedly proves nothing of the sort.

DOES NOTHING to Resolve the Issues

It's easy to describe my attitude toward Farhad Manjoo's response to RFK Jr.'s Rolling Stone opening-of-the-public-debate on election protection: "I disagree with what you say, but I defend your right to say it." That should be part of the response of any freedom-loving person, for without the freedom to be wrong there's no freedom at all. Totalitarians highly value their intense commitment to "truth" and commitment to stamping out "falsity." Sadly, we are hugely behind in this public education area, and the consequences are apparent all over the place.

That being said}, Manjoo continues in the tradition of one "school's" de facto rule regarding election protection: If there's any plausible reason for it, ignore the arguments, facts or inferences that it's fraud. (or even that it's just an inaccurate result, though no one wanted it to be inaccurate specifically).

The enormous problem with this is that if someone IS going to attack the election process (and there's enormous motive and stakes to do so) then they are highly likely to choose a method for which there's a plausible innocent explanation. They would do this of course in order to avoid getting caught and make the election seem legitimate.

An election rigger or hacker wants, probably more than anybody, to create an impression that the false is the true. Given that lies must appear as truth, riggers or hackers MUST cover their tracks with plausibility and secrecy or some combination thereof. They are going to steal votes in ways that appear, to Manjoo and election officials in this particular school of thinking, to be innocent "glitches" or random glitches were both sides got hit about "equally" or, in any event, they would rationalize that "there's not enough evidence to change the result."

Under the Manjoo school of government-thinking, there's a sort of enforced naivete centered around denying that anybody would want to control a county or even the world's sole military superpower and wealthiest country simply by moving some voting machine electrons around.

This naivete is like having a gorgeous wife and denying that anyone would ever cheat "regarding" that wife. It suggests that this enforced naivete is strong enough to overcome both good sense and knowledge of human nature. One could suggest, although I in reality do not with regard to Mr. Manjoo, and be more aggressive by questioning Manjoo's love for his country, just as one might question the love for one's spouse if one was indifferent about the threat of the spouse cheating AS WELL AS NAIVE about the signs of an affair, all of which are readily denied and rationalized as being innocent. (oh honey, just some smeared lipstick, hot weather today...)

Given all of world history and its sagas and hijinks bids for control, the only rational response is to call for a full investigation, and not just by the government, but by a funded group of outside activists, academics and attorneys. The government, in administering elections, has a profound conflict of interest in presiding over the determination of ITS OWN legitimacy, and to then, on top of that, have the goverment INVESTIGATE the bona fides of its own legitimacy via election is just too much insanity to bear, once one is clear about the precise context here.

The key is that all of the plausible explanations Manjoo offers don't add up to anything, because the (most likely) multiple fraud attempts will almost surely attempt to pass themselves off as legit.

The bottom line:

END QUOTE

C'mon guys, c'mon Manjoo wake up PLEEEEEZE: To call something "fraud" means that there is deception going on, making the bad look like the good. For Manjoo to continually write articles as he does reminding us that the fraud "looks like it was good" is the epitome of nonresponsive irrationality.

Until Manjoo and others check out the more accurate contextual frame above, they are doing their country a deep disservice by muddying the atmosphere when we need to see as clearly as possible in order to defend democracy.

let me say what I think is the engine of the "enforced naivete" I'm talking about: the universal attacks on "conspiracy theorists." The fear of making educated guesses about something that is not as yet totally known, or the fear of crediting something not reported in mainstream Pravda publications.

This fear of being castigated as a conspiracy theorist causes (for Manjoo, for example) all of RFK's arguments to disappear from Manjoo's mind the second Manjoo constructs a plausible alternative innocent explanation. This disappearing argument effect occurs even though the "plausible" counter-explanation doesn't have any greater status in most cases, and is not in many cases even suggested to be authoritative. So, what's the force or power that can trump two theories or explanations entitled to a presumption of equal status?

Again, the dynamics of "conspiracy theory" fears. Here's how the "texas two step" of conspiracy attacks works, and why it's so unfair and insidious:

1. A wall of secrecy is set up or exists. (FOr example, vote counting is made secret or trade secret vote counting)

In response to the wall of secrecy this necessarily occurs:


2. Then comes step two in the Texas Two Step: Citizens are denounced as "conspiracy theorists" regardless of whether a real conspiracy is or is not alleged.

Clearly, when you think about it, the attacks on people considered "conspiracy theorists" are ATTACKS ON THINKING PEOPLE ENGAGING WITH SECRECY.

Got that? ATTACKS ON THINKING PEOPLE ENGAGING WITH AND TRYING TO UNDERSTAND WHAT'S BEEN MADE SECRET.

The People are in control in a true democracy. The only question is whether vote counting is any of the people's business or not. Or whether extra-constitutional actions are any of the people's concern or not. If they are a proper concern, then the problem is not conspiracy theory the problem is the lack of information because of improper secrecy. If the concern is not a proper concern of a citizen, then their theorizing is a minor invasion of privacy somewhere.

But I can't see any basis to DENOUNCE people like we see. That's the thought police in action (though all of us have probably used this "conspiracy theorist" attack a few times because it's a powerful way to shut people down)

Because, as John Dean of Watergate fame says, this government is the most secretive in US History, we've necessarily got a lot of conspiracy theorists running around talking about the secrecy. If those "theorists" or educated guessers are suppressed or denounced, there is not a single evil of Tyranny that our Founders belabored to protect us against that we are safe from. On the contrary, Tyranny rushes in like fools rushing in where angels fear to tread, because Tyranny as the Founders understood it, is a form of absolutism where we really believe we are right and really have to protect something good and wonderful from a greater evil. Thus, from that perspective, due process is wasteful and absurd, if you're convinced of your own truth, freedom of speech is irrational because it allows lies, and if you're convinced of the necessity of your own freedom of action, you set up secrecy to insure that what you do will not be interfered with.

It's time to fight back for freedom.

of your post to Manjoo's piece. Manjoo expressed disappointment that Kennedy had wasted an opportunity to make a decent case, by failing to check his facts and sources, and Manjoo is absolutely right. It may be thrilling for activists who want to believe that Kennedy is right about the stuff he is wrong about (and there is much, frustratingly, that he is right about) but it makes the task of persuading people that there really are serious questions about the conduct of American elections that much harder when poorly researched and prominently trailed pieces by a Kennedy fall at the first fence, which, predictably, was the exit poll story.

Why this obsession with the exit poll story? It was an interesting idea back in November 2004 - it was worth investigating - it was worth prodding thoroughly. But it's been prodded and it's dead. OK. I know that many here don't believe me, and I won't rehearse the arguments here. But Kennedy should have, and he should have consulted at least someone other than Freeman and Baiman before sticking exit polls at the front of his story. Even if he'd stuck them at the back, it would have been less crass.

So, as far as I am concerned, Manjoo has provided a much needed gloss to Kennedy's piece. But how much better it would have been if, instead of one lazy, sensationalist piece plus a corrective, we could have had a single, well-researched plea for a serious campaign to right the wrongs of 2004?

But who the heck has done any DENOUNCING? Who are the thought police? Kennedy (and Manjoo) have as much right to be wrong in public as anyone else, sure. But are you suggesting there is something repressive about pointing out mistakes?

making inferences based on available information, because the rest of the data is made secret.

STEP ONE. A wall of secrecy is set up.
STEP TWO. Concerned citizens using available information make educated guesses about what's behind the wall of secrecy.
STEP THREE: The thought police come along to denounce these "conspiracy theorists" who are mere citizens doing the best they can.

If it wasn't for the secrecy, people could come along and give out a cite for where the information missing was freely available.

Ah, but when it's not available, we must guess. An educated guess is a theory. The theory results in the conspiracy theory attack if there's any negative inferences to the theory.

Enforced naivete results. Pollyanna raised to political principle.

People use this conspiracy theory attack usually unwittingly because they disagree and want to "win". But the de facto function of conspiracy theory is to *protect* *secrecy* by punishing the thoughts and arguments of those thinking about what's been made secret.

Probably not intentionally, but Manjoo in part is part of the thought police in his ridicule at points.

He could at least make the point there's no evidence the elections were proper, either. But he didn't make that point, did he?

the electronic ballots or the exit poll data?

What wall of secrecy are you talking about? Do you simply mean that the exit poll data was not made available? That might be unfortunate, but "wall of secrecy" scarcely describes the confidentiality constraints. A great deal of data was made available. Except I think you might be talking about the electronic ballots. In which case I quite agree with you.

And who are the thought police supposed to be?

And sure, in the absence of data, people must guess. I spent a great deal of time guessing last March, using data simulation techniques. Fortunately, more information is now in the public domain, thanks, to some extent, to my own guesses.

I have no idea what you mean by "Pollyanna raised to political principle". And as far as my own case goes I have no desire to "win". I simply want to know what went wrong in the 2004 election. I now know a heck of a lot more than I did. And I have tried to present what I know, because I think it is important. If millions of votes were not stolen by digital vote theft, that is rather important information, not least to the Democratic party.

And I believe Manjoo made the point you require him to make on the first page of his article.

in case you didn't notice, like in the cartoons when the foundation falls out from under a cartoon character and they hang they for a long moment, the whole foundation of your argument has collapsed by your own admission. Now it will have to fall.

1. You state in paragraph 2 of post 101 (link just below) that you “quite agree” with me that the secrecy or absence of disclosed data as to electronic ballots is a "problem". <http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=431251&mesg_id=431529>

You then state post 101 (link above), a couple paragraphs further down regarding the “data” which includes electronic ballots, as follows (asterisks are mine):



So, the “data in the public domain” is your paper which is based to some extent on your own "guesses.”

How is it, anyway, that “guesses” become “data”?

I don't think it's allowable science to "guess" as to data withheld to science intentionally (regardless of the justifications therefore), for these reasons:



Once you admit that the wall of ballot secrecy exists with regard to electronics, it’s all on faith. You can't do science or solid math based on it.

If you can, then please tell me why.

we do seem to have a serious communication problem. I shall blame the Atlantic.

Way back last year, in March (perhaps my fault was not specifying 2005) I wrote a paper in which I used modelled data, using mean signed and unsigned error values from the E-M report to demonstrate that WPE was a poor measure of within precinct discrepancy, and that an alternative measure was a better one, and that use of the WPE in the E-M report might have led to misleading mean values that were being interpreted in a way that was not valid. As a result of that paper (originally posted via DKos), Mitofsky re-analysed his data, and presented the results at AAPOR 2005. He also hired me to run further analyses on the actual data.

This data was not released to me - I am as bound as the data is. However some of my findings are now in the public domain, and I have presented, and attempted to explain them, on DU, and also at AAPOR 2006. These findings are based on real data, not modelled data.

Modelling, however is a useful way of exploring the properties of data where one does not have the actual dataset, but aggregated data points, such as means and variances. A lot can be learned this way, and indeed, some of my analyses of the real data involved data modelling, in order, for example, to compute values expected under the null hypothesis.

So no, "guesses" do not become data, although inferential statistics involves fitting "models" to data, and sometimes simulations ("bootstraps" or "Monte Carlo" simulations) are used to compute sampling variance and probability values. It's one of the joys of being able to use modern computer speeds to conduct inferential statistics.

This may be gobbledy gook, so the take-home message, in English English if not American English is: the findings I have presented were findings in the actual data, including the actual precinct vote totals, and the actual precinct tallies neither of which datasets are in public domain for ethical reasons I have spelled out.

This has absolutely nothing to do with secret electronic ballots, which of course should not be secret, and must be rendered auditable if your democracy is to be preserved.

Salon is not the independent journal people seem to think--it is part and parcel to the infortainment arm of the corporate media.

They have not written a single article on the election integrity concerns, but they've written a lot of pieces dismissing it.

Wise up and cancel.

I challenge you to cite one thing that Manjoo has written, here, or earlier, that dismisses election integrity concerns.

Let's see what Manjoo wishes Kennedy had done:



Does that sound like a dismissal of election integrity concerns to you? Honestly?

it's hard to sort out such mixed messages and give them proper weight. Yes, Manjoo makes the quote you state (i assume). But that does not detract from the overall meaning, which is to scour RFK and diss him.

NO REGULAR READER, not a single one, would have the resources or time to do this kind of stuff, which normally only gets done by opposition research -- the "intelligence" arm of political campaigns to attack their own. So the "bad reputation" Manjoo is said to 'fear' when 'poorly researched' claims get published simply doesn't exist and can't exist in the average reader's mind -- they lack the necessary information to critique it. Even regular rank and file "enemies" don't attack like Manjoo, they have to be funded and motivated enemies. This at least mean that salon's editors grant him te editorial freedom and for whatever reason Manjoo chases MCM and RFK like a pit bull chases a squirrel. I'm sure that his intent is pure.

No, actually I'm not. At the very least Manjoo wants to score points by being perceived as bringing down a Kennedy. There are intellectual "kill" points for that. But Manjoo has this kill instinct for the entire movement, basically. He addresses his attention to the movement only for purposes of firing upon it.

So why does Manjoo go public? Why not castigate Kennedy in personal email? A publicity stunt for both? Not likely since it seems the MSM cited Salon's and Mother Jones (Hertsgaard) dismissal of MCM as a good reason not to bother. A very risky publicity stunt and a deceptive one, at best.

On the other hand,

Febble is a bright talented statistical professional with a lot to offer, and who has spent much time on election issues. In the main she agrees that secret vote counting's a huge problem, and that there must be reform, and the fact that people can readily disagree about elections is itself strongly indicative of their fundamentally flawed nature.

Febble you also tend to get a bit exercised when I explain to you why people get upset at how you and Manjoo approach things with your commitment "to truth" or accuracy.

ELection reform is a little baby that we're trying to get its wings spread, once it flies it's fair game. No individual owns this issue so the idea that there MUST be corrections immediately made to any given person is bogus.

It's more like one's child in a first public performance, literally with a somewhat halting voice from RFK Jr, and we all want this child to fly and then real soon there will be full fledged voice competitions and so forth.

But really, in what is equivalent to the first halting performances of "our baby", the extreme nitpicking critique is not going to be well taken. It's just not.

And then, it's made much worse by an "explanation" that one's "just being honest, we all should be committed to the truth" stuff, while, at the same time, you're not seen as building the movement just critiquing it, however honestly it may be.

With Manjoo in particular attacking every mainstream peek the movement gets, it's like he's playing Whack-A-Mole with the election movement. It' doesn't appear that any of the few cheering Manjoo on are ready to pick up the pieces of the movement and start fresh and accurate. So, on the impression, people smell a rat, rightly or wrongly.

There could be a compromise here: critics agree not to strangle this movement in its very crib, and in exchange for holding off, if the movement gains substantial strength or there's a debate, you'll get a VIP panel-presented invitation.

Manjoo, Febble: True or not, the reaction to perceived falsity is strangling the movement in its crib, even if it's making some "mistakes" in precisely the ways claimed. There's simply no way I can imagine that this constructive criticism is assisting this debate in moving to the mainstream.

MAYBE RFK is wrong on a bunch of things.

FOR SURE Manjoo is hurting this with the MSM, just like he hurt MCM.

I defend the right to speak and say these things, but I must say that whether true or not, it is the experience of a child strangling. It could lead to the end of or a worsening of democracy for lack of any correction. One could rationalize that by blaming RFK (again), but to take any risk like that with democracy without it giving pause, some serious pause, is troubling to me.

You see, when asked "why do you do this?" I don't think it's a sufficient answer under the circumstances to say "Because it's honest and what I believe is the truth." Does the baby not get any more chances? Why can't Manjoo say OVERALL that the issue deserves widespread MSM attention instead of saying that overall RFK wasted an opportunity with falsity?

So now, the upshot of this particular post i've written here is that it seems to me that friends of a young movement, somewhat like parents, will indulge and cut some slack to "the kid" so the kid can learn to fly, then it's open season and fair game. INstead, Febble, Manjoo and others UNLEASH their intelligent minds to attack flaws and interstices of arguments while the movement is still in its baby steps. I've compared that quite unfavorably to child strangulation. But I know that it would be "irrational" to believe this post of mine to be critical of Febble, because after all it is undeniable that several paragraphs above I stated quite positively regarding Febble:



So, there we have it, I can't possibly be considered to have been overly harsh on Febble. I shall now be mystified rather completely if anyone should respond and find this post somewhat harsh on Febble. It's honest, and truthfully indicating an important point: Possibly the movement critics may be right that the exit poll arguments have technical fatal flaws that might damage the movement in the MSM. But if the child is killed in its crib, it doesn't have even a chance to make it.

If there's any reporters who decided not to cover the issue because of exit poll stuff, I'd like their names and numbers and I'll take to them and get the story back on line. I don't think a battle of experts with Ph.D.'s can possibly be resolved in favor of one side or another by reporters. INdeed, that is exactly the conflict or drama reporters are supposed to thrive on to make a dry story sexy.

So again, we're trying to remove the drama/conflict from the movement because ONE side in an unresolvable battle of experts says it's wrong?

harsh on me, and I appreciate your comments. I do think you are reading Manjoo's article from a strategic point of view, and failing to read, literally, what he is saying, but we can perhaps agree to differ on that.

I accept that views also may legitimately differ as to what should be written, when, and by whom. I happen to think that a cause is damaged by bad arguments, and helped when someone who shares the commitment to the cause (as I am prepared to believe Manjoo does, given his opening statements, as well as his past statements) points out the flaws. If the argument is going to have credibility it needs (IMO) to get clear of the junk. If we let people who are against us (yes, us) dismiss our case by pointing to easily refutable arguments, then we have only ourselves to blame. The point, to me, is that the case for election reform, and the case against digital voting systems (unauditable ones at least) does not depend on whether or not millions of votes were digitally stolen in 2004. And every time the argument for the need for reform or investigation appears to derive its force from the claim that millions of votes were stolen (for which the exit poll provides only quantitative evidence) then the argument is damaged. I think that matters.

And I don't actually think we have an "unresolvable battle of experts". Actually, the "experts" (of whom I don't count myself one), as OTOH has shown, agree. The literature by the relevant experts in the field is unanimous that the electoral vote was unlikely to have been stolen, and the popular vote certainly wasn't. The battle is between the amateurs - me, Baiman, Freeman, TIA, the NEDA guys.

I don't know if the media have ignored the story because of the exit poll story. I was appalled that the media ignored the challenge to the Ohio electors. But I am pretty sure that if you want a story to run, it helps if it's properly researched. Bad research sometimes hits the headlines (cold fusion comes to mind) but it doesn't generally do the researchers any long term good. If you want this story to make it into the mainstream, it strikes me that bad arguments are your enemy.

And I happen to think that those who point out the bad arguments while still supporting the core cause are your friends.

Your friend

Lizzie

To clarify, I don't mean affected by DRE theft (we know not many in Ohio since there were few DREs) and I don't mean just digital votes. I mean all votes of whatever technology affected by all tricks of all types at all stages in the process.

Another clarification, I don't mean how many votes can we prove were stolen, I mean how many votes do we think may have been stolen. Since there has been no serious investigation into any of the serious allegations it is not appropriate to use proof as the standard. At this point, sad as it is, the standard should still be suspicion.

Are more than 118,000 votes in play (or half of 118,000 for any tricks that amounted to a vote flip)?

I think that a few thousand votes, the majority of them Kerry votes, were lost to DREs in Ohio, simply because they were in short supply in predominantly Democratic precincts.

I don't think the exit poll in Ohio tells us much at all. Forty nine precincts have very little statistical power. Nonetheless, there isn't a hint of a swing-shift correlation, as the ESI study showed. Which means that we can't use the exit poll shift to quantify theft.

To be honest, I don't see how any realistic quantification can be done. I mean, we could add up the lost votes for which there is good evidence, and we could tweak our thresholds of conviction so that the total could rise to engulf the margin, but I'm not sure I see the point. The scandal that should be shouted from the rooftops is simply is that we don't know whether or not Kerry would have won a fair election. There was no recount in which anyone could have any faith, and there is no way of counting, accurately, votes that were never cast, but would have been if it had not been for the inequities of the way the election was run.

And what this means is that we cannot be sure that the present President of the United states was elected by an electoral college that represented the votes that would have been cast and counted had the election been just. Which is quite outrageous.

So my own view is that counting the lost votes, while an interesting exercise, and possibly even a profitable exercise in terms of PR as long as its presented with care, is a potential PR pitfall. Electoral justice shouldn't depend on whether disenfranchisement would actually have affected the results. Democracy depends on the consent of the governed, specifically the consent of those who voted against the winner to be governed by the winner, and if the votes of the minority, even if it is a genunine minority, are not counted correctly, then the consent of the minority cannot be presumed.

This is why it distresses me that attention has been so focussed on whether it would have made a difference to the outcome, although up till the inauguration, of course, I was still hoping for a deus ex machina to declare Kerry president (at least we had the Ohio Challenge). But even in November, I kept wishing that people would stop arguing about whether Kerry had won Ohio, as though this was germane to the question as to whether he should involve himself in the recount. It seemed to me, and still seems to me, that Kerry should have protested Ohio, not because he thought he might have won, but on behalf of all those poor voters who stood for hours in the raind whether or not they voted for Kerry.

That's why I keep posting links to Shanikka's wonderful DKos rant, Folks don't get it. The thing that is fundamentally wrong with American democracy is not The Wrong Man Is President (that can happen even in a fairly run election), but that a country that prides itself on being a democracy systematically disenfranchises the least powerful of its citizens.

For me "unlikely" does not mean impossible. It just means unlikely. It doesn't conflict with my view that there is legitimate room for doubt.

This is probably a symptom of statisticitis. I think I'd call something "unlikely" if it was p<.05. If it was p<.0001, I'd buy that as as good a certainty as I'm likely to get.

it's that there's no basis to conclude the vote count was proper, or improper, in electronic jurisdictions.

We should also compare an earlier statement made regarding exit polls:

COMPARE: Exit polls are NOT intended nor are they evidence of fraud.

with the current opinion above that

WITH: the majority of of statisticians are unanimous that it is unlikely that election was stolen. (or whatever Febble said 2-3 posts above)

BTW, what's the cite and source for the "unanimous" statement? Is it published?

for unanimous. I simply have not found a peer-reviewed paper that has called the popular vote win into question. Ask OTOH for a lit review, he will have better access to the relevant academic databases.

Really, LS you are making gold out of straw here. I have said, quite clearly, that there is no basis for confidence that the Ohio would have been won by Bush had the election been fair. It is my opinion that the evidence for voter suppression is much better than the evidence for vote switching fraud, and that the fact that we cannot know whether Kerry would have won on a level playing field is a scandal. I devoted a fair bit of time to doing my own bit of research on the Franklin county story. I don't think it matters whether or not Kerry would have won or not. What matters is that a) we don't know and b) that the system was unjust. Both those things are important, whatever the unknowable answer would have been. You can't count votes that were never cast.

Exit polls are not intended to monitor US elections. They can, nonetheless, be milked for evidence of fraud, and I've had a damn good try - so did ESI. We found the cow dry.

In short: the election stinks; the system stinks; the stink probably didn't change the outcome; whether it would have done or not is less important than the fact that we don't know and that poor people were, as usual systematically disenfranchised, in a myriad of ways.

That's my best offer.

Oh, and Bush lied about WMD.

.... that you've described. Is that correct? And the exit polls milked for evidence confirm that? Is that correct?

On edit: can you name the names of all who are "unanimous"? (this is another form of "request for data" when the data is not in written form, yet)

what you mean: "the elections fairly measured the injustice...."?

No, I can't name names. I can only say that the literature on why Bush won the election is extensive and I don't know of any peer-reviewed paper that argues that he lost the popular vote. That, if correct, would make the assumption that he won it among analysts of his victory unanimous. But if you can find one peer-reviewed paper that argues that he lost it, then you'd prove me wrong.

I do not have access to a political science database however, so I may be missing something. Ask OTOH.

Not only do I not know of a peer-reviewed paper that argues that Bush lost the popular or electoral vote, but I do not know of a conference paper or working paper by a political scientist that argues it. Poli sci is a very large field. But I could not say that I can discern any debate on this point. (It may be that political scientists were so disheartened by the initial rash of quantitative nonsense that they turned away too soon -- in which case RFK is doing his bit to make sure that political scientists never treat this issue seriously. Thanks, Bobby!)

That isn't to say that the election is a closed book. Mebane is probably the most prominent researcher working in this field. (It is not my impression that he is overly enamored of RFK's article.)

Keeping in mind that the whole issue, for me, of winning and losing is a false issue. I note you haven't weighed in on the thread in GD-P, that's like #2 on the greatest page right now

<http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=132x2661019>

If the whole issue of winning and losing is a false issue, and if RFK chose to pitch his article as an attempt to demonstrate that Kerry won Ohio, then....

I don't really see anything to weigh in on on the GD-P thread. I support transparent election systems. There are lots of lines of inquiry about 2004 that I've treated here already. It's not as if I'm a newbie here.

I don't know how many you have to have to 'count'

DOn't have any real good ideas as to where such a thing fits or would get published.

Oh yeah, OTOH has read it too, but that's not peer review (or is it)?

It's not national and not exit poll related so it doesn't fit your criteria. But, I don't know where it would be peer reviewed and published if it existed (I"m not saying it doesn't, but i don't remember one)

it doesn't even address the national election. Also, I found it unpersuasive. But you could run it by other political scientists and try your luck.

I was getting the impression you thought Bush definitely won Ohio so I'm glad you clarified that part.

I'm still not clear what you're saying about the unanimous experts with regard to the electoral vote. Are you saying that the experts are unanimous that Bush won Ohio? Taking into account the dozens of different types of irregularities? I'm thinking that's not what you really meant but, if it is, which experts are we talking about? Are they experts who think there is no reason to look into irregularities? Have they looked into them and decided there is nothing there (if so, how and why)? Have they decided there may be something there but have somehow managed to add up the numbers (how would they do that?) and decided there weren't enough votes in play to call into question the 118,000 vote margin? I just don't see how such a belief is supported by the facts so, admittedly, if you can show me that the experts are really unanimous on this point then, to me, it says a lot about the "experts" and adds nothing to my understanding of the facts, but still, who are these experts? Or maybe you meant to say "popular vote" when you said "electoral vote"?

It looks to me (but I'd want to check) that there is a complete absence of political scientists proposing that the popular vote was stolen, and many writing about why Bush won. But there seems less confidence that Ohio was unstolen, although the conclusion of people like Mebane, and ESI are that it probably wasn't, despite manifest injustice. But then a lot depends on what you call stolen. I don't think anyone has any reason to be confident that Kerry wouldn't have won Ohio on a level playing field, even if, adding up the suppressed votes only bridges the margin on very generous assumptions.

But I have not done a thorough survey of the literature, so I can just go on what I've read (and from talking to people at the recent AAPOR conference). OTOH will know more. I'll page him....

I can say, with confidence, that after 2000, a bunch of political scientists wrote articles saying, in various ways, that Gore 'really' won Florida; nothing like that has happened in the wake of 2004.

I have not encountered any article (in any form) by any political scientist who argued that Kerry 'really' won or might have won Ohio. I can't interpret this as certainty about any positive claim, just as perplexity about how to build a cogent case for Kerry.

for one thing, the consortium of media outlets that recounted florida would find it simply impossible to do so now, with the touch screens in place.

DO you agree that makes peer reviewed papers harder to do?

Most of the quant work on Florida 2000 didn't depend on ballot inspection.

Nevertheless, in general, I do think that DREs tend to make it harder to do forensic analysis, yes.

Just based on the margins alone (Florida 2000 vs. Ohio 2004) it is easier to make the case for Gore. Also, the fact that the media consortium examined every undervote and overvote ballot in Florida 2000 and made the data public means that you could just count the ballots and see that he won by a hair. If you wanted to find a more comfortable margin you could look into just one type of irregularity and there you have it.

For Kerry in Ohio you would need to look into multiple irregularities (number TBD) and add up the votes to build a case, which is not an easy task (which you and I discussed a couple of weeks ago). On the other hand, it would not be any easier to build a case that Bush really won since you would face the same difficulties (or even more difficulties -- since it amounts to proving a negative you can't look into just enough allegations to build your case; you have to look into all allegations). So I can buy perplexity about how to make a cogent case. I don't buy unanimous agreement that Bush won Ohio if the election had been fair.

_{Edit: minor wording}

I think (at least I meant) that I perceive a unanimous view that Bush won the popular vote.

I didnt meant (so sorry if I implied) unanimity over Ohio. I think there may be a consensus in political science literature (but not, clearly, in the popular literature) that Kerry is "unlikely" to have one Ohio, but clearly there is much less certainty, even in the consensus. ESI merely concluded that non-response bias was more likely than fraud to have contributed to the exit poll discrepancy.

If there is a consensus in political science literature that Kerry is "unlikely" to have won Ohio then I suspect that it is more of a presumption than a researched conclusion. If there is anyone (just a single expert, much less all of them in the consensus) who's done a serious job of researching all the Ohio irregularities and has some plausible set of arguments that lead to the conclusion that Bush would "almost certainly" (the complement of "unlikely"?) have won in a fair election then by all means bring that paper forward.

done on Ohio. More than that, I'd like to see prosecutions, and those wouldn't depend on adding up the numbers to give Kerry a victory (though it would be nice).

I think we are on the same page here. Strategically, I think it would be better approached as a crime investigation rather than as an investigation into whether Kerry would have won. After all, if criminal intent lay behind the voter suppression, it must have disenfranchised Bush voters as well.

But really, all I ask is for is an honest and dispassionate approach to the evidence. I think we agree!

Allons.

It is very hard to say this clearly. Saying that there is a "presumption" that Kerry is unlikely to have won Ohio may be just about right. Political scientists (even me, despite rumors to the contrary) are not positioned along the barricades ready to attack anyone who makes the case that Kerry should have won Ohio.

But it is important (well, I don't know what is important, but I sure would like...) for folks not to misunderstand what this means. It doesn't mean that political scientists are going to pick up Rolling Stone and say, "Ohmigosh, RFK is right, lookie these exit polls! lookie this Connally race! lookie these lost-voter estimates! why didn't we think of that sooner?" Maybe some will, but I doubt it. On the contrary, I think some will say, "Oh, no, here we go again." Indeed, because political scientists are only human, I fear that they are more likely to be put off by the article's flaws than stimulated to further thought or research.

I can only hope that they don't come over to DU to watch people taking shots at Stewart and Mebane. I mean, really, who needs this?

Manjoo: “Kennedy finds these "missing votes" in the damnedest places. He counts 30,000 voter registrations that were deleted from voter rolls, in keeping with state law, as mostly Kerry voters, though it's impossible to know if those were even real people.”

http://www.easternecho.com/cgi-bin/story.cgi?3505

<snip>

Blackwell has put forth two edicts now, with less than a week left for voter registration in Ohio, that will deny the vote to thousands of poor people, predominantly African-American, registered for the first time. The first requires that all voter registration applications be submitted only on 80-pound paper, and the second requires that provisional ballots be denied to potential voters.

In Montgomery County, where Dayton, Ohio, is located, deputy election board director Steve Harsman told the Dayton Daily News that up to 4,000 voters could be affected by the paper requirements alone, even though "there just is no reason to use 80-pound paper," says Harsman.

Obviously, Blackwell disagrees. The 40-pound paper had been sent out because of an approximate 250 percent jump in new registrations this year, but Blackwell feels that voters who mail the 40-pound registrations in could have them mangled in the mail. That is an excellent reason to disqualify registrations that have already been received by the election boards and have not been mangled.Because if we don't stop those folks, everyone will think they can register with the forms sent to them by the Ohio Secretary of State.

<snip>



Manjoo: “As Kennedy documents, in the months prior to the election, Blackwell issued a series of arbitrary and capricious voting and registration rules that could well have disenfranchised many people in the state.”



Notice his use of the phrase “could well have”. It did disenfranchise voters, and he is a lying sack of shit. Nothing is proof to this asshat.

You seem to like Manjoo's opinion, you seem to think he is honest. It makes people question your opinion.

but to accuse someone of lying because they say "could well have" instead of "did" is to redefine "lying".

at least not the one to which Manjoo refers. RFK was referring to purges of previously registered voters, not to new registrations.

But as long as we are talking about new registrations: has anyone actually followed up with the BoEs to see how many, if any, registration applications were actually rejected because they were submitted on the wrong weight of paper? The Conyers report doesn't seem to come up with any. For that ruling, at least, I find it hard to quarrel with "could well have."

I believe you were quoted in Manjoo's article: several times.

If so, can you now be described as a public figure? Say in the same regard as, oh, Bev Harris? Is that not fair? Or would you rather stay anonymous?

Please, let me know before I proceed further. I'd hate to cause trouble with the mods. Oh, if somehow OTOH misses this, I will be asking again.

Public figure can have a limited scope to it, as well, or it does have a limited scope (one can be a public figure only as to election issues, but not for other purposes)

Allowing one's self to be inserted into the media or "injecting one's self" into the public media is a waiver of privacy on a limited level. This insures free, full and robust debate (though it doesn't always insure "fair" debate because of the respect for free, full and robust)

in theory, just because this is an interesting hypothetical thought line, I don't think the Supreme Court has ruled whether blogging or posting on a place like DU, all by itself, could make someone a public figure for either general or limited purposes.

FWIW i've not talked or written with BeFree about this at all.

....always looking at the constitutional aspects of stuff. <big grin>

No, we did not discuss this before now, LS got that right.

My line of thinking was purely in the DU community aspect. There is a code of conduct required of members here, separate and unequal to US constitutional conduct.

It is in that regard that I asked whether it would be fair to OTOH to proceed with a line of questioning that if DU admin found itself uncomfortable with, I could make a case that OTOH was only being treated as any other publicly declared member of DU, such as Bev harris, or even William Pitt, for that matter. And that OTOH was a-ok with it.

I don't think of myself as "anonymous" here; I simply haven't made a practice of using my name in posts.

Quoting the rules:
You are permitted to offer constructive criticism of my activities as a public figure.

Does that help?

Just so we understand. I would hate for it to be too big a surprise.

Basically, your site has some disturbing info on it when looking at some numbers.

C'ya

I really missed this:

The literature by the relevant experts in the field is unanimous that the electoral vote was unlikely to have been stolen, and the popular vote certainly wasn't. The battle is between the amateurs - me, Baiman, Freeman, TIA, the NEDA guys.

Do you mean Warren Mitofsky and other polling experts...?

The recount was bogus, we haven't had an even barely decent investigation of 'the vote' from any 'relevant experts'. Only what journalists have revealed about Damschroder mis-assigning voting machines (he says because Blackwell drug his feet) and so forth...

After 2000 - a court ruled that all Florida votes had to be recounted and from this we learned that Gore had won the Presidency -- if all votes were recounted, by any standard. Gore's Victory...
<http://www.consortiumnews.com/2001/111201a.html>

If we aren't allowed to use the polling data to say that something stinks in Denmark, then IMO, you aren't allowed to use it to say the election probably wasn't stolen. At least not until Mitofsky *releases the raw data!*

although I may mean other polling experts. But what I actually meant was academic papers on the 2004 election by political and social scientists who include public opinion researchers. But the proportion of those that have specifically considered the exit poll evidence is something I don't know.

A number of studies on Ohio have concluded that it probably wasn't successfully stolen (in other words that it is unlikely that the injustices would have affected the result) but it seems to me that it is by no means cut and dried (hence my use of the word "unlikely"). Sure, the recount was non-legit, but there are other ways of determining whether an election was fraudulent or not, including the study of historical patterns of voting.

As for the question as to whether the popular vote was stolen (i.e. whether several million votes were stolen) then the fact that no peer-reviewed paper is claiming that it was, while other peer-reviewed papers tackle the question as to why Bush won more votes, seems to me to imply unanimity of view that the popular vote was won by Bush. In other words, the multi-million-vote heist theory doesn't appear to have any traction in a field of study that would be expected to give it traction if it had any. In contrast, check out Walter Mebane's home page for papers tackling election 2000, and in particular his paper, The Wrong Man is President.


http://macht.arts.cornell.edu/

Right, to your last paragraph:



You are certainly allowed to use polling data to say that something stinks in Denmark, as far as I am concerned. I used it myself, in the sense that I looked at the polling data, and said to myself "something stinks in Denmark" (although it wouldn't have stunk so badly if the election had looked half way fair in Ohio, and if you used handcounted paper ballots like any other sensible democracy).

Indeed the polling data is extremely informative. It could well have told us that the election was stolen. In fact the reason the swing-shift correlation was done (by me, as it happened) was precisely because, had it indicated a relationship between swing to Bush and redshift, it would have been an extremely strong indicator that at least some of the redshift was attributable to something that was making Bush do better in the count, vote-switching fraud being all too plausible a candidate. The trouble, though, with testing an a priori hypothesis, is that you have to take your lumps. Even a small, positive, but insignificant correlation would have given food for thought. But a robustly insignificant correlation that is actually slightly negative is a very high bar for the fraud theory to jump. And the finding that older technology, not DREs and optical scanners, were, in urban areas, associated with greater discrepancy, while difficult to account for in terms of response bias (iand it could well indicate higher rates of Democratic residuals votes), is also a difficult bar to jump for the hypothesis of massive digital fraud.

And on the other side of the coin is the really rather good evidence that factors that were a priori hypothesised to be proxies for selection/non-response are strongly correlated with redshift.

So, in short, yes, exit poll data can be used to help ascertain whether the exit poll discrepancy was caused by fraud, and the answer appears to be that not only does the data not support that hypothesis, but that the theory is contra-indicated by the data; whereas the selection/non-response bias is well supported by the data. But it could have gone the other way! Both theories were valid, and testable.

However, the exit polls are a poor tool to check on the election in any given state (very low statistical power) and in any case are powerless to detect the kinds of disenfranchisement that clearly occurred, in Ohio in particular.

Finally - the data release! I've said this before, but I'll say it again. The data is released. What is not released is the precinct identifiers, and this limits its usefulness for your purposes. However, the reason the precint identifiers are not released is that it would jeopardize the confidentiality of the respondents. This is not a lame excuse. It is a central tenet of survey ethics (and indeed of any ethical body governing studies with human beings) that confidentiality of the respondents is paramount. It may seem like a stretch, but some of those precincts are fairly small. The questionnaire data contains detailed respondent data, and in some cases the respondents could be easily identified from that data - and their confidential information, including their vote, known. This is not acceptable. To give an extreme, but perfectly plausible (and hypothetical) example: if you were the only black male thirty something in a small white rural precinct in some deep red state - would you want anyone in your community knowing that you had voted for Kerry? Or indeed that you were gay?

This whole thing with the exit-poll data is like playing cards and the dealer handing you only the cards she wants to hand you.

All folks are asking for is the name/number of the precincts that were exit-polled. 30-40 per state. That's all.

It's quite obvious that any further use of the data by those who hold the data, to say this or that about the data, is patently unfair.

And it really leads to a credibility problem, mainly because the truth is being kept hidden.

I trust TIA and RF and Freeman, when they say that what is available to us -- those out of the Mitofski loop -- shows that the exit-polls describe fraud.

I believe Febble has seen the data, and I think OTOH has seen the hidden data, why can't someone like Freeman see the data? It makes no sense, and leads to disbelief of any further statements made by those so use that hidden data to make their arguments.

Its the same idea as requiring links from posters here. No link? It cain't be trusted.

I DO NOT KNOW THE NAME AND NUMBER OF THE PRECINCTS THAT WERE EXIT POLLED IN EACH STATE.

Got it?

Next point:

Yes, I have seen the data for the simple reason that I was hired by Mitofsky to analyse it. Edison Mitofsky staff have also seen the data, as have the owners of the data, the networks. Moreover the questionnaire data has been publicly released.

But no more can be released because IT WOULD VIOLATE THE CONFIDENTIALITY OF THE RESPONDENTS. To do so would be UNETHICAL.

This matters.

I know you are just a minor player and don't have the names/numbers, Mitofski does.

But you miss the point... Mitofski releases the data and then once everyone has seen it, we can have an honest discussion. The discussion so far is one-sided. You have seen it, but we haven't. Therefore the discussion can't be labeled the TRUTH.

anyway. We are having an honest discussion. I am honest.

The only problem is that you can't prove it. It nonetheless happens to be true.

But there is no way you can prove it because the data can't be released in any form that would allow respondent questionnaires to be matched to precincts. Do you really not understand this point?

And even if it was released, would you believe that it hadn't been tampered with?

Would I believe it hadn't been tampered with?

Given the time between now and then, and given the nature of the data and its' importance to certain, very powerful interests, that really is not a question that educated people should ask.

Anyway, you seem to be claiming that I should trust you when you say it is "true". I think you know the answer to that.

so I don't know why you keep asking me questions.

Because you keep claiming to know the inside scoop, and I want to know the inside scoop.

Because the more you write, the more I believe that the baby is gonna grow up fast, is growing fast.

I'll tell you. I've actually told you already. But I'll tell you again if you want.

As you have said, you can't and remain in good graces with the contractor. I dig that.

I place the responsibility with the contractor, and only bug you because you are seemingly representing the contractor here.

Loved that LS reference to you being out of reach of the laws of the US. Because, if there ever is a real open investigation you're free, eh? Smart move. Well planned by the contractor.

Bad for the US, but what else is news?

Still, the POINT IS, in all fairness and honesty, using secret evidence in this public forum is not being fair and honest. Its no big deal, we've been screwed time and again, but the truth always comes out, one way or another.

_{spell check edit}

to data.

That's yet another reason why "lack of evidence" claims are unfair, even if true, in this context. Applying scientific rules of proof under conditions where data is not freely available is illogical, unfair, and misleading, albeit unintentionally.

And whose claims?

Can we get rid of this passive voice please?

This is making no sense.

My bad.

I have nothing against it in principle, but it often leaves me at sea as to who the implied subject of your sentence actually is.

and I have no idea what LS is on about. I am not worried about legal action against me, and I would be no more worried if I lived in the states.

But this conversation is taking a truly bizarre turn. We are not playing some game where one side has an unfair advantage. We are on a forum where we debate what might have happened during the 2004 presidential election and how to stop it happening again. If you don't want to hear information you can't personally check, then put me on ignore. I shall continue do my best to help people interpret the evidence in the light of what is in the public domain.

Lizzie

that violates the rules or expectations of real science.

Even if it doesn't it certainly prevents me from hiring anybody to double check ALL of Febble's work. that's unfair.

what's so difficult about that??

The subpoena stuff was a hypothetical legal analysis of what it would take in order if the EAC or the Supreme Court or somebody wanted to get this data withheld for use in a very important trial about democracy. In that case, I'm suggesting the legitimate rights of survey respondents would or should yield to the public good.

At what point in the above do you lose me?

and you can't double check it. It's why I treble and quadruple checked it myself. And sure, it's unfair. Would you rather not know about it at all?

I agree with you about subpoena - I already said that. I expect you are right that the rights of the survey respondents would yield. But they certainly shouldn't yield unless a court determined that it was for the greater good.

... just like the paperless DREs.

I agree, and is what I have been saying, thank you, "it's unfair".

So, are you going to continue to present unfair information?

I do look forward to the day that I could say: finally, it's fair. But since we will never(?) do it here, I will always label that presentation as unfair. You won't mind, will you?

One side can say that:



But, on the other hand



Thus, so long as you agree with me that the data is limited and therefore unverifiable, claims regarding BUsh and regarding Kerry wins RISE AND FALL TOGETHER. It's the same data.

The difference is not science or argument, it's the party line that is required to be towed, and the thought police enforcing their denunciation penalties from those who think freely.

Under the present electronic system, Bush and Kerry should be Saimese twins in terms of their legitimacy.

And Manjoo can't fairly attack RFK without also attacking Bush and the whole system for hiding the data that Manjoo in part faults RFK Jr for. (though to be fair not every manjoo attack is affected by this, but important ones are)

How can this not be a double standard?

to present unfair information, because failing to present it would, IMO be even more unfair.

You don't have to listen.

.... is a double standard. I don't have to listen, yer right.

But I will read it and debunk it every chance I get.

You see, the "proof" or "truth" ain't always what it seems to be. Especially when it comes from what might be considered a tainted source, and given that many, very smart people have made many, very big mistakes, thos 'facts' behoove a response.

My vote, and my fellow countrymens' votes, are being squished in this whole convoluted process; that is undeniable. So when ANY so called proof is laid out there which supports that squishing, it deserves, aye, even requires, to be squished right back.

In due time, once the major obstacles to clean election counting have been obliterated, I might venture to be more tolerant and not be so biased for, or against half-truths. But in the current crisis we find ourselves at, and at the crossroads we find ourselves, any deterent that tends to kill our forward movement will be met with what force I may acquire and use.

Nothing personal, mind you, but your 'evidence' is questionable and is in the way, know what I mean?

I shall be more than happy to listen.

Except that if all you can say is that I must be wrong because I am a tainted source, then we won't get far.

Sure, I could have made a mistake. I am constantly aware of that possibility as all conscientious scientists must be. But I am not in any way supporting the squishing of votes. I have done some serious research into the squishing of votes and I am totally agin it. I think it is quite outrageous.

All I am saying is that the exit poll evidence is not supportive, and even contra-indicates, the case that several million votes were digitally squished. It says nothing about the case that voters were squished, and I believe that they were.

But I don't think the case that the exit poll discrepancy was due to massive digital vote-squishing holds water at all. Which is GOOD NEWS. You don't WANT your vote squished, do you?

And if you think my evidence is "in the way" - well, truth is occasionally inconvenient. But I'd still have thought it was good news. It means you have more chance of having an unsquished vote next time.

duplicated and reproducible, the claims made regarding that data do not qualify as something that can be considered by *science* as true, or proven.

From the scientific standpoint, one would then be peddling lies, passing off conclusions expressly or impliedly stating that they have scientific validity when the unavailability of the data makes those claims impossible.

Similarly, I can't say "I've got the goods, * stole the presidency!" and then ask Be Free to "verify" the claim, and he/she says "Yup, LS has got the goods, holy mackerel!!!"

Then when others come asking I say "Sorry, I promised confidentiality to my source."

To be clear, these claims can exist as POLITICAL claims, and be argued validly on that level. But then they should not be in any way capable of being misunderstood as statistical claims. Yet, they *must* be, I think, when statistical claims are specifically made, and when the ethics of survey statisticians are being asserted we are undeniably in the area of professional expertise and not the (freer) personal opinion area.

I think this general issue is perfectly clear (disregarding the new concept in the paragraph immediately above). BeFree gets it. Febble's very smart lady so I don't know if it's any longer reasonable for her to not understand this issue. Sometimes stepping back and taking a deep breath does the trick. Febble?

"science" does not "consider" anything. People do.

I seriously do not understand what either of you are on about. But let me tell you some facts on the off chance I will address one of your issues.

The exit poll data belongs to Edison-Mitofsky's clients, the networks. They paid for the data, as any other client pays a polling firm for data. It belongs to them. It is commercial data, in other words. However, pollsters are also bound by the ethics of their own professional organisation, AAPOR, which, like all professional organisations involved in human research, has strict ethical guidelines regarding the release of data to the public that would violate the confidentiality of the respondents. Exactly the same is true in psychology.

The data have not been "released" to me, any more than they have been "released" to any other analyst contracted by Edison-Mitofsky. I was simply contracted to run some analyses. However, some of my findings are now in the public domain - not the data, the findings. And as those findings are of interest to those concerned in the conduct of the elections, I have attempted to report them, and to explain them. And as I actually conducted them, I am in a fairly good position to do so. As OTOH worked with me on developing the appropriate analytical techniques (he does not, I should make it clear, have access to the data) he has also been in a position to explain them

But I am no more at liberty to release data than anyone else contracted to work on it.

But you may note, a colossal quantity of data has been released, and is in a public archive, together with many years worth of similar data, and is a hugely valuable academic resource for anyone interested in the political science, history, or public opinion research. In addition, a "blurred" dataset was commissioned by ESI for Ohio, and I understand similar datasets for other states may also be prepared. The blurring, however is a fairly intricate process, and requires new data collection in order to preserve the statistical properties of the primary data.

Now, if you have any questions, please ask. Please use the active voice, and eschew metaphor. I don't want to have to deal with any more strangled babies today.

Cheers

Lizzie

...and OTOH's.

But since it is YOUR's as it is colored only by you, and without other's analysis of the same data, YOUR coloring makes the use of that data invalid for use in our discussion.

Besides, the rbr explanation has so many holes in it, it no longer even floats. RFK was but another who poked holes in that colored explanation.

(rbr= reluctant bush responders)

don't.

But as you've responded: you claim that "rbr explanation has so many holes in it, it no longer even floats. RFK was but another who poked holes in that colored explanation."

Can you say what those holes are?

... but look at the link below to the KOS article, it does a much better job than I ever could, and so did RFK, and Truth Is All, and....


As far as discussing it, I will be free to do so at my leisure, continuously. I will attack it at every opportunity and will be, as you can see, relentless. Nothing personal, mind you. Except that it is MY VOTE.

article that refutes "rBr" - it is a critique of Manjoo's critique of Kennedy.

Kennedy cited Freeman and Baiman, and I think that Freeman and Baiman are wrong. TIA did not refute rBr, he simply claimed that the samples were random because they were supposed to be random. Unless I missed something.

If you want to debunk "rBr" you will have to be more specific. There is excellent evidence to support the theory that Bush voters were slightly less likely to participate in the polls than Kerry voters. Some of it was presented in the E-M report.

the confidentiality of a few survey folks, then Damn, the possible breach of confidentiality wins hands down. :sarcasm:

Look, courts balance constitutional interests that compete all the time, Febble. YOU LOSE, HANDS DOWN, AND BIG TIME, in asserting the absolute secrecy of this data. You are balancing a mere professional ethic not present at the time of our Constitution against democracy itself, the integrity of which has long been held a "compelling interest" of the government and the people.

As a professional you may have an ethical obligation, whether that ethical obligation has exceptions at extremes or not I don't know but they often do have them in lawyers' ethics.

I trust you'll be secretly pleased, if there were justice, if you got a subpoena compelling you to reveal that information. It would present a dilemma. :dilemma:

But in reality, Mitofsky chose someone to do the work safely beyond the jurisdictional powers of American courts, so you're safe, even if this wasn't just a hypothetical about a subpoena.

I'm so glad you're up front about your work for Mitofsky otherwise I'd get into real trouble as a consiracy theorist in fearing that you were. :fear: But wait, if it woulda been a conspiracy theory to think Febble was on the payroll for Mitofsky, does that mean if she admits to that, it's not true?!? :dilemma:

Dammit, LS, we have been through this over and over and over again. It is wicked cool to think that some "secret data" hold the key to "the integrity of democracy" -- but if those data exist, the exit poll data are not the ones.

they exist, or are supposed to exist, but we dont' get that data, or any of the software used, or anything else.

Yet your tone ridicules "secret data" as if there's no basis to believe it exists. It certainly exists, it just isn't being coughed up.

This conversation is becoming surreal. I thought you wanted the exit poll data.

You don't suppose I, OTOH or Mitofsky have the electronic ballots do you?

use a subpoena. But without a subpoena, the data stay confidential.

A "mere professional ethic"? ?????

Since when was an ethic "mere"?


Edited to add:

You are aware, aren't you, that the confidentiality in question is the confidentiality of the respondents (i.e. not that of "a few survey folks")?

That if precincts could be identified, it would be perfectly possible, in a small precinct, for certain respondents to be identified from the questionnaire data?

As I asked in another post: if you were the only black male thirty something in a small community in a deeply red state, would you like it if the information that you divulged in confidence to the exit pollsters were made public and revealed, amongst many other personal details, that you had voted for Kerry? And were gay?

Jeez.

But constitutional rights yield all the time to other, weightier constitutional rights.

I'm saying SURE, there's an interest in confidentiality here.

THe interests of Democracy and election integrity are greater.

Unless you can make the two not conflict at all, one has to yield. Which one yields? To me the one that is the more important interest is obvious.

Analogy: reporter confidentiality, sometimes courts say they gotta identify their sources, after going through the exceptions to the rule based on the courts' reluctance to subpoeana reporters.

1. I expect a professional like Febble to maintain this posture, as would an attorney with the attorney client privilege. that's ethical assuming the rule is what it's advertised to be, which i've not reason to doubt.

2. HOwever, the confidentiality of survey responders is not even, to my knowledge, at the level of a statute (i.e. it's not law) so it will be hard pressed to prevail against the constitutional level interests of voting.

3. Therefore, I used the word "mere" with respect to survey responders rights, as a relative term. I acknowledge on faith that there is some binding ethic that requires you not to disclose the data.

So, I see Febble here as a helpless actor with respect to this issue, much like i'd be helpless if asked to reveal a specific client confidence. I'd be obligated to clam up, and thus "helpless", I'm playing a role to act out rules of professional ethics that compel me to do what I do.

But there still may be possibilities.

I'd be curious to know if there had been any ATTEMPTS to contact survey respondents and see if they'd consent. It might be hard to do so but I'd look for the existence of largescale honest attempts. I think the survey respondent controls the privilege if there is any, and the waiver of it, and you might find that the survey respondents would, once they knew, gladly agree to waive the privilege as a sacrifice for their country. Or, perhaps they don't care. But we won't know until we ask, will we?

the point is, by (if I assume correctly) not contacting the survey respondents and instead asserting THEIR RIGHTS on their behald WITHOUT CHECKING.

The normal rule is that third parties may not assert the rights of others. That's the main normative purpose behind standing rules that toss cases out of court where the plaintiff's not been legally damaged. (There are other purposes too)

Well, ethics schmethics, an attorney can get disbarred for making important decisions on behalf of clients without consultation. What's the rule for statisticians? This defense might be swiss cheese.

In other words, I don't think survey confidentiality is as bullet proof a way to keep the truth from fully coming out as it might seem. Courts, which are not at play here now, are "truth-seeking" institutions, or are supposed to be, so exceptions that undermine that truth seeking function are to be narrowly construed.

I am completely at sea here. What on earth are you suggesting?

I'm suggesting the survey respondents be contacted to determine THEIR intent. I can't recall whether names are collected at all or not, but even if not, the law has procedures for contacting people. Full page ads, media support, efforts could be made to find them. Certainly you can determine which precincts were sampled, and write a letter to everyone in that precinct asking them to contact Mitofsky using the voter reg list from the County, even if you've got no address or contact information at all.

So contact CAN BE DONE. THe only question is, WOULD IT BE DONE?

well I'm glad you said.

I have no idea whether "it would be done". Who would you envisage doing this thing?

That does not sound so good.

Do you mean it is ethical for him to have the data but no one else?

If you know, what gives him this special ethical status?

First of all, I have not seen the "hidden data," as Febble and I have explained repeatedly. So of course I find it offensive that you continue to assert it.

Second of all, the same data used by Freeman indicate that John Kerry won New York by 29.7 points (IM WPE) or 31.3 points (Best Geo estimator).
http://www.exit-poll.net/election-night/EvaluationJan192005.pdf p. 22
appliedresearch.us/sf/Documents/2004ElectionOutcomeExitPolls.pdf

The four pre-election polls in the final week put him ahead by 15 to 18 points.
http://www.electoral-vote.com/2004/info/allpolls.csv

And the official returns, per Leip, put him up by 18.3 points.
http://www.uselectionatlas.org/RESULTS/

So, you can "trust" whoever the heck you want, but an actual glance at publicly available information might point in a different direction.

The Charles Stewart analysis didn't depend on access to "hidden data," either.
http://www.vote.caltech.edu/media/documents/Addendum_Voting_Machines_Bush_Vote.pdf

Is it possible that you call for more data as an excuse to ignore the data already available?

I did not assert you "did".

But then, Febble did just say she had seen the data!!

Once again your statements are, to put it mildly, confusing and contradictory.

But were you to say something like: yeah, we all ought to see the data -- it ought to be released to someone else. Then at least we'd be taking a step forward to an honest discussion of what it all means. But as long as it remains a hidden secret I don't see how anyone can use that data to say it means this, or, that.

The data seen does point to fraud. It is incumbent upon those claiming otherwise to give up using the hidden data as a basis for their arguments.

OTOH and I are not the same person. There is a large ocean between us. I met the guy for the first time a couple of weeks ago.

Then why does he keep butting into our discussions, and why do you, too?

Of course I can tell the difference betwixt you too, but I can also determine the similarities.

I have to ask myself: why did she respond in this way, with that post? What does it all mean? Why would she think I think he is her? I dunno, it makes no sense.

that because I was hired to analyse the data doesn't mean that OTOH has access to it. But sure, we have collaborated closely on the analytical techniques.

http://www.dailykos.com/story/2004/11/22/19755/343
Rebuttal of Stewart

http://www.notablesoftware.com/Papers/MITvsMercuri.html
Ms. Mercuri rebuts Stewart



CHARLES STEWART, III, is Professor of Political Science and Associate Dean of Humanities, Arts, and Social Sciences at MIT. For the past three years he has been a member of the Caltech/MIT Voting Technology Project, a multi-disciplinary effort to study the electoral system in the United States and to foster new research into the appropriate application of technology to voting. Within that project, he has focused on measuring voting machine performance and early voting measures.
++++++++++++++++++++++++=

Seems Mr. Stewart has some serious faults. One, he's taking money from HAVA and his analysis of the link you posted was a whitewash of electronic voting. Too, the data he used in that report was corrupted.

Its academics like Stewart who have gotten us into this mess.

Both your links predate the paper I linked to. Right?

Yes, right, all the academics agree with you, except for the ones who don't, who must be corrupt. No need to read their papers. No need to think. Nothing to see here, folks, move along.

.... by showing his bias and that fact that he's a champion of e-voting. So looking at his paper, with the foreknowledge of his bias, one can see where he twists things around to make it seem the machines had no bearing on how the votes were counted.

Why do some claim to be against the machines, yet link to papers containing machine supporting arguments, support articles supporting use of the machines and argue against arguments describing the machines as election altering?

It is a black or white issue.

Knowing that the machines are easily alterable, and knowing that there are powerful people who will use the machines to alter votes, it is unconceivable to me how anyone who claims to want machines taken out would want to support machine supporting characters like Stewart.

people who look at data, and people who don't.

You haven't laid a glove on Stewart's data. You are just trying to kill the messenger.

His reputation precedes him.

Like we are all aware of, two people looking at the same data can come up with two different conclusions. That is exactly why conclusions from one person can't be trusted until it is tested by others and/or peer-reviewed.

As far as looking at data, the only data I'd like to see is that which you and Febble got your hands on to come up with the infamous rbr conclusion.

Now, I know neither of you cough it up, but somebody can, and that somebody should. Hey, here's an idea, let Stewart look at it.

compared to my response below.

First, your concluding statement, which in this context I think is a false argument:


Febble, you just defended your Masters. How 'bout if I lobbied your professors publicly and privately to say that your "application" so to speak for a Master's degree was highly flawed, inaccurate and untruthful? But, I think your at your "core" you should get a Master's, just that this particular application of yours is incredibly flawed. Am I your friend??? BTW, in the course of my comments I point out that you last application for a master's, like you find Kennedy's Thimerosal arguments to be, was unfounded, without basis and damaging to legitimate interests of other people. My comments are of course in this hypothetical assumed to be just as honest and well-intentioned as yours or Manjoo's. How can you wiggle out of that one?

You can try to quibble a bit with the parallel nature of the analogy but in the end the level of critique here is inconsistent with the level of friendliness claimed.

My additional response, amplifying what I mean by "application" in the paragraph above, is a cut and paste from a response thread of mine in my own GD-P post on the R.F.K.-F.M. controversy:



So, dear Febble, you don't think I'm being harsh when I say you and Manjoo would strangle the baby in its crib, you instead DEFEND the timing and strategy of statements?

So, per the highlighted paragraphs above, what I can not abide with is Manjoo's (metaphorical) intervention to try to stop (in effect) RFK Jr's application for a warrant for a public investigation he calls for in Rolling Stone. Manjoo makes that investigation materially less likely, while being able to claim that he's called for an investigation too so he's in "good faith". I claim, since you're in cognitive psych, that Manjoo having multiple political personalities in tension with the other personalities' actions does not establish "good faith" on the aggregate level.

To clarify, my understanding of a proper commitment to free speech means no suppression or censorship or prior restraint, but that does not mean that anything said at any particular time is polite, proper, strategic, friendly, or non-damaging, or immune from a statement that "you should have kept yer mouth shut if you truly wanted an investigation."

Democracy. We should have a vote on whether Manjoo helps or hurts the investigation he claims to support.

Febble is waiting to defend her PhD. If you actually had strong reason to believe that, say, she had plagiarized her dissertation, then I think most academics would agree that she did not deserve the PhD. Friendliness really has nothing to do with it, or shouldn't.

An historian named Michael Bellesiles published research about guns in early America that was initially lauded for authoritatively undermining certain gun rights arguments. He ultimately resigned his post after a number of historians, including a good friend of mine, found persuasive evidence that some of his research was falsified. "Yes, but the gun nuts are so much worse" was not deemed an admissible argument. I don't liken anyone in the present debate to Bellesiles, but the principle remains: a poor argument in a good cause is scored as a poor argument. And a willfully poor argument can destroy a career.

You have no argument here, just florid and tendentious metaphors. Manjoo has written more good stuff on election integrity than any other journalist I can think of. What does it mean to say that he "would strangle the baby in its crib"? or that he has "multiple political personalities"?

Of course you can say to Manjoo that "you should have kept yer mouth shut," but if everyone who saw the big problems with RFK's arguments kept ther mouths shut, you would be complaining about that, too. If you want to come out and say "you should have spun" or "you should have lied," at least we would all know where we stood.

I'm saying, to make an analogy to a courtroom and the entitlement of BOTH sides to their "day in court" that Manjoo can say anything he wants LATER ON at "trial" so to speak. (our day in court is our time of national media exposure on all networks)

But instead of waiting for a day in court or pushing hard along with Kennedy for a big investigation, he's firing away now so we don't really ever get a day in court, and that's troubling.

RFK Jr and Rolling Stone can get us a hearing, which is only a chance, regardless of whether the article is all right or mostly wrong. The flaws, if any, in Rolling stone are invisible to the average fairminded swing voter, if you will. So, Choosing to interfere in this manner at this stage is premature especially SINCE HE SAYS HE SUPPORTS INVESTIGATIONS, and damaging to the cause of the neutral scientific dispassionate investigation you and FEbble claim devotion to.

Put another way, excess honesty is not moral perfection, it's passive aggressive. Can even be brutal. Excess honesty especially when combined with some emotion can get a person tombstoned. Ya know what I mean? It's not that hard.

If someone else that's a "neutral" so to speak (nobody is, totally) and says I wasn't clear, then I'll reconsider. But right now I'm puzzled at the lack of understanding here.

See, because if you don't comprehend or understand all of my points, your claims of "no evidence" may simply be a lack of reading comprehension on your part and/or a failure of communication on my part. But see, in both cases, the "no evidence" claim being made in this hypothetical would be wrong, albeit due to failures of communication.

I have no idea what particular "investigation" you think could not proceed before RFK's article, but now can. It seems quite late in the day to criticize a salon.com reporter for derailing the "investigation."

at least for me. I'm so afraid that the media attention on this issue won't go anywhere because people will kill it before it has a chance take root. This is the reason I posted that "Senator Kerry report for duty" thread. In two (the only two I'm aware of) interviews on network television RFK, Jr has been dismissed because the talking heads say that Kerry even said there was no fraud (killing the questioning of the process before it has a chance to take root). In my frustration immediately after the last interview I posted that thread, and them it became a pro/con Kerry thread that muddied the water. All I asked was for Kerry to state his view in the MSM to clarify it.(He doesn't seem to have a problem stating it privately.) It then became a "what Kerry has done/did" thread that didn't address the issue at all. For most of us in this forum, I'm willing to bet that this is our baby. We've spent countless hours here and don't want to see this die in the MSM before it has a chance to get to the general public for questioning.

Not brandy.

I think RFK's article would have had more traction if it hadn't kicked off with the exit polls. Lead with your best arguments, not your worst.

and let him get strong before we whack him. I've don't give a crap what it leads with.

that it leads with crap.

Potty assassination has been substituted for potty training.

:cry: This hurts me more than you, dear baby, but yer just poopin' too much. {sound of gunblast} {fade to black} :cry:

O8) O8) O8) O8) O8) O8) O8) O8)
Democracy is better off now, she's in heaven.
O8) O8) O8) O8) O8) O8) O8) O8)

Yeah, killin' your babies is a term for writers getting rid of pet paragraphs that don't really fit, but Manjoo instead of applying this principle to his writing applies it to the election protection movement pretty much as a whole. I've got a half dozen article ideas for things not about Ohio, exit polls or stolen elections, he can prove me wrong by doing a positive article on the real issues that GOES SOMEWHERE.

"The first and most important thing I want to stress: WE MUST NOT SET OUT TO PROVE THAT KERRY WON."

Well, if we could or can prove that, I myself would be all for it -- as long as it didn't derail us from all the other work. But I don't think this movement is so "infant" that it couldn't have learned something from those prescient words of almost 19 months ago.

The "nitpick" meme is malarkey. RFK loudly, needlessly embraced a passel of arguments that have been rebutted repeatedly and extensively for a long time now. If he felt this is what he had to write in order to get buzz, well, I am happy to be part of the buzz. Maybe it will actually do some good.

I wrote the foreword for his ebook. Check out his new site, it's great and available along with the transcript of his recent Thom Hartmann appearance, at <http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x431050>

But really, Guvwurld would definitely state his mantra here:



To which I would add the word "rational" before "basis", because one can always choose to believe them on faith, and therefore have confidence - of a sort.



But since you cite guvwurld with approval, and seem to adopt his "prescient" work, then it would seem likely you agree that there's no basis for confidence in elections because they are nonrepeatable, nonobserved, nonverifiable events (under the same or similar conditions as existed at the time of elections, a condition of scientific validity).

But it's important to note the flip side of the absence of evidence:



The problem with "no evidence" arguments is that it proves too much, and goes to far. It certainly covers the elections under electronics, regardless of whether we believe it covers RFK Jr. There's no evidence we have real elections. Anything could be happening, or not happening. We just don't know, if we're COMMITTED TO HARD EVIDENCE.

Why doesn't Manjoo say something like that?

Given how wide open the whole vote counting thing is in this, the most public of all possible fora (namely that of vote counting), how can one accuse concerned citizens making educated guesses about elections like RFK, Jr to fill in the void of evidence, WITHOUT ALSO AT ALL TIMES POINTING OUT THAT IN MANY JURISDICTIONS THERE'S NO EVIDENCE OF A FAITHFUL COUNT, JUST REPORTED TOTALS NAKED OF ALL SUPPORT???

Vendors and elections officials don't want to be held to the elementary and junior high school level standard of "show your math."

But Manjoo thinks the important thing is to blast RFK for making "baseless" claims. This is about as responsible as the elementary school student who, in a world of electronic calculators available to cheat, refuses to show his work and then chides the teacher who suspects foul play for a "lack of evidence."

Not cool, Manjoo, not cool at all.

But wanting to help, I could ghostwrite an approach for Manjoo, cuz i'm kinda a "big picture" type of guy some people say, and Manjoo is more a master of detail it seems, so here it is:

Critics could start out a more constructive critique like this:




But of course, instead of something like the above, Manjoo does find copious fault with the "application" for the investigation "filed" by RFK in Rolling Stone, and he does not suggest friendly amendments TOGETHER WITH THE EVIDENCE HE CLAIMS TO CHERISH and find so indispensable that it must be added in addition to over 200 footnotes, else Rolling stone must be dismissed as unproven and unfortunate.

I do have a problem with that lack of perspective by Manjoo. Yes, I do.

I am happy to say that the facts are otherwise.

Farhad Manjoo, 6/23/05, "Bungling the vote." http://dir.salon.com/story/news/feature/2005/06/23/ohio_vote/index.html

Farhad Manjoo, 11/18/04, "Third World democracy." http://dir.salon.com/story/tech/feature/2004/11/18/voting_problems/index.html

Farhad Manjoo, 10/29/04, "Fun and games in Florida." http://dir.salon.com/story/news/feature/2004/10/29/florida/index.html

Farhad Manjoo, 10/15/04, "Seeing red in Florida." http://dir.salon.com/story/news/feature/2004/10/15/florida_voters/index.html

Farhad Manjoo, 8/24/04, "The downloading of the president '04." http://dir.salon.com/story/tech/feature/2004/08/24/machines/index.html

Manjoo actually has standing to say that RFK is making the wrong arguments, because he has worked harder than any other writer I know of to sort out which is which. Trashing Manjoo is shooting the messenger, IMO.

of modest contribution by Manjoo. These articles make the situation worse though. It's as if Manjoo is saying "i've been here, done that, been in the movement, and i'm telling you it is ALL MESSED UP, even RFK jr's got it wrong"

SO, this is why i'm questioning his intent because it certainly is damaging and he's killing chances for national debate. I don't think that can be characterized as being an unavoidable tragedy brought about by the publication in Rolling Stone followed by an string of consequences including Salon that were practically predetermined and unavoidable.

No, clearly Manjoo chooses to assassinate the claims of anybody writing on stolen elections.

I'd debate him in a heartbeat. But he'll decline and tell me i'm nobody and nobody would attend. Alas.

I wonder if salon will publish-they appear to have their minds set.

Exactly. This issue is destined to remain "forever on the backburner" until someone with Kennedy's PR clout takes the time and trouble to write a decently researched article that makes the case properly.

The case is too urgent to sabotage with crappy evidence. He had a platform. Why couldn't he be bothered to make a decent case?

limitations of time would seem to explain it. Learning curves.

until he'd done his homework. Sorry, don't buy that excuse, especially as he claims he has considered the evidence "carefully".

Did he have a deadline or something? Like Manjoo?

1. What from the Kennedy article did you think was decent?

2. What should he have emphasized, in your opinion?

3. What about Lou Harris' concerns about the poll results as quoted in the article? I will admit that I still don't understand how 100's of polls across the nation could be so wrong when the sampling method was random (every 20th voter) and when the 'plan' to make the poll results right - by mixing them with election results coming in from the precincts - seems like preparation for committing fraud. By 100's of polls across the nation I mean that Mitofsky-Edison had multiple polls in each state (many polls in each of the swing states) and there were Zogby polls and internal Dem polls that all showed Kerry in a landslide.

4. What do you think about Bill Chirolas' response to Manjoo?



Oops - a fifth question: Have you looked at the recent work of Richard Hayes Phillips at Ohio Free Press? He has been counting ballots and finding evidence that suggests that ballot were shifted between precincts so that votes marked for Kerry in one precinct were counted for Bush in another precinct. <http://www.freepress.org/columns/display/3/2006/1355>

but in the mean time, here is a piece that I largely agree with:

http://moritzlaw.osu.edu/blogs/tokaji/

Cheers

Lizzie

Smart guy overall but doesn't seem to understand that new technology = new paradigm. New paradigm means different ways to cheat and malfunction. Not understanding that, he leads the ACLU through pro-DRE lawsuits that find paper technologies unconstitutional (punchard and central count opscan) while upholding DREs and precinct based opscans (for now). The test used is total "residual votes" which is overvotes plus undervotes. Overvotes are prohibited by DRE interfaces, and undervotes both discouraged (through reminders to vote unvoted races) as well as supppressed (when folks interpret those reminders are necessary to "cast vote")

It's not apples and organges, it's apples and rocks. But on the surface you can kinda see how a court gets there with a "gee whiz, DREs "lose" fewer votes than others". The DRE residual vote rate is just assumed to be the "real" undervote rate, those people who legitimately don't want to vote a particular race.

I don't see how it makes any sense to respond to Tokaji's critique of RFK by complaining that Tokaji isn't worried enough about DRE security, even if it's true. His arguments don't seem to hinge on his beliefs about DREs.

This is the problem with arbitrarily sorting folks into With Us and Against Us. Tokaji has done excellent work on voting rights. If he ends up Against Us, then we are drawing the circle pretty damn small. (For that matter, if Manjoo ends up Against Us, that should be an air raid siren that it's time to change the categories. But you and I have been through that before, I think.)

... is that he sucks at the teat of e-voting vendors.

The one thing we here all seem to agree on is that the e-voting machines are bad news in elections, yet Tokaji seems to think they are just dandy.

Another thing about all this is that we have made no enemies, they have made themselves, and Tokaji seems to be working hard at being just that: an enemy of free, fair, and open elections.

I would be very careful about aligning one's self with that sort of election expert.

"The DRE residual vote rate is just assumed to be the "real" undervote rate, those people who legitimately don't want to vote a particular race."

Who is doing the assuming in this sentence?

Different voting technologies will have different problems. We know fairly well what the problems are with older technologies, and they are serious (they cost Gore the presidency). Just because DREs have different problems is no reason at all to assume that their residual vote is any more "real" than any other residual vote rate. Presumably it would be possible for DREs to understate the "real" undervote rate, to the advantage of a particular presidential candidate.

Oh dear, Land Shark, we seem to be having a very peculiar conversation here. You seem to assess the value of an opinion by very strange criteria, to my way of thinking. I'm a simple soul - if someone says something I agree with, or tells me something I didn't know, I am content, even if it is something I wish I didn't think, or wished wasn't true. I have told you that the exit poll evidence suggests that far from being associated with greater exit poll discrepancies, digital voting technologies were associated with smaller discrepancies than older technologies, in urban areas. This is information. But you seem to regard it as dangerous because it might encourage people to think DREs are an improvement over the old technologies. And you seem to think that Tokaji's views on Kennedy are suspect because he takes an erroneous view on DREs. And that Manjoo is damaging the cause because he isn't cutting Kennedy enough slack.

Well, my criterion is not whether a view is politic but whether it is true. But I am Machiavellian enough to be aware that if we insist on what is true, we stand a better chance of being politic.

The only flaw is the wimpy labeling of the third category; Hard to Swallow.

Ohio as the centerpiece should have been powerful, with residual impact if RFK had wounded Blackwell, who is running for governor this year. I was thrilled when the previews of RFK's piece highlighted the word Ohio. How could you screw that up?

When every news report was citing exit poll crap again, I needed the rehab. I post on several non-political sites with secondary political forums, and partisan balance among the posters. Without exception, those forums have dismissed the RFK article as more babble about exit polls and Diebold.

I posted a brief summary similar to Tokaji's Real Problems on one of those sites but it didn't earn a single reply, instead buried between dozens of wisecracks about Democrats and machine paranoia. One guy made a post eerily similar to something I mentioned on this forum the other day, that Pennsylvania was close to Ohio in terms of vote count and margin: "I claim Bush won Pennsylvania, Both times. You know why? Because I want him to."

When I webmastered VerifiedVoting, he was always on the wrong side of this issue. I just confirmed with VotersUnite.org that this remains the case.

Anything that Tokaji says should be considered in that light.

Lou Harris doesn't seem to have said anything about polls in the piece, and I have no way of knowing what his opinions are. He may have stated them somewhere else, but I haven't found it.

I am happy to put my professional reputation on the line -- not that that counts for much on DU, but whatever -- that weighting to official returns is not preparation to commit fraud. It is standard practice and is likely to yield better estimates.

(On random sampling -- look, we know that there was non-response bias in the poll by age; that, the interviewers were able to measure so the analysts could attempt to compensate for it. I don't really understand how folks get hung up on the assumption that there couldn't possibly be non-response bias by vote choice, which of course the interviewers couldn't measure.)

Chirolas's numerical example seems to have little to do with what the DNC report actually says about Ohio. The provisional ballot survey (which was limited to Cuyahoga County) indicates that new registrants were about 3% of the electorate, and that about 1/4 of them cast provisional ballots. Overall, about 3/4 of all provisional ballots were counted. I can't find anything in the report that points to a massive franchise die-off of new Kerry voters. Maybe RFK can, but his "astounding fact" doesn't get there.

If that RHP work is what I think it is, I think RHP is still assuming that people Just Don't Vote for Bush and Connally, or for Bush and against Issue 1. That won't work. (EDIT TO ADD) I do like the idea of inspecting actual ballots; I would be especially interested in evidence of 'death by stiletto,' lots of overvotes. I think the spoiled ballots he finds are pretty equivocal and few -- and the aggregate returns from these counties don't seem especially anomalous -- but the more we know, the easier it is to evaluate what we are looking at.)

1. What I liked.

I will refer directly to Dan Tokaji's article, because I liked the same things. He was more measured than I was, because, I confess, I was put off by the fact that the first numbered item in the article was the section on exit polls, of which more later. But the fact that there was much to like was partly what made me so cross. I just think the article could have been so much more powerful had it not started by citing the most easily refutable part of the case.

2. What should he have emphasised.

I think he should have emphasised the stuff that I liked (duh...) But more importantly I think he should have emphasized that that that stuff about systemic disenfranchisement, of Blackwell's conflicts of interest, and all that Went Wrong In Ohio, is an outrage whether or not it made an iota of difference to the result If he wanted to quantify it, I think he should have been more circumspect - provided a range of possibilities perhaps, and I think he should certainly have checked his numbers, but mostly I think he should have emphasized that the point is not whether it affected the result but that it happened at all. Even if every disenfranchised voter was a Bush voter, it matters. It matters because democracy matters. Sheesh, I could almost rewrite the thing myself, if only my name was Kennedy.

3. Lou Harris and polls.

I can't comment directly as the cite isn't referenced, but I would like to know what he is referring to as "sticking out like a sore thumb", and what polls he is referring to. It may be that there are anomalies at county level, but this wouldn't show up in the exit polls, as fewer than one precinct per county on average is selected. So to address your more general comments about the exit poll: the sampling tries to be random (actually the longest selection interval was 10, the shortest was 1, i.e. every voter). However, even if strict random sampling was achieved, the sample would only be random if those who refused (or who were missed) were not "drawn from a different population" as we say in statistics from those who participated. If they were, the polls would show the well researched phenomenon of "non-response bias". Moreover, we know from the E-M report that redshift was greater where the selection interval was longer. The longer the interval, the easier it is to make an error, which won't matter if the errors are random. But the more likely a random error is, the easier it is for non-random selection to happen as well. If, for example, an interviewer is struggling for willing respondents, if the 10th voter looks surly, and the 11th looks friendly, and they are more or less abreast, then it is not difficult to see how you might end up with a sample more heavily weighted with friendly voters, so in that sense, it can induce another form of "non-response" bias, although "selection bias" is a better term.

As for the "mixing" with the vote count. There is simply no reason to infer preparation for fraud from this. If you look on the E-M website, and the FAQ, you will see exactly how the projections were always going to be made, which is the way they are made every time, and that page was up well before the election. It is simply not a secret. I think the misunderstanding here arises from confusion the meaning of "projections". All polls try to measure something that will tell them something else. By measuring how people think they will vote before an election, the pollsters, after a fair bit of data manipulation, will "project" the future election result. Therefore, as well as a margin of error due to sampling variance, there is also error due to the assumptions made by the pollsters about, for example, how likely each respondent is to vote. The exit polls are absolved from that assumption, but they still need to weight for other things likely to affect their "projections" including bias in the poll (non-respone bias, for example). This is done initially by observing the age, race and sex of those who were selected to be in the poll but who were not (they were missed because the interviewer was busy, or the voter refused), and weighting the sample for any mismatch in the age, race and sex proportions between the sample of voters interviewed, and the non-responders. However, while the pollsters might be aware that there were, for example, a smaller ratio of men to women in the respondent sample than in the refuser sample, and can therefore weight by the voting preferences of men in the sample, there is no way the pollster can no the voting preferences of the men not in the sample. It may be Bush-voting men are more likely to refuse than Kerry voting men. We know there is non-response bias by gender; we can't rule out non-response bias by vote. So, before the vote counts come in, they also weight according to whether the sample is in line with pre-election expectations. If it isn't, they suspect bias in their sample, and reweight accordingly. Then, once the vote returns come in, they also have further information as to bias in their sample, and they use this too.

The trouble, of course, is that this procedure assumes the vote returns are correct, although as the polls are designed to project the counted winner, not as a check on the count, this is only a problem if you want, as is of course reasonable, the poll to be an independent check on the count. But it isn't designed to be. The fact that it isn't designed to be, is not, in itself, cause for suspicion. It's been done this way for years, because the "projections" are designed to call the winner of the counted vote. Ironically this is probably why people think they are so uncannily accurate.

4. What do you think about Bill Chirolas' response to Manjoo?

I think he is less fair to Manjoo than Manjoo is to Kennedy. I've never bought into the "he who is not for us is against us" argument, and I think when your own side points out the inconsistencies and flaws in your argument, they are helping, not hindering the cause. But it may be a minority view around here ;)

5. RHP

I have a lot of respect for Phillips, and exchanged data and analyses on Ohio with him early in this whole saga. I don't think he's a statistician, but he is a diligent collector of data, and I think he's done an invaluable job. I think some of his inferential analyses fail to take into account some important factors, but I don't think inferential analysis is his strength. He prides himself on using nothing more than a handheld calculator, and I think that is an excellent approach - simply adding up the votes he thinks were lost. It's once we try to multiply them using inferential methods that we can up with wrong answers (although good inferential statistics are what we need).

Because I've always suspected that any Dem who steps forward about election fraud will have Clinton's team treating them immediately as delusional. And THAT is a huge problem for Dems, because the Clinton people get the most access to media.

I haven't had time to fully absorb Kennedy's piece, much less Manjoo's -- but I think this article might have a few points of interest to those who want to find and accentuate the best of Kennedy's article...

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=103x215073

1. OP on this thread is dead-on. Arguments about who will untimately be determined the winner of Ohio are moot in the face of overwhelming evidence that the will of citizens in our democracy was subverted. This is unacceptable in microscale, much less in macroscale as seen in recent elections.

2. While I do agree with some of the points made in the Salon article, Manjoo is guilty of "skeptic's disease". This is a condition where the skeptic thinks that providing a plausable alternative to their opponent's assertions is good enough to disprove the assertion. Ironically, the skeptic has committed the same sin as his opponent in a dance of rhetorical hypocrisy. Take, for example, the "swamp gas" explanation for UFO's. It is true that some sightings of UFOs can be attributed to swamp gas, but not all of them can be. What's more, if the event cannot be measured directly, both the believer and the skeptic are making conclusions with no data on the event in question. This is the case with the election...if one doesn't really have all of the information about the true intent of the vote, then one doesn't know where to start to verify or debunk the results. The whole debate becomes a war of experts with their own criteria and their own set of data (none of which reflect the TRUE intent of the voters).

3. Manjoo makes the erroneous assumption that citing a reference indicates 100% agreement with that reference. This is not true. Kennedy can be free to cite a fact from a report even though he disagrees with the conclusion. I am not confortable with Manjoo's use of this tactic to discredit Kennedy. I am also not comfortable with Kennedy not stating clearly that he disagreed that voting machine allocation discrepencies favored Republicans and Democrats evenly (from the electoral maps, it is clear that this trick favored Republicans..regardless of what the "Democrat's report" concluded).

4. I am a little disappointed that Kennedy's arguments were not as solid as he could have made them, but Manjoo's arguments are sparse when compared to the totality of data that Kennedy presents. There is a LOT of evidence in Rolling Stone beyond Manjoo's arguments. The purpose of the Salon article seems to be more of an attack Kennedy piece than a point by point refutation.

5. The governer's race of Ohio in 2002 argument that Manjoo uses to counter some of Kennedy's numbers is really, really weak. In 2002, Governor Taft-R was running against a Democratic challenger (Tom Hagan) that was not ready for prime time, not endorsed by his party, not given adequate funds to campaign, and made numberous gaffes before the election. Plus, 9-11 made Bush, Taft, and all Republicans at least 10-15% more popular than they were in 2004. Turnout was dismal for the election, and Taft predicatbly won. Using this election as a benchmark for a Presidential election where the country is polarized and the issues are huge is disingenuous....at least on the level of which Kennedy is accused.

6. Even if Manjoo is right on every point (which he has not proven sufficiently..only provided doubt), there are still thousands upon thousands of votes that were not counted...and thousands and thousands of people who wanted to vote but were turned away (add 'em up, and the numbers sure do seem to me to be in order of questioning the outcome). The fact that we cannot simply know these numbers for sure speaks volumes about how our democracy has been subverted. We shouldn't be left with speculation....and we should not be left with Blackwell running for governor and still in charge of Ohio's elections.

1. I agree with this point, but I don't agree that the OP is dead-on.

2. I think your analogy inadvertently makes the point: I want to talk about election integrity, not about UFOs. Many of the arguments in the RFK article are way too close to ufology.

3. Sure, Kennedy is free to cite a source he partly disagrees with -- and Manjoo is free to document a pattern of cherry-picking.

4. There certainly is a lot of evidence in RFK's article, but there isn't a lot of evidence that Kerry actually won or should have won Ohio. That was the standard that RFK set for himself, not something that that meanie Manjoo thrust upon him.

5. Gee, go slug it out with Walter Mebane. But before you do, you might actually want to read Mebane's work. I don't know what to try to explain first.

6. I agree with this, and I believe Manjoo does too. That is why we are both (I think this is a fair reading of Manjoo's intro) so pissed at RFK.

1. The OP is dead-on because it acknowledges that the REAL issue is not who won, but why we allow people to lose their voting rights for partisan reasons by setting the bar as high as "prove Kerry won".

2. You are not addressing my point. Let me state it more clearly...just because something is established as "possible" does NOT mean that it is probable. One cannot debunk an assertion by simply coming up with another possible explanation in the absence of proof. That is "skeptic's disease".

3. My original point is that the citation of a fact from a piece without agreeing with the conclusions of the piece is not considered "cherry-picking", so Manjoo's accusation of academic fraud is itself, fraudulent. The fact that there were long lines that made voters leave is a fact in the citation (there was evidence of that)...the assertion in the same piece that these voters were equaly Bush and Kerry voters is an opinion. There were no facts in the DNC to back up the "equal" claim. That might explain why RFK didn't address it.

http://www.dailykos.com/storyonly/2006/6/3/224759/7766

Here is another who agrees with this assessment.

4. The preponderance of evidence still tilts heavily in RFK's direction. Manjoo had four points, and three of them are fairly weak. RFK's article covered far more than four points, so Manjoo, even if right on every point, barely dented the totality of RFK's work.

5. Telling me to argue wth the guy who came up with the comparison is an appeal to authority, which is a logical fallacy. I really do not care about who came up with the idea to compare the Kerry vote of 2004 with the Hagan vote in 2002. The comparison is still disingenuous. Hagan 2002 was a gubernatorial election, it was an off-year, Hagan wasn't popular or well-known, Hagan wasn't funded, and the political climate in Ohio (and he rest of the country) was drastically different. As an Ohio voter, the comparison stood out to me like a sore thumb (as it would any Ohioan who was here for both campaigns).

6. A fair analysis of RFK's work should not be done by someone who is "pissed"...that goes for Manjoo as well as the average DUer.

1. My problem is that RFK framed the issue as "Was the 2004 Election Stolen?", and proceeded to offer some very iffy evidence in support, tending to obscure better evidence for other important claims. Some folks may think that posing the issue in that way is a neat way of getting people interested (and under some circumstances I might even take that view myself). But I think it is only fair to Manjoo to point out that he isn't the one who framed that issue. And I agree with you that it is the wrong issue. And I see no reason why the issue had to be framed in that way.

2. In all seriousness, what exactly is "something" in this context? Are we attempting to assess whether people were disenfranchised, whether (and under what counterfactuals) Kerry won Ohio, or what? At this point I cannot tell what, if anything, we actually disagree about.

3. I don't think that Manjoo alleged "academic fraud," nor would it make much sense to lodge that complaint against Rolling Stone. But I do agree with him that it was misleading of RFK to cite findings in the DNC report that supported his case while not so much as mentioning broader conclusions that directly opposed his.

The assertion that the voters who left lines were equally Bush and Kerry voters is not 'merely an opinion,' unless the analysts weirdly made it up altogether. It would be an inference from the survey data, just like all the other inferences. I don't see how it makes any sense to accept the inferences that RFK likes while disparaging the ones that he doesn't. That said, it should be noted (as, if memory serves, I noted here soon after the report came out) that since the total sample size was only about 1200, we don't have a lot of statistical power to estimate how people who left the lines would have voted. We can certainly say that the DNC analysis doesn't support the conclusion that these voters overwhelmingly favored Kerry.

4. Well, I thought the totality of RFK's work was intended to establish that Ohio was stolen, and I don't think he did that at all. It seems to me that once one chucks out the bad evidence, the good evidence adds up to fewer than 118,000 votes. If he hadn't set himself the challenge of reaching 118,000 votes, this would not have mattered.

5. Since you evidently have already made up your mind, I'm not sure it matters much what I appeal to. Nevertheless, I can point out that you seem to assume that someone argued that the Hagan race was like the presidential race. No one did. That is why I suggest that you read Mebane's report, if you haven't already, and then criticize arguments that he actually made.

6. Hey, whatever. If I'm not mistaken, quite a few people are venturing analyses (or at least criticisms) of Manjoo's work, and they generally seem "pissed."

from Manjoo's article

"The tendency to vote for Kerry in 2004 was the same as the tendency to vote for the Democratic candidate for governor in 2002," their report noted. "That the pattern of voting for Kerry is so similar to the pattern of voting for the Democratic candidate for governor in 2002 is, in the opinion of the team's political science experts, strong evidence against the claim that widespread fraud systematically misallocated votes from Kerry to Bush."

That is where the camparison is made between Kerry 2004 and Hagan 2002. It is not a fair comparison for all of the reasons stated above and it seemed to be the basis for their conclusions.

And, I haven't made my mind up (so please do not assume). I donot have an agenda as much as I have questions. I am saying that Manjoo is just as guilty of sloppiness as RFK is. Only, Manjoo launched his criticism more as an indictment of Kennedy than an honest analysis of the issues, which is hard to do from a glass house.

Also, read the Kos entry I cited, which went into detail regarding the DNC report, the "equally hurting Bush and Kerry" claim and the number of people who left the registration lines.

I will accept that you haven't yet made up your mind. I drew a faulty inference from your attribution of a "disingenuous" comparison, which struck (and strikes) me as quite unwarranted.

Everyone would agree with you that the gov race was very different from the pres race, but that doesn't make it useless as a statistical baseline. What tends to happen in comparing partisan races is that while the support levels may vary sharply, the results are still highly correlated -- and the DNC analysis confirms that this is true for these two races. If you want to argue that actually the races were too highly correlated, and that this in itself points to fraud, then you have some work to do. My guess is that if there is a good argument for fraud lurking in the data, it in no way hinges on the unrepresentativeness of the Hagan race.

I've read the Kos entry, and it gives me no reason to change my view. I'm not pounding the table here, just working with what I know so far.

I don't agree that Manjoo is just as guilty of sloppiness as RFK. It probably isn't worth debating this, because I don't know how to measure it.

http://www.dailykos.com/storyonly/2006/6/3/224759/7766


A Semi-Comprehensive Quizzing of Manjoo's Rebuttal of RFK Jr.
by malcolm
Sat Jun 03, 2006 at 07:47:59 PM PDT
This will be my first ever diary on Daily Kos, but I don't expect anyone to go easy on me.
This afternoon I read with great interest Farhad Manjoo's rebuttal in Salon of Robert Kennedy Jr.'s article in Rolling Stone. I would ask you to believe me when I state that although I have been of the opinion since Nov 2 2004 that it was very likely that the election was fraudulent, I approach articles such as Mr. Manjoo's with a strange hope that this time I can have it explained to me in a way that allows me to finally get a good night's sleep, secure in the knowledge that nothing overtly untoward was going on, that the process of democracy in the United States is still hobbling along, flawed but basically intact. Despite my huge admiration for RFK Jr., and my lasting dream that he will one day soon be the most effective and honorable head of the EPA that this nation has ever had, deep inside I harbored a hope that he was wrong on this, and that Salon and Mr. Manjoo were going to show me the way.
Sorry to say, I'm just not there yet.
What I'm going to do is outline 3 out of 4 of Manjoo's rebuttals, and explain in each case why I just don't think they wash. I wish to do this in a spirit of openness, avoiding any hostility or undue sarcasm, and I would hope that people would afford me similar. I'm happy to be proven wrong on any or all of this. I admit straight up that I am in no way an expert in the fields of elections, statistics or politics and don't wish to pose as one. Statistics, in particular, I did three years of, but that was twenty years ago.

<snip> more at link

I agree with the kossack that the Kerry-Connally discrepancy is problematic because of the fact that Connally was not identified as a Dem. I want to see this investigated, however, because it reeks that Connally got more votes than Kerry even though she was down-ticket.

Kind of a depressing Sunday - after feeling happy on Saturday that Kennedy was at least out there slugging...

:(

they always have some merit on each side.

Scoring *apparent* points that could later be reversed, where a greater investigation is being called for, is not grounds for depression or sadness. Have a great Sunday!!

Voters were disenfranchised in Ohio. It is difficult to take people seriously who deny this.

They seem to wheel this guy out anytime they need to take down someone getting too much attention...

Daily Howler has the scoop on Farhad Manjoo's cheapshots trying to derail Krugman:

http://www.dailyhowler.com/dh060405.html

Thanks for the link.

the other articles of Manjoo can have relevance. Manjoo's developing an impressive (assumed) progressive "kill list": Krugman, RFK, Jr. Mark Crispin Miller, Steve Freeman.

It goes beyond elections and it's impressive

Dear DU Friends,

I just sent this to Manjoo to post on Salon.

In his June 2, 2006 Salon article “Was the 2004 Election Stolen? No”, Farhad Manjoo claims to have “thoroughly debunked” a June 15, 2006 article in Rolling Stone by Robert F. Kennedy Jr.: “Did Bush Steal the 2004 Election”.

As one of the (applied) statisticians cited Kennedy’s article, I find that Manjoo’s “debunking” is either superficial spin that is easily refuted by adding more detail to Kennedy’s already very long piece, or simply factually erroneous. I will focus on the Manjoo points that relate to official return and exit poll data - my particular area of expertise.

a) Making the point that voters sometimes do vote more often for a “down ticket” candidate is hardly a “rebuttal” of the 12 County Ellen Connally “anomaly”. First of all this was not a normal election but rather one of the hardest fought Presidential races in recent history in a key battleground state. Moreover, Connally was an under-funded and largely unknown African American, liberal, pro-choice, pro-gay rights, anti-death penalty, woman from Cleveland, and all of these counties are far from Cleveland. But even if we disregard this, anybody who wishes to check the numbers posted on the Secretary of State of Ohio’s official website will find that: a) nine out of only 14 counties (out of a total of 88 counties in Ohio) where Bush did abnormally well (over 43% better) relative to Moyer (the Republican Supreme Court nominee) just happen to be the same 12 counties where Kerry inexplicably did worse than Connally, b) this was not a result of voters in these counties switching from Moyer to Connally as Connally’s vote is worse than the Ohio county average in all but one of these counties and Moyer’s vote is better than the county average in all but one of the 12, c) the “lost Kerry” vote of 81,599 (relative to the average Kerry/Connally ratio) just happens to be just about equal to the “excess Bush” vote of 75,766 (relative to the average Bush/Moyer average in the 12 counties), and d) when one looks at the county by county pattern of “lost” Kerry votes to “excess” Bush votes, they match almost perfectly in their orders of magnitude – see the table below:

"Excess Bush" and "Lost Kerry" votes for Counties where Connally vote is less than Kerry's.
2,712 4,070 Auglaize
2,264 2,771 Brown
26,913 25,123 Butler
10,077 13,855 Clermont
2,829 4,078 Darke
1,751 2,130 Highland
3,579 4,027 Mercer
3,710 5,882 Miami
2,974 2,013 Putnam
4,069 4,096 Shelby
1,376 1,968 Van Wert
13,512 11,586 Warren
75,766 81,599 Total


As Kennedy notes such an approximately 80,000 vote shift from Kerry to Bush would have given Bush an additional 160,000 votes, more than enough to steal an election which Bush won by about 119,000 votes. In other words, just this one “anomaly” is enough to support a “stolen election” hypothesis and I challenge anyone to come up with a plausible explanation of this set a)-d) of remarkable statistical “coincidences” as reflecting anything other than a pattern of vote shifting!

b) Manjoo’s claims that exit polls are not always accurate and that Mitfosky has an “explanation” for the large discrepancies (that he acknowledges were both highly significant statistically and the largest since 1988 in the January report) are both irrelevant when one considers the pattern of the exit poll discrepancy in Ohio – see graph below:

(Can't be reproduced on DU - see Table on p. 6 of ElectionArchive.org/ucvAnalysis/US/exit-polls/Ohio2004-US-future.pdf )

These are “within-precinct discrepancies” or ((Kerry -Bush Official Vote) minus (Kerry – Bush Exit Poll shares)) for each of the 49 precincts for which exit polls were taken in Ohio based on data released by the Election Science Institute in a report in which Mitofsky is listed as an “assisting author”. (Note that there are not 49 bars as some precincts have the same official Kerry vote share so that the bar is an average WPD.) Negative WPD reflects large Kerry exit poll “overstatements” relative to official vote counts. Note that almost all of the discrepancies are negative (against Kerry). In fact, as Kennedy reports, 20 of the 22 statistically significant discrepancies (large enough so that they could not plausibly be the result of random sampling error ) are against Kerry. But even more striking is the pattern. On the right side in “high Kerry” precincts (above 57% official vote count – right of the red vertical line) there is a more or less random pattern of Bush (positive) and Kerry (negative) discrepancies, whereas on the left side (below 57% official Kerry vote – left of red line) almost all of the discrepancies (and all of the statistically significant discrepancies) go against Kerry. Moreover the precincts on the left are densely clustered whereas they are more dispersed on the right.
This strikingly non-uniform pattern cannot be explained by either large but “unbiased” (not one-sided) exit poll discrepancies – the pattern on the right, or by a “reluctant Bush voter response to the exit Poll” (“rBr” - Mitofsky’s hypothetical explanation) as this would not even produce the pattern on the left, not to mention the un-biased discrepancy on the right. Those who have been following the 2004 exit poll debate know that rBr would produce a “U” shaped pattern of discrepancies that are larger in more competitive precincts and taper off to nearly zero in highly partisan precincts. There is no such pattern on the left of the graph. But the pattern displayed is perfectly consistent with “vote shifting” from Kerry to Bush that would “move” precincts by their official vote count to the left thus producing large negative Kerry WPD and much smaller Bush WPD, and a clustering of precincts on the left. In fact Manjoo’s (or Lindeman’s) calculations of overall exit poll response averages are simply mathematically wrong as they don’t take into account the larger proportion of Kerry or Bush voters in the “high Kerry” or “high Bush” precincts respectively – for accurate exit poll response rate estimates, and an explanation of why Mitofsky’s data contradicts his own “rBr” hypothetical see NEDA reports.

c) Manjoo’s claim that Mitofsky’s “rBr” hypothetical has not been decisively disproved is bunk. We applied an “optimal” (in the sense that this explained the WPD in more precincts than any other level of “rBr”) 59% to 50% excess Kerry to Bush voter completion rate to the Ohio data and found that 30% of Ohio’s exit polled precincts still had significant discrepancy, that these were still overwhelmingly against Kerry (11 out of 15) resulting in a 4.3% WPD against Kerry that was still more than double Bush’s 2.1% margin of victory in Ohio – see NEDA January report.

d) Finally, Manjoo’s claim that WPD’s in Ohio, Iowa, Nevada, and New Mexico were “not outside the margin of error” depends on how one calculates the “margin of error”. In particular, based upon a very liberal 30% “cluster factor” increase in the margin of error proposed (after the election) by Mitfosky, there was only about a 1 in 100 chance for the approximately 6.8% (weighted statewide) WPD in Ohio. This is well over any reasonable margin of error (typically set at about 1 in 20). Moreover, even using this very liberal margin of error, the odds of the Kerry exit poll discrepancies being as large as they were in the three key battle ground states of Ohio, Florida, and Pennsylvania simultaneously (and not just any random three states) is an impossible 1 in 133,000 (see Baiman Ohio Court Challenge affidavit).

In short Manjoo’s “debunking” of the official vote and exit poll analysis parts of he Kennedy article is superficial and erroneous nonsense. Moreover, we have been through this with Manjoo and many of the people that he cites before. He and they either do not understand the relevant statistics and mathematics of rigorous exit poll analysis and should not be reporting or prognosticating on this topic, or they simply refuse to accept their errors of interpretation and understanding.

my own debunking of the paper by you that Manjoo cites is here:

http://www.geocities.com/lizzielid/TheGunIsSmoking_Review.pdf

Enjoy.

Cheers

Lizzie

a) Ron has made no rigorous effort whatsoever to substantiate his case. We can all agree that 2004 was "not a normal election," but then what basis does he have for arguing about the expected number of counties in which Connally should have outdrawn Kerry (which he claims is zero)? There isn't actually a statistical argument here to rebut, when one comes down to it -- only Ron's seat-of-the-pants expectations.

I do think it is likely that most of the so-called "Lost Kerry" votes in these counties were cast by people who voted for Bush and for Connally, and so I am not surprised that they tend to correspond with so-called "Excess Bush" votes. If we accept that Connally was "under-funded and largely unknown," what is the point of enumerating her other characteristics? Not even Connally's partisan affiliation appeared on the ballot. Why is it inconceivable that Bush voters pulled the lever for a judicial candidate named Connally? Well, it isn't. And probably quite a few Kerry voters pulled the lever for her Republican opponent. Thus, we would expect Connally to do best vis-a-vis Kerry in Republican counties, which have more Bush voters than Kerry voters. Any analysis that doesn't even attempt to consider these points is inadequate on its face.

b) Ron is welcome to write lots of words explaining why he thinks there is a "pattern" in the Ohio exit poll discrepancy. However, ordinarily when statisticians want to argue for "patterns," they state hypotheses and report the results of significance tests. So, again, I see no statistical argument here to rebut.

However, we can formulate one. Ron seems to assert that the precincts with less than 57% official Kerry vote share are disproportionately likely to have red shift. (Why 57%? Well, it's cherry-picked so that a couple of precincts with big red shifts fall just below the line.) Interested readers can verify (from the same ESI data used by Ron) that below 57%, 30 precincts have red shift, 6 have blue shift, and 2 have zero shift. Above 57%, 5 have red shift, 3 have blue shift, and 1 has zero shift. Thus, we can perform a chi-square test on either 30 : 6 vs. 5 : 3 (red shift : blue shift) or 30 : 8 vs. 5 : 4 (red shift : no red shift), obtaining p values greater than 0.10 in either case. And this is using Ron's cherry-picked dividing line. If we divide the precincts into two equal halves, or into "majority Kerry" versus "majority Bush," this "pattern" is even less consequential.

I could keep going, but I imagine that I have two kinds of readers: folks who already see what is wrong with Ron's arguments, and people who will agree with Ron no matter what any of us actually writes. If Ron's failure to conduct a basic statistical test -- and to note that it undermines his argument -- doesn't raise eyebrows, I don't know what would. That said, I encourage interested readers to consider Febble's review (to which she has linked) and feel free to ask questions.

c) As one would think Ron would have understood by now, his attempted refutation is irrelevant, because it assumes that non-response bias should be constant in all precincts. The Edison/Mitofsky evaluation report, released in January 2005(!!), presented evidence that non-response bias varied across precincts, depending (for instance) on the interviewer's distance from the polling place. Ron is refuting a straw man.

d) Ron's apparent quotation of Manjoo, "not outside the margin of error," does not appear in Manjoo's article, thus providing another instance of refutation by misquotation. Manjoo does not claim that the WPDs were within the margin of error. What Manjoo states is that "the exits didn't show Kerry ahead by the margin of error, meaning, statistically, that his lead wasn't secure."

The margins of error in exit poll projections are not calculated by applying a cluster effect multiplier to a formula based on raw sample size. They are based on observed deviation in the precinct samples. The standard errors (for two different estimators) are reported in the E/M evaluation report. For the Best Geo estimator based on interviews alone, the exit poll gives Kerry a margin of 6.5 with a standard error of 3.9, p > 0.09. The estimate incorporating pre-election expectations is even less favorable to Kerry. Manjoo is correct.

Manjoo's reaction to the "660,000 to 1" statistic used in RFK's article also applies to Ron's "impossible 1 in 133,000." If someone were arguing that the exit polls were accurate within sampling error, then this statistic could be used to demonstrate otherwise. But no one is arguing that. Manjoo did not state that the Ohio exit poll was accurate within sampling error. Once again, Ron is rebutting a straw man.

If these are Ron's best arguments, then I think we are done.

on the expert level without a trial, so it's just claims in shorthand of an expert community. That certainly includes claims that another professional is not "rigorous", etc.

Maybe febble and otoh should write a 300 page book, that could contain enough explanation to possibly educate and then we could decide.

As I point out in the fitrakis thread in ER, the plaintiff's election contest case involves the totality of the facts being the winning formula, like a prosecutor. The defense, like Febble and OTOH always try to pierce individual facts. It's almost always possible to do that.

But don't let individual fact piercing, even if successful, take your eyes off the "totality of the circumstances" price. IN that sense, given that there's not been a single place like a trial to sit down and hear evidence, the "plaintiffs" here haven't really had a fair chance to FULLY make the case. The evidence can not be held in mind all at once without such a sustained effort, but that's the plaintiff's case here/


on edit: let me be perfectly clear: the defense almost always contests each and every fact with a plausible explanation. that does NOT mean there isn't a ROCK SOLID case for the plaintiff in the end, at summation.

We don't need a 300-page book to point out, again, that the pre-election polls said that Kerry would win New York by 15 to 18 points; the official returns say he won by about 18 points; and the exit poll interviews "say" he won by 30 or 31. I've had people tell me they think it is perfectly possible that Kerry did win New York by 30 points, but no one yet has told me straight-faced that s/he believes it.

That's just one simple argument against the accuracy of the exit polls, no matter what one thinks about 2004 election fraud in general. Even if you somehow can manage to believe that the exit polls are accurate, at least in the states where you really want them to be accurate, it will be hard to persuade others to share the belief.

The "totality of the circumstances" indicates that Bush reneged on his Air National Guard obligations, AFAICT -- but we still got hosed when CBS featured a memo that gave the appearance of being forged.

It is obvious that the statistical power of NY exits is minute because the numbers of voters polled were so small. Something like 700 people were polled in NY, while 2000 were polled in NC - a state almost one third the size in population.

NY is an outlier which by most standards of analysis would be cast aside, and never again even mentioned by an honest statistician.

You appear to be out by a factor of nearly 4.

The sample size in New York was 1476*. You show no evidence of knowing what "statistical power" is, nor warrant for anything else you said.

If you really want to chuck out New York as an "outlier," then presumably it would be because there was error in the poll. Hmm....

*EDIT TO ADD: That's the number of interviews for which all the responses were phoned in. The total size is probably substantially higher, but I don't have a list of total Ns handy.

The sample size in North Carolina was 3 times larger, per capita, than the sample of NY.

Anybody can see that if you interview 30 people out of a hundred, you stand a chance of getting a better matching result than if you interview just 10 out of a hundred, as was done in NY. Not really done in NY, or NC, but that's how the numbers work out - three times as many people per capita (voters) were interviewed in NC as compared to NY. Simple math, really.

So, using that skewed measurement (NY) is like using a really bad statistical outlier, which messes everything up. Heck, even I can see that.

but if you don't believe us, then check TIA's posts, because this concernes the Law of Large Numbers. Or even Wikipedia here:

http://en.wikipedia.org/wiki/Margin_of_error

Where you have a large population, your chance of getting a matching result depends entirely on the size of the sample not the size of the population.

The sample size in North Carolina was almost exactly the same as the sample size in New York, and the product of Kerry's vote share times Bush's vote share was similar in the two states (.25 for NC, .24 for NY). From the sample size and the product of the two voteshares you can compute the standard error, and from that, the MoE.

You have the wrong sample size and the wrong calculation. This means you are wrong. NY is not a statistical outlier in terms of the statistical power of the poll. The statistical power in the two states was virtually identical.

Peace

Lizzie

There were 2099 sampled in NC with a voting population of 5.5M.

The size in NY was 1476 with a voter population of 11.8m.

Apples to oranges. Using NY is a bad idea.

numbers. OTOH got 1476 from the Roper subsample - is that where you are getting your NC number?

Well, the point about the subsample is that it is about half the total tallies (but not exactly half).

What I am telling you is that the total tally count for the two states was virtually identical. I probably shouldn't be telling you, but I am. And for the last time, the population of the state is completely irrelevant to the statistical power - power doesn't depend on proportion unless the proportion is quite large, way larger than these proportions.

There is absolutely no reason to suppose that NY was a statistical outlier. The statistical power was huge. The only conceivable explanations for the large NY discrepancy are fraud or polling bias. Fraud is pretty hard to argue in NY, given the expectations and the technology, so it looks like polling bias.

Are you seriously trying to say that the discrepancy in NY was due to sampling error because the sample was too small?

The sample size of NY, compared to the sample size of NC, was three times smaller when related to the population of voters.

I can't believe it takes me to show you that.

Please read the Wikipedia link, BeFree, pretty please?

In short Manjoo’s “debunking” of the official vote and exit poll analysis parts of he Kennedy article is superficial and erroneous nonsense. Moreover, we have been through this with Manjoo and many of the people that he cites before. He and they either do not understand the relevant statistics and mathematics of rigorous exit poll analysis and should not be reporting or prognosticating on this topic, or they simply refuse to accept their errors of interpretation and understanding.

http://www.baiman.blogspot.com/

http://www.geocities.com/lizzielid/TheGunIsSmoking_Review.pdf

Baiman does not appear to "understand the relevant statistics and mathematics of rigorous exit poll analysis" either.

I would not dispute Ron's right to prognosticate, but I agree with him that those who do should not "simply refuse to accept their errors of interpretation and understanding" especially when these ahve been repeatedly pointed out.

As I have tried to explain before, and as you could probably verify from any basic statistical textbook (or would someone like to try a Google?), the statistical power of a sample depends more on the sample size than on the sample percentage, unless the percentage is rather large. The exit poll samples are all tiny fractions of 1%.

This is not something I invented to torture you. It's just the way it is. I really don't like it when you say that "(a)nybody can see" statistical principles that actually aren't true.

If you think about it, you might realize that something is wrong with your intuition. National phone surveys often interview something like 1000 or 1500 people, right? OK. Now, let's say that Ohio contributed 5% of the national electorate (I think it was a bit less). Should a poll of Ohio only interview 50 to 75 people in order to be equally accurate? New Hampshire is less than 1%; how about a poll of 10 to 15 people?

A sample of well over 1000 people (probably well over 2000) has a lot of statistical power no matter how large the population is. The One Small Catch is that to be accurate, it should be a random sample. The less random it is, the less likely it is to yield accurate results, even if it is huge (like the 1936 Literary Digest mail-in poll).

(EDIT to correct confusing phrase)

..."sample size". The size of the sample of NY is three times smaller than the sample size of NC when you factor in the number of people the sample size is supposed to represent.

And since neither sample was random, the factor increases.

To be an equal sample, the numbers of the voters in NY sholud have been three times greater to be equal to the sample size of NC. It is just simple math. Any comparison otherwise is apples to oranges. NY is a really bad outlier.

would you check that wikipedia link? Or even TIA?

YOU DO NOT "FACTOR IN" THE NUMBER OF PEOPLE THE SAMPLE SIZE IS SUPPOSED TO REPRESENT when you compute the margin of error in a poll.

You don't. Really, you don't.

The formula for the standard error is sqrt{p*(1-p)/N}, and to get 95% confidence limits interval you times that by 1.96. p is the vote share for one candidate (take your pick) and N is the sample size.

As you can see, the population size doesn't appear in the formula. If the population is really quite small (say you want the standard error for a precinct, with a population of 500, and a sample size of 100) then you do use a slightly different formula, but when your sample is a small proportion of the total population, it is not worth using that formula, because the answer will be the same as the answer you get using the one I have just given you.

As for the non-random nature of the sample - I agree, both samples were probably non-random, i.e the were biased. However, you cannot compute this. You infer it from your lack of match. But an alternative theory is that the sample was random but the vote count was biased.

I thought that was the theory you liked?

Sorry, wrong answer, as explained in the preceding post. You might want to reread the post and try again.

Ah, heck, let me try one more time. You now argue (I think) that a survey of, let's say conservatively, 1500 New Yorkers out of a population of about 19 million should be ignored because it is too small a proportion of the population. You don't say what sample size would be adequate, but can we assume that it would be no smaller than 1900 or so? So, 0.01% seems to be a conservative estimate of the Smallest Acceptable Sample in your view.

The 2004 US population was estimated at a bit under 294 million, of which 0.01% would be 29,400. So, based on your standard, the Smallest Acceptable Sample for a national survey would be about thirty thousand people. Does that sound about right to you?

Does anyone have any questions?

is that faced by two expert opinions, neither of which you understand, you might as well go with the one whose conclusion you like, lest the other one strangulate the baby.

OK, go for it.

But the best way to determine which is the more convincing of two experts with opposing views is to subject both those views to peer-review. If Ron believes his re-analysis of the ESI Ohio data will pass peer-review in a relevant journal (i.e in the fields of political science or public opinion research) then I suggest he submits his paper to such a journal.

The ESI paper from which he derives his data, is, I understand, currently undergoing such a review process.

The arguments of one side are sophisms (plausible but misleading or fallacious arguments).

No reputable science journal has agreed to publish the disproven (but highly creative) sophisms that mascarade as 2004 exit poll analysis of Mitofsky, the Election Science Institute (including Fritz Scheuren former ASA President), Liddle, or Lindeman - nor, I predict, will they be published by any reputable journal.

The National Election Data Archive recently published short compendium of failed attempts to dismiss the evidence of vote miscounts in the 2004 election:

http://electionarchive.org/ucvAnalysis/US/IncorrectElectionDataAnalysis-06.pdf

If you can understand those six pages, then you will KNOW who the real exit poll analyst experts are.

Why has Mitofsky not released the data for any standard, customary, accepted for generations, analysis of his exit poll data - which simply requires precinct-level (blurred would be fine) vote counts, sample sizes, and exit poll results - for even one state?

No scientific theory is proven until it can be replicated by independent analysts.

Why has Mitofsky not released the data (exit polling factors that would not risk voter privacy) that he claims proves his case?


The reason Farhad's article is rife with innacurate information, is that he openly relies on Mark Lindeman (and also Mark Blumenthal) as his primary sources, who in turn rely on Liddle. Read the above short paper to know why none of Liddle or Lindeman's exit poll analysis is correct.

Always go back to the primary sources to check anything that Lindeman or Liddle claim. Farhad neglected to do that.

to go back to the original source of this post.

But let me ask, as you seem to be in a position to know:

Have any of the NEDA publications been published in a "reputable science journal"?

And do you have any comments to make on the fact that the NEDA document you link to has three easily confirmed misattributions on the first page alone?

I, like you, am a believer in checking "primary sources", which is why, when I checked the "primary sources" cited in the linked document I found that they did not contain the statements attributed to them.

Careless, I call it.

Elizabeth Liddle

Thought this might be of interest to you folks:

MONDAY, JUNE 05, 2006
Manjoo Critique in Salon is Superficial and Erroneous Nonsense

In his June 2, 2006 Salon article “Was the 2004 Election Stolen? No”, Farhad Manjoo claims to have “thoroughly debunked” a June 15, 2006 article in Rolling Stone by Robert F. Kennedy Jr.: “Was the 2004 Election Stolen?”.



As one of the (applied) statisticians cited in Kennedy’s article, I find that Manjoo’s “debunking” is either superficial spin that is easily refuted by adding more detail to Kennedy’s already very long piece, or simply factually erroneous. I will focus on the Manjoo points that relate to official return and exit poll data - my particular area of expertise.



a) Making the point that voters sometimes do vote more often for a “down ticket” candidate is hardly a “rebuttal” of the 12 County Ellen Connally “anomaly”. First of all this was not a normal election but rather one of the hardest fought Presidential races in recent history in a key battleground state. Moreover, Connally was an under-funded and largely unknown African American, liberal, pro-choice, pro-gay rights, anti-death penalty, woman from Cleveland, and all of these counties are far from Cleveland. But even if we disregard this, anybody who wishes to check the numbers posted on the Secretary of State of Ohio’s official website will find that: a) nine out of only 14 counties (out of a total of 88 counties in Ohio) where Bush did abnormally well (over 43% better) relative to Moyer (the Republican Supreme Court nominee) just happen to be part of the same 12 counties where Kerry inexplicably did worse than Connally, b) this was not a result of voters in these counties switching from Moyer to Connally as Connally’s vote is worse than the Ohio county average in all but one of these counties and Moyer’s vote is better than the county average in all but one of the 12, c) the “lost Kerry” vote of 81,599 (relative to the average Kerry/Connally ratio) just happens to be just about equal to the “excess Bush” vote of 75,766 (relative to the average Bush/Moyer average in the 12 counties), and d) when one looks at the county by county pattern of “lost” Kerry votes to “excess” Bush votes, they match almost perfectly in their orders of magnitude – see the table below:

<snip>

http://www.baiman.blogspot.com/