Left2Right

Comments

Posted by: Tom Perkins

For various reasons I don't care to elaborate right now, the night of Nov. 1st I drove to Columbus Ohio from northern Viginia to work the phones for the Republican GOTV campaign, the morning of the 2nd I drove back to work. I've never done anything like that before--I'd never given money to any political campaign, only to to the Libertarian party. I do not at this time anticipate supporting the Republicans or any other party with like efforts in the future, I'll wait and see what and who they come up with. I was on the phones from about 7am to 7:15pm. I can tell you from personal experience that at about 5pm and on, I increaqsingly called teenage children and babysitters who reported that, "Mommy and Daddy went to vote for the President, they're not hear right now."

That's nothing I heard even once before 5pm. I do believe the earliest polls exagerated support for Kerry. It was a sampling error, likely caused by the work or other personal habits of the people voting that early in the day.

It is no mystery to me that early exit polls were wrong, the Bush voters were working for their living, they voted late. At 7:15, I was still telling them (by voice messages, gennerally they weren't home) on the phone that if they got to the polls by 7:30, they (the poll workers), had to let them vote.

I know lots of Republicans in rural areas waited in line to vote in cold wet weather as late as 11:30--it was not a problem unique to heavily Democratic urban areas.

That's all there is, there's no conspiracy to steal OH from Kerry.

I wouldn't be so adamant about it if I wasn't there to see some of it for myself.

You lost. GET THE FUCK OVER IT!

Yours, TDP, ml, msl, & pfpp

Posted by: Tom Perkins | Jan 31, 2005 10:23:43 AM

Posted by: Tom Perkins

I should add that I am sure that each party did their level best to win, and that I do not think that there were more shenanigans than usual in this election.

But. I agree that voter verified paper records of votes are required for electoral integirty to be maintained. This is obvious.

It is also obvious to me that the Democrats resist equally obvious measures required for the integrity of the electoral process.

So I tell you what (although I think agreement should be untroublesome to all parties of honest intent), we'll trade you voter verified triplcate records for voter registration mechanisms which ensure legaly valid voters vote only once and in their precincts of legal residence, and only if they are of good legal standing (not felons, are citizens).

How does that strike you?

Yours, TDP, ml, msl, & pfpp

Posted by: Tom Perkins | Jan 31, 2005 10:36:43 AM

Posted by: Paul Deignan

Exit polls rely on voluntary participation and exposes the respondent to costs and privacy concerns that the actual vote does not. Furthermore, the exit polls were conducted by one firm at a particular time and particular locations.

Now, consider that some other polls were wildly off on the eve of the election -- esp. Zogby. Other polls tended to be consistent over the run up to the election such as Pew.

Note also that there is a persistent trend among some to systematically make very very poor predictions, e.g. the outcome of the Iraqi elections, Afghanistan elections, Nicaragua, El Salvador, the USSR's demise etc. There is a trend here.

Furthermore, it has been found that there is a statistically significant discriminant between the characteristic information processing methods (personalities) of conservatives and liberals in such a way as would explain the systematic failure of liberals to predict events.

The question is, "Are we ready to deal with reality?" This is a very good time for examination and analysis. The US functions best with two functional political parties.

Posted by: Paul Deignan | Jan 31, 2005 10:46:23 AM

Posted by: Paul Deignan

Here is a link to polling leading up to the election -- there are internals here as well: Polls

Posted by: Paul Deignan | Jan 31, 2005 10:50:58 AM

Posted by: noah

How about random bias in favor of Kerry on the part of the exit-poll takers? I have read that there is evidence of such bias among the younger cohort.

There is much more palpable evidence that the Washington governors race was stolen by the Democrats...why does this not concern you? HMMM!?

Posted by: noah | Jan 31, 2005 10:52:37 AM

Posted by: CDC

Fraud is possible, but not likely. Too many people with loose tongues would have to be involved and too many LIBERAL DEMOCRAT reporters would be hot on the trail. It positively would come out and the feeding frenzy would make any other political scandal in American history seem trivial. The potential downside so outweighs the potential upside, and the odds of success are so small, that a fraud of this type would be stupid to attempt. Rove isn't stupid.

I think it's much too early to give up on case 2.

Posted by: CDC | Jan 31, 2005 10:55:42 AM

Posted by: Tom Perkins

Ah shoot, that should read, morning of the 3rd. Right around where 522 cuts south from I70, I was starting to get a little blurry headed. If Bush pulls of some lese majeste with respect to the scared cow of Social Security, it'll be well worth it. Yours, TDP, ml, msl, & pfpp

Posted by: Tom Perkins | Jan 31, 2005 10:58:26 AM

Posted by: pedro

Very informative post, Paul V. It's interesting and dispiriting to witness how the problem of making sense of the discrepancies between exit polls and vote tally--a problem which requires technical statistical knowledge to be tackled sensibly--is easily explained away, with formidable simplitude, by people with little knowledge of the statistics involved, let alone of the data.

Posted by: pedro | Jan 31, 2005 11:09:46 AM

Posted by: Tom Perkins

Pedro, I was there on the phones. Were you? Yours, TDP, ml, msl, & pfpp

Posted by: Tom Perkins | Jan 31, 2005 11:13:15 AM

Posted by: pedro

So, I'm to rely on your anecdotes, Tom? Do you really think I'm naive enough to dismiss the observations of professional statisticians simply because you worked the phones for the GOP?

Posted by: pedro | Jan 31, 2005 11:55:24 AM

Posted by: D.A. Ridgely

As one of those formidably simplitudinous folks who can’t tell a chi square from a chi rho and whose knowledge of statistics doesn’t extend much beyond confidence in the belief that 49% of all statisticians are below average, put me down for “all of the above.”

I would never rule out the possibility of some voter fraud in any election. If anything, I think I would be more inclined to rule out the possibility of no voter fraud. Even so, Mr. Velleman doesn’t seem to be considering that there could be multiple factors contributing to the disconnect between the polls and the results. Moreover, the fact that plausible explanations have not yet been forthcoming is hardly the same thing as there being no such plausible explanations.

In any case, and I say this as someone who is sufficiently innumerate that I actually do break down and buy a lottery ticket from time to time, can’t a penny randomly turn up heads a hundred times in a row? I mean, don’t flukes happen?

Posted by: D.A. Ridgely | Jan 31, 2005 11:58:47 AM

Posted by: CDC

Pedro: "...a problem which requires technical statistical knowledge to be tackled sensibly--is easily explained away, with formidable simplitude, by people with little knowledge of the statistics involved..."

Thank you for your concern but the report seemed pretty straight forward.

Posted by: CDC | Jan 31, 2005 11:58:57 AM

Posted by: Tom Perkins

My anecdote would I think be repeated as the experience of all present.

You can conclude we are all in on the conspiracy or conclude the initial exit polls were at best accurate for the people having voted at the time--who were not nearly everyone who was going to vote that day.

What says Prof. Velleman with regard to my personal experience. Has he interviewed other GOTV workers? Would he regard the results of those interviews as credible?

And what would he say about both parties multilaterally disarming with respect to their putative favored vote fraud methods? A trade of voter verified paper records for systems making it very difficult for someone to vote more than once and then only if they are citizens legally able to vote in the precinct where they show up, and actually prosecuting fraudulent voters?

Yours, TDP, ml, msl, & pfpp

Posted by: Tom Perkins | Jan 31, 2005 12:17:49 PM

Posted by: Will Curtis

I think the US Counts Vote brief is interesting and their hypothesis is plausible. But if I were a betting man I would propose the following hypothesis of my own:

Instead of suggesting that their was a nationwide effort to systematically shift the vote in favor of Bush, why not suggest that there was a systematic nationwide effort to shift the exit poll data.

This hypothesis is more likely, in my view, because whereas vote counting is overseen and scrutinzed in the various states and counties by people *from both parties*, the collection of exit poll data is not done with any overt counter-fraud measures. Thus, it is the *exit poll* fraud which scares me, not the much less likely vote-count fraud.

How much easier would it be to skew exit poll results? Think about it.

Posted by: Will Curtis | Jan 31, 2005 12:20:26 PM

Posted by: mikec

As a libertarian who voted for neither Bush nor Kerry, I saw a huge difference in attitude between Bush and Kerry supporters. Kerry supporters assumed I was one of them. They started conversations with statements that would have been insulting had I been a Bush supporter: implying that Bush was stupid and evil, and that if I supported him, I was, too. Bush supporters were much more reticent about their views; in fact, they usually didn't state a preference at all unless they got some encouragement.

My guess is that, by election day, quite a few Bush supporters were tired of being insulted and demeaned and decided to keep their mouths shut.

Posted by: mikec | Jan 31, 2005 12:48:28 PM

Posted by: pedro

"You can conclude we are all in on the conspiracy or conclude the initial exit polls were at best accurate for the people having voted at the time--who were not nearly everyone who was going to vote that day."

Or you can refrain from jumping to conclusions before the data is examined properly. Nobody is arguing that it is pertinent to conclude right now that there was significant fraud in the election. But it is perfectly reasonable to seriously scrutinize all possibilities, including the possibility that exit polls coincidentally happened to lose accuracy in precincts with a particular kind of voter technology.

In the aftermath of the 9-11 attacks, a number of financial experts speculated that it would be possible to detect risk of attacks by examining patterns of unusual option market activity, in particular fluctuations in the case of put/call volume ratios. A normal put/call ratio is in the neighborhood of 1, yet in the days before 9-11, both put/call ratios for American and United were exceptionally high. Does it make sense to analyze the data so as to have predictors of risk of attacks? Yes, it does, regardless of how promising the method may be.

Similarly, it behooves us to analyze patterns in the voting data to evaluate risk of fraud. Whether systematic, widespread, significant fraud occurred is very hard to know, but I'm inclined to disbelieve it. But a careful examination of the data by professional statisticians, and a careful evaluation of the voting technologies by professional engineers is entirely appropriate.

Posted by: pedro | Jan 31, 2005 12:54:55 PM

Posted by: Paul Deignan

I perceive a political bias in the MSM that makes me wary of pollsters. Why? Generic fear of contributing to an effort that might misrepresent me such as the fear of being misquoted and scandalized by unethical partisan reporters. Unfounded? Unfortunately not. We can read the pages here (pedro -- you are guilty of this). The fact that this tactic is employed by one party disproportionately is a function of their internal dynamics (see the blog).

I will bet, that there is a politically related dependency between response rates and party affiliation/voting preference. In fact, there are reports that tend to support this proposition.

There is a reason for the polarization -- that is not by chance either as the red-blue maps should make clear to all.

Posted by: Paul Deignan | Jan 31, 2005 12:57:16 PM

Posted by: Paul Deignan

Here is one link: Exit Poll

Posted by: Paul Deignan | Jan 31, 2005 1:14:13 PM

Posted by: Paul Deignan

BTW, for those interested, the reaon for the intimidation is directly related to the personality shift between liberals and conservatives. Liberals feel a need for agreement and ratification while conservatives are more comfortable with rational disagreements.

The results are here: Personality Study Results

Posted by: Paul Deignan | Jan 31, 2005 1:20:57 PM

Posted by: Tom Perkins

pedro, it is only prudent to expend resources scrutinizing serious possibilities. There is as yet only thin statistically derived evidence of any vote fraud/irregularity in OH, and I can with personal experience provide testimony that would tend to explain such thin statistical evidence as resulting from something other than fraud.

The put/call ration for the corporate stock you mention would first have to be shown to be at least a few standard devaitions away from the delta of such ratios in like circumstances before I would call it predictive.

Will Curtis' mention of "exit poll" fraud deserves at least as much attention as investigations of "vote" fraud.

And in the vein of that meme, "a careful evaluation of the voting technologies by professional engineers is entirely appropriate" Prof. Velleman certainly may be a professional statistician, by appearing in this forum in this fashion, he also declares himself to be a partisan one. Curtis' possibility would I think have been possibility number 4 in a list of explanation compiled by a perceptive, uninvolved person.

This list:

"1. Chance error,
2. Bias in the exit polls, and
3. Inaccurate election tallies, or (say it softly) election fraud."

Is incomplete.

Yours, TDP, ml, msl, & pfpp

Posted by: Tom Perkins | Jan 31, 2005 1:21:16 PM

Posted by: pedro

Paul D.: as usual, you make little sense to me. (If only you took it upon yourself to write more clearly.) You accuse me of behaving like an unethical partisan reporter, yet you provide no argument to support your ridiculuous accusation. I imagine you must be reading between the lines.

Posted by: pedro | Jan 31, 2005 1:22:35 PM

Posted by: CDC

Pedro: "But a careful examination of the data by professional statisticians...is entirely appropriate."

Fine. Then publish it.

Posted by: CDC | Jan 31, 2005 1:24:12 PM

Posted by: Doug

Dr. Velleman, I had read a (to this layman) plausible analysis of the exit poll failures on Mystery Pollster a few weeks back. The systemic factor he identifies is young, undertrained exit poll workers. He provides a couple of hypotheticals about how a poorly-trained college student might unwittingly pick a non-random sample or undercount voters with which she disagreed.

Do you agree with that analysis? Or if you don't feel it's supported by the data you examined, why not?

Posted by: Doug | Jan 31, 2005 1:24:35 PM

Posted by: AlanC9

The issue isn't that the exit polls underreported one side or the other. That could have a lot of a reasonable explanations. After all, Bush and Kerry voters are different from one another.

There's an issue if the variation between the reported vote and the exit polls is related to the voting technology. That shoudn't happen, unless there's some sort of demographic distinction between districts with electronic voting machines and disctricts without them.

Is there? Or is the variation not really related to the voting technology? D.A., Paul, tom.... any explanation for me? (Believe me, the last thing I want is to actually have to think there's a problem here.)

Posted by: AlanC9 | Jan 31, 2005 1:42:48 PM

Posted by: Paul Deignan

AlanC9

That shoudn't happen, unless there's some sort of demographic distinction between districts with electronic voting machines and disctricts without them. Is there?

Yes, yes, yes, yes, yes,yes, yes, yes, yes!

There is a statistically significant intrinsic difference between partisan democrats and republicans -- personality!

This factor is exactly the right factor in the right proportion to exactly explain what we are seeing here. Believe it or not, this is an original observation (the fact that the difference is measurable and significant).

Posted by: Paul Deignan | Jan 31, 2005 1:55:32 PM

Posted by: pedro

Tom: unsurprisingly, some research has indeed been conducted regarding the violation of put/call parity in the leadup to 9-11. I'm not sure whether there's any predictive value in looking at substantial deviations from put-call parity, but the issue is one which competes the relevant professionals to evaluate, and it seems like they have evaluated it. Allen Poteshman, for example, has been studying how rare violations of put-call parity are in the stock market.

As for the possibility that only pollers in precincts with certain type of voting technology chose to commit fraud on the public by reporting biased numbers, I wouldn't find it very compelling. But I don't wish to speculate on the data, because I don't have it. Precinct-level data will help statisticians evaluate the possibilities.

Posted by: pedro | Jan 31, 2005 1:56:23 PM

Posted by: AlanC9

Paul, interesting study on your blog. Be nice if you broke it down by [i]types[/i] of liberals and conservatives, since those two labels aren't all that coherent. Maybe next time.

Posted by: AlanC9 | Jan 31, 2005 1:56:53 PM

Posted by: AlanC9

Paul, what does that have to do with the voting method in the district? Are majority-conservative districts more likely to have touch-screen voting?

Posted by: AlanC9 | Jan 31, 2005 1:58:48 PM

Posted by: Tom Perkins

AlanC9,

Actually, the issue IS "explaining the exit polls."

Prof. Velleman writes: "E/M have not released precinct-level data, which would be necessary to determine whether voting technology is a factor." So that is not an issue being examined by anyone yet, I suppose.

I personally am unaware of any assembled data at all breaking down precinct results vs exit poll discrepancies by polling technology for OH. Numerous examples are available for FL on this page, although they mention "by county" not "by precinct" data.

http://yalefreepress.blogspot.com/2004_11_01_yalefreepress_archive.html

Perhaps Prof. Velleman, or Dr. if he prefers, will able to supply that complete with any fudge factors or coefficients used in calculations, along with the data and equations/methods used, if and when it is available.

Yours, TDP, ml, msl, & pfpp

Posted by: Tom Perkins | Jan 31, 2005 1:59:39 PM

Posted by: AlanC9

(that's what I get for going back and forth between boards with UBB code and HTML boards)

Posted by: AlanC9 | Jan 31, 2005 1:59:48 PM

Posted by: Paul Deignan

AlanC9,

It was necessary to have only a single output dimension in order to have statistically significant results from that sample. A future study involving thousands under better control should include a political ideology questionnaire as you suggest.

I don't think any correlation with voting technology is causal in the exit polling discrepancies. I would, however, expect that the discrepancy be nonuniform and dependent on the subcultures of the areas and also upon the pollster demographic (there is probably insufficient data to draw any such relationship).

Do expect the discrepancy to be nonuniform and locality dependent.

Posted by: Paul Deignan | Jan 31, 2005 2:04:50 PM

Posted by: Paul Deignan

Pedro,

Paul D.: as usual, you make little sense to me. (If only you took it upon yourself to write more clearly.) You accuse me of behaving like an unethical partisan reporter, yet you provide no argument to support your ridiculuous accusation. I imagine you must be reading between the lines.

Just waiting for you to bite:

Here is your post: Pedro's post

Posted by: Paul Deignan | Jan 31, 2005 2:13:46 PM

Posted by: pedro

Paul: my post was a response to this.

Posted by: pedro | Jan 31, 2005 2:31:05 PM

Posted by: Chris

When Lou Holtz was head coach at Notre Dame they played someone (was it Michigan?) and they won the game. However, during the post game press conference the reporters being the expert statisticians that they are were quizzing Holtz on the fact that the opposing team had more first downs, more total yardage and the like. Holtz’s reply was that he thought it was the score that mattered and not the number of first downs and that if the winner was based on first downs then he would have made sure that he had more first downs than points.

Posted by: Chris | Jan 31, 2005 2:34:17 PM

Posted by: pedro

Incidentally, Paul: I would agree with someone who contests the idea that homosexuality is like being tall. My point was simply that even people who choose to be homosexual ought to have the same rights as those who choose otherwise.

Posted by: pedro | Jan 31, 2005 2:39:29 PM

Posted by: Paul Deignan

Pedro,

We do not claim that our interpretation (false or true) of another's statement is their own. Your intent was clear. I have seen it used before. It is a defense mechanism, advertisement of group membership, and warning to others.

Note here that you are directly addressing me in a show and tell/ask manner. This is the normal mode of rational disscussion when the purpose is to exchange information. Here we drop our defenses and open ourselves to ideas. "Do you see what I see? How do you think that? Why?"

The two methods of communication have different objectives. We can communicate on either level, but a mismatch of methods leads to noise and polarization. This is an implication of the personality study.

Posted by: Paul Deignan | Jan 31, 2005 2:58:46 PM

Posted by: Paul Velleman

A few quick responses to questions and observations raised here:

1)Tom P: The E/M data are from their third wave of polling, accumulated at
the end of the day. Any observations about who voted early or later are
irrelevant and cannot explain the patterns. Read the E/M report: they make
that clear.

2) Paul D: There is indeed some evidence about differences between
Republicans and Democrats in responding to polls. Pew Research has twice
performed the experiment in which they execute a standard 4 or 5-day
telephone poll and at the same time execute an intensive 3-month effort to
contact a parallel sample and ask the same questions. The first poll gets
completed questions from about 25%, the intensive one from about 75% of
the sample. They find that the telephone polls match the more thorough
survey with few differences, but one difference was that Republicans were
*more* likely to respond to the first poll -- by about 6%. If you know of
a study showing otherwise, I'd like to see it. This is the best study I
know of thus far.

3) Noah, Doug (and others): E/M looked for bias patterns vs age of the
interviewer (their interviewers coverd the full range of ages) and found
none. See their report.

4) Paul D: As I pointed out (and see the USCountVotes report) the only
pattern against party affiliation goes slightly in the *wrong* direction
for the E/M explanation. There's no evidence otherwise that I know of.

5) Exit poll "fraud" is an interesting thesis, but again, it calls for
some corroborating pattern. Should this have happened equally in all
states? More in Red or Blue? Could it possibly be related to voting
technology? I don't see any patterns in the data that support this and
some that seem to contradict it.

6) Voting technology. Yes, the E/M report includes a table (repeated in
the USCountVotes report) showing that voting technology has a strong
relationship with the bias. That is the most suspicious result in their
data both becuase it suggests fraud and because it contradicts most of the
alternative explanations.

7) Could the errors we've seen have happened by chance? Basically, no. As
a statistician, it is the virtual impossibility of this happening by
chance that caught my attention and that demands explanation. And E/M
readily acknowledge this in their report. It is the failure of the E/M
explanation (or others such as those proposed by discussants here) that
concerns me and raises the question of other causes.

Posted by: Paul Velleman | Jan 31, 2005 3:13:06 PM

Posted by: John F. Opie

Hi -

Gee, only three choices?

I don't mean to be snarky, but there are at least two others:

4) Incompetence on the part of the exit polling companies, their pollsters and statisticians

and

5) deliberate miscalculation to negatively impact the election results.

Now, I don't know whether this is the case: but ignoring these possibilities is at least as politically biased as your 3.

And having a great reputation doesn't protect you from incompetence and politically motivated bias. Even the best of companies can make mistakes.

Personally, I think it was bias in the exit polls. But if the exit polls were biased, this raises the question of 4 or 5: someone screwed up.

And after reviewing the article in question, I think that you are perhaps missing the point: exit polls don't tell you for whom people voted, it tells you what people are telling exit pollers for whom they voted. That's perhaps the fundamental flaw.

John

Posted by: John F. Opie | Jan 31, 2005 3:23:57 PM

Posted by: No Labels Please

The argument cited in the article that the sampling error should go the other way [response rate sightly higher in pro bush distrincts] is badly flawed.

What's important is not the absolute response rate for all voters, but the relative response rate between Bush and Kerry voters.

For example, it's easy to hypothesize that in smaller, more rural areas [more RED/BUSH areas] all voters are more amenable to being polled than voters in NYC and Los Angeles [I don't know that this is true - but it passes the smell test]. ***This however, would not imply anything about the relative repsonse rates of Kerry/Bush voters in these areas or in any other areas.***

And therefore, this overall increase in resposiveness in "Bush" states wouldn't necessarily lead to any bias at all in the calculation done.

This does not reflect favorably on the "statisticians" who wrote this article.

Comments, Mr. Velleman?

Posted by: No Labels Please | Jan 31, 2005 3:27:45 PM

Posted by: No Labels Please

I'd also like to propose an alternative explanation for the data:

Many people who actually voted for Bush lied and said they voted for Kerry in the exit poll.

It seems to me that supporting Bush became socially stigmatized for many people - ie extremely unpopular and "uncool". This was a result of the widespread unpopularity of the war and the virulence of the anti-Bush opposition

For example, many of us probably noticed that although the pre-election polls were about 50/50 Bush/Kerry, you would think that Kerry would win in a landslide if you observed the vocal anti-Bush sentiment at social gatherings, in academia, in the press, with the Europeans, etc...

My suspicion is that many people voted for Bush who may have not wanted to admit it in an exit poll becasue of latent fear of social stigmatization or actual genuine embarassment as their support was lukewarm.

Posted by: No Labels Please | Jan 31, 2005 3:42:55 PM

Posted by: frankly0

Paul Velleman,

My problem with the notion that it is election fraud that lies behind the exit poll discrepancies lies on the other end, in the inherent plausibility of your 3). I just don't see how election fraud could have been pulled off on such a massive scale that it wouldn't leak out.

Were partisan Republicans truly in positions where this scope of fraud might be perpetrated? How might this have been carried out? If you can't come up with a scenario whereby this might have happened, why should we entertain it? It has been primarily a failure to come up with such an account that has put me off from considering the possibility of election fraud.

I understand the political sensitivity of the point, but, honestly, I don't think it's quite right to simply suggest the possibility without providing some sketch at least of how it might have been made to work.

Posted by: frankly0 | Jan 31, 2005 3:45:13 PM

Posted by: Paul Deignan

2) Paul D: There is indeed some evidence about differences between Republicans and Democrats in responding to polls. Pew Research has twice performed the experiment in which they execute a standard 4 or 5-day telephone poll and at the same time execute an intensive 3-month effort to contact a parallel sample and ask the same questions. The first poll gets completed questions from about 25%, the intensive one from about 75% of the sample. They find that the telephone polls match the more thorough survey with few differences, but one difference was that Republicans were *more* likely to respond to the first poll -- by about 6%. If you know of
a study showing otherwise, I'd like to see it. This is the best study I know of thus far.

One such link was provided in post immediately following and directly relates to this election.

4) Paul D: As I pointed out (and see the USCountVotes report) the only pattern against party affiliation goes slightly in the *wrong* direction for the E/M explanation. There's no evidence otherwise that I know of.

One cannot draw this conclusion from the trend showing higher exit polling turnout in nominally Republican precincts. We need to know who responded and who balked -- cannot assume that they were representative of the precint unless under these conditions. That is a grave error.

Since the exit polls are voluntary, it is impossible to measure an unmeasurable. You need to capture a random sample -- not a self selected sample. This is always one of the greatest problems with these type of polls.

Posted by: Paul Deignan | Jan 31, 2005 4:16:23 PM

Posted by: Tom Perkins

While I have only scanned this document online:

http://uscountvotes.org/ucvAnalysis/US/USCountVotes_Re_Mitofsky-Edison.pdf

I find no data here about when a third wave of exit polling took place, nor any comparison about differences between inaccuracies of exit polls taken at different times of the day.

This document:

http://exit-poll.net/election-night/EvaluationJan192005.pdf

has such information, but it perforce cannot say whether my experience was coincidental or material. It can only state what the error was, and does not show any correlation between polling technology and error, other than to show that the electronically recorded ballots had a much higher error than the maligned paper ballots--but the error corresponded strongly to the size of locality wherein the precinct was located. Larger locales had larger errors, and these were larger across the board, not correlated to polling technology. If this pattern held as strongly across strongly blue and strongly red states, it militates against such a correlation being related to fraud--unless suspicious persons are willing to credit a VERY large conspiracy with members almost having to come from differing parties.

There is also no evidence the error was particularly large in especially close states, or is states of an electoral college size enablng them to decide the election, had they gone another way.

Weather had almost as large an effect on the WPE (within precinct error) as whether a state was a swing state--not something I would expect if fraud was at issue. You'd think if someone were to commit a felony of this magnitude, they's eleiminate weather as a variable.

Also, the average error within precinct for OH and for CA were the same(10.9), and SC (10) was very similar. In each case, a state is either uncontestedly Dem (CA) or Rep (SC) or a swing state. Why are there errors so similar? Why is this pattern repeated over many election cycles? Why is UT consistently a low error state, and can that performance be duplicated?

Again, this is a first impression after skimming the report.

Exit poll errors generally increased in the year 2004, and did so without regard to the advantage to any party in any state.

Something has changed and the work done in the 70's and 80's to make exit polls more accurate needs to be done again.

And not to take an off topic track without good reason, but would Dr. Velleman endorse both voter verified paper records and rigorous mechanisms to prevent people from voting illegally?

Does Dr. Velleman think my experience of a sharp rise people voting for Bush after 5pm (many of whom would not have exited a polling place until long after 7:30pm, owing to the lines) is coincidental? Or irrelevant?

Yours, TDP, ml, msl, & pfpp

Posted by: Tom Perkins | Jan 31, 2005 4:26:13 PM

Posted by: Paul Deignan

Paul V.

Note that the Pew poll was telephonic – not a face to face interaction. I will wager you that the discrepancies were especially bad in the precincts where a young pollster was stationed (I have not seen the data). Not that the pollster was necessarily inexperienced, but that the respondent identified the pollster with partisans Democrats. If Democrat -- a higher response rate, if Republican -- a higher balk rate.

So, perhaps if we knew the balk rate and the age/gender of the pollster we might make some use out of the nominal composition of the precincts. Of course, we would also need to normalize for time of day.

I'll also wager that we have insufficient data to say much here.

Posted by: Paul Deignan | Jan 31, 2005 4:28:45 PM

Posted by: Paul Deignan

So on that Pew poll -- the more face to face, the more likely the Republican to balk.

This is the type of results that support my hypothesis of a social interaction in balking by political affiliation.

Posted by: Paul Deignan | Jan 31, 2005 4:45:24 PM

Posted by: Paul Deignan

Here is that link again: http://www.wm.edu/news/?id=4027

It also lends support to the social interaction theory (not otherwise). Right now, we have a pattern.

Posted by: Paul Deignan | Jan 31, 2005 4:47:49 PM

Posted by: oliver

Actually, in theory the biases could be due to chance, although as Prof. V says that's really really unlikely. What they can't be the result of is a fluke. "Fluke" suggests the thing defies analysis. But for every event bigger than the quantum mechanical movement of electrons all there is to deal with is simple cause and effect. Statistics isn't the sharpest tool in the drawer, but analysis is far from hopeless.

Posted by: oliver | Jan 31, 2005 5:52:21 PM

Posted by: oliver

Well, I guess there are big non-linear systems like the atmosphere that give rise to mathematical chaos, which looks kinda flukey. I retract.

Posted by: oliver | Jan 31, 2005 5:59:37 PM

Posted by: oliver

We might be able to exclude chaos as a cause of these biases based on the short time interval over which they happen. Chaos I think would require that people's votes depended on each other, which of course is true, but your friends probably don't change your mind in the course of an hour on election day. Yeah, I think we can rule out chaos along with quantum mechanics.

Posted by: oliver | Jan 31, 2005 6:05:55 PM

Posted by: Paul Shields

My theory is that there is a critical mass of shy Republicans

Posted by: Paul Shields | Jan 31, 2005 6:08:31 PM

Posted by: oliver

About exit polling methodology, the report says "In constructing a prediction for statewide outcomes, algebraic weightings are used tocorrect for the observed demographic composition of the sample. For example, responders are re-balanced
by race and gender in this process to assure a representative sampling of the state."

I find this troubling. As I read it, the analysts make up for differences between the demographics of their sample versus the region or state as a whole by assuming that people in demographic category X in the sample vote Bush v. Kerry at the same ratio as the county or region or state overall. I guess nationally race is a meaningful demographic category for African Americans, given reports that 95% or whatever vote democratic, but how many such high-quality predictors are there, and are they good not just nationally but county by county? e.g. Are the women here really going to vote just like the women there?

Posted by: oliver | Jan 31, 2005 6:26:30 PM

Posted by: pedro

Paul Deignan: How on Earth is my post a warning to others, when I agree that it is a mistake to essentialize homosexuality? I might as well have said: Paul and I seem to agree that it is not a good idea to say that that all people do not have choices when it comes to homosexuality, yet I view this as entirely irrelevant. Perhaps my choice of verb was obscure: to contest to me is synonymous with to dispute. You seemed to be disputing the idea that homosexuality is not--in some level--a matter of choice. I do not disagree. I simply think that it is irrelevant.

As for your psychoanalysis, it is dreadful, and guilty of the very thing you seem to object to: attributing to others intentions they reject. Forgive for having mischaracterized you--if indeed I did do so--on the post in question. I thought I was agreeing with you in some level, and so I saw no reason to take offense.

Posted by: pedro | Jan 31, 2005 6:39:48 PM

Posted by: oliver

I accidentally made my last comments under the wrong post. They were....

Ah, I see now they analyzed for what was troubling me about weighting and found no trouble. I can believe that.

Posted by: oliver | January 31, 2005 06:29 PM | Permalink

Posted by: oliver

Wow. Just finished the report. I find it very persuasive, and very troubling. Definitely calls for deeper inquiry--as well as a response from the group that offered the initial explanation of the bias. Are these exit poll statisticians generally a bunch of total losers who couldn't get a real statistics job? I thought this business attracted bright people.

Posted by: oliver | January 31, 2005 06:42 PM | Permalink

Posted by: oliver

I hope the panel of statisticians who produced this reanalysis was bipartisan or at least had a token Republican. Otherwise I bet there will be some heat and the general public might feel Republicans who oppose more analysis would have a talking point.

Posted by: oliver | January 31, 2005 06:46 PM | Permalink

Posted by: oliver | Jan 31, 2005 6:49:38 PM

Posted by: oliver

"My suspicion is that many people voted for Bush who may have not wanted to admit it in an exit poll becasue of latent fear of social stigmatization or actual genuine embarassment as their support was lukewarm."

But if their motivation is as you hypothesize, then one would expect the exit poll response rate for people who claim to be Republican would be less in precincts where most or all people were Republican--to the tune of the 6% difference we need to account for. But the graph in the report shows that's not so.

Posted by: oliver | Jan 31, 2005 6:56:28 PM

Posted by: oliver

Sorry, that was confused thinking on my part above about the implications of the graph for the lying Republican hypothesis. I take that back.

Posted by: oliver | Jan 31, 2005 7:05:28 PM

Posted by: Paul Deignan

Pedro,

It is possible that I am just a skittish conservative that misinterpreted your mode of communication. If this was an invitation to discussion, however, I am at a loss to account for our lack of follow up thereafter.

BTW, I wasn't thinking of psychoanalysis so much as herd behaviors. The herd model seems to account for a good deal of human behaviors as well.

Posted by: Paul Deignan | Jan 31, 2005 7:06:16 PM

Posted by: pedro

Paul Deignan:

I didn't follow up because I was confused by your response. You said that I ought to consider the argument better, and I frankly did not understand where I had got you wrong. I find it difficult to follow you sometimes. Perhaps it is mutual? Please accept my apologies if I misrepresented--in your mind--what you said. I certainly thought I hadn't understood your point at all when you replied the way you did, and rather than trying to parse your meaning, I decided to drop the subject.

Paul V.: You said, in your original post, that E/M has not released precinct-level data. I wonder if you have any idea whether there is some pressure for them to do so soon? I do think it would help illuminate or dispell possible problems with technology.

Posted by: pedro | Jan 31, 2005 7:21:28 PM

Posted by: Paul Deignan

Pedro,

You might like this one: http://courseweb.stthomas.edu/kwkemp/Papers/HR.html

It was left as "an exercise for the reader". So maybe there is more that we agree about. To reach a possible consensus and to explore the various ideas, it would have been necessary to continue the discussion. There was an inconsistency with your point that was mentioned in the above link and that we later discussed in that thread.

Note that you stated a position and claimed a piece of ground for yourself in the discussion. That was very instinctual.

Posted by: Paul Deignan | Jan 31, 2005 7:31:29 PM

Posted by: NetworkSpokesperson

We now manufacture all election results.

We are not obligated under law to disseminate the truth about anything.

Posted by: NetworkSpokesperson | Jan 31, 2005 7:34:40 PM

Posted by: Detached Observer

Interesting stuff. Thanks for bring the E/M and USVotes reports to our attention.

One issue not addressed here is exit poll biases in previous elections. These tended to be very consistently for the Democratic candidate, though never on as great a scale as 2004. But if fraud is accepted as a hypothesis, then we must accept that such fraud occured in each of the past five elections, which is considerably more difficult to believe.

Posted by: Detached Observer | Jan 31, 2005 7:57:36 PM

Posted by: DBCooper

I think New Hampshire is key in understanding what happened. New Hampshire was the battleground state that exhibited the greatest variation between the early exit polls and the final result….Kerry led by 10.8% in that exit poll. In the end Kerry won by only 1.3% margin, for a deviation of 9.5% from the predicted value.

A statistician for the State of New Hampshire analyzing the results precinct by precinct, found what she believed to be anomalies in 11 voting precincts across the State. She also found that those areas used the Diebold Optical scanner machines for tabulating the vote. For those who have not followed the voter fraud allegations, proponents like Blackboxvoting.org claim Diebold is the lynchpin in their case for fraud, as it is an Ohio based Company that is owned by an outspoken Republican supporter (among other reasons). In short, every element existed to bolster the claim for fraud in this State (even though, in fact, Kerry still won). Ralph Nader with this information in hand then petitioned to have a hand recount in the 11 precincts that were identified as being statistically significant.

A hand recount was done, overseen by the statistician, the Nader camp, and a bi-partisan electoral staff. No material discrepancies were found after a hand recount of the ballots occurred. The statistician concluded that what appeared to be anomalous patterns, were just the realities of the voting behavior of those from New Hampshire.

If you believe the exit polls were correct, and fraud explains the discrepancy, you must also believe that GWBush stole around 73,000 votes in New Hampshire, but didn’t bother to steal extra 9275 votes to actually secure the win in the State. An absurdity.

Posted by: DBCooper | Jan 31, 2005 8:16:38 PM

Posted by: neal

I do not assert there was extensive fraud.

No, but you amply suggest it.

I guess you aren't supposed to be a sore winner, but there is evidence of fraud on the part of the democrats, from slashing van tires to giving cocaine instead of money for voter registration. I think this shows an endemic disease in the liberal party that will stop at nothing to achieve their ends.

Normally I wouldn't say this has anything to do with the democrat party, but as liberals pointed out, abuse commited by a few low level people in the Abu Ghraib torture scandal went "all the way to the top."
http://www.google.com/search?sourceid=mozclient&ie=utf-8&oe=utf-8&q=%22to+the+top%22+Abu+Ghraib+

Results 1 - 10 of about 43,400 for "to the top" Abu Ghraib . (0.25 seconds

Of course, I suspect that relationship doesn't interest you anywhere near as much as the poor quality of the exit polls.

What's also interesting to me is how the exit polls on a number of sites including CNN.COM suddenly changed as the results came in, after the polls were closed. That's kind of interesting too, isn't it?

Posted by: neal | Jan 31, 2005 8:20:06 PM

Posted by: D.A. Ridgely

I’ll repeat my acknowledged ignorance of statistics by way of prefacing the rest of my remarks, but I have to say that a 6% delta between polls and election results, even if it does correlate positively with type of voting machinery being used, doesn’t strike my lay mind as out of the realm of possibility of being a fluke. With all due respect for Mr. Velleman’s credentials, whenever someone tells me basically that the model can’t be wrong so there must be something wrong with the facts, I get a bit worried. Put differently, I just don’t have all that much confidence in the experts’ confidence about their expertise. Put a third way, I’d rather be ignorantly skeptical than ignorantly gullible.

As the natural scientists never tire of pointing out to the social scientists, people behave oddly, making both experimentation and confirmation of experimentation highly problematic. It’s one thing to say we’ve got the math down solidly enough to do sampling rates for analog-to-digital conversion for a CD, another to say we are entitled to that same degree of confidence with polling techniques with quirky human beings. And, contra Oliver, flukes happen all the time in the trivial sense of long shots paying off. The fact that the odds of winning the lottery may be 10 million to one doesn’t mean someone isn’t going to win. Perhaps the well has been poisoned to the point where more people suddenly do lie to pollsters. I always lie to them as a matter of principle, and yes I know they theoretically take folks like me into account, but who knows for sure that attitudes toward pollsters haven’t shifted faster than the pollsters have been able to adjust their models for? Maybe the polling data were corrupted by evil hackers. Maybe, as one rumor had it on election day, Democrats with inside information were planting themselves as polling subjects to try to start a bandwagon effect. I know, that’s absurd. But so are some of the other theories that have been floated around here. In fact, a facially absurd explanation might nonetheless turn out to be the right one.

Look, I have no serious explanation and I do find the disconnect between the polls and the election results at least prima facie suspicious. If there is credible evidence sufficient for probable cause to believe there may have been election fraud, it should be investigated. Is that what is being alleged? If so, is there any material evidence beyond the statistical abnormalities?

Posted by: D.A. Ridgely | Jan 31, 2005 8:33:00 PM

Posted by: pedro

D.A. Ridgely asks: "If so, is there any material evidence beyond the statistical abnormalities?"

I think this is an important question. I have to preface my own remarks by saying that I'm simply not familiar enough with the data to speak authoritatively on the matter. (I believe that skepticism is a healthy attitude when it begins at home, i.e., when one practices it with respect to one's own half-baked ideas.) But if there were evidence that the discrepancies occurred largely in precincts in which a particular type of technology was used, naturally I'd be worried about the possibility of some form of techie fraud, notwithstanding the assurances of experts in technology (guided by the very skepticism with which D.A. Ridgely sees the confidence of experts).

Because Mr. Ridgely's question is very important, I think it is crucial to have more knowledge about the ways in which the different technologies operate, and whether there are paper trails or other forms of reliable trails (not susceptible to smart hacking, like electronic voting might be).

Posted by: pedro | Jan 31, 2005 8:50:54 PM

Posted by: Tom

How about this as explanation 2a) Bias on the part of US Count Votes which analyzed the E/M data.

A cursory review of the sustaining members and contributors to US Count Votes revealed that in every single instance where I could find a match on hometown and name at opensecrets.org the member or contributor to US Count Votes was also a contributor to the Democratic Party or one or more of the Democratic Candidates in the primaries. There was one exception - a David Dodge who contributed to Lyndon LaRouche.

My examination was not exhaustive, however I found 8 matches in pretty short order.

My question to you Dr. Velleman is this - What are the chances that I would randomly check names on the US Count Votes contributor/member list and not find one contributor to the Republican Party? I'm betting close to 100%.

Posted by: Tom | Jan 31, 2005 8:57:47 PM

Posted by: oliver

For evidence, you might note the from the US Count Votes report this mention of
"a sworn affidavit by a Florida computer programmer who claims he was hired to develop a
voting program with a “back door” mechanism to undetectably alter vote tallies"

My Web surfing led me to this thing purporting to be such an affidavit:

http://www.buzzflash.com/alerts/04/12/images/CC_Affidavit_120604.pdf

and this is a related blog by the guy doing the purporting

http://www.bradblog.com/archives/00001047.htm

Posted by: oliver | Jan 31, 2005 9:00:57 PM

Posted by: Paul Deignan

DBCooper mentioned one audit. However, I don't think that paper receipts are the way to go (look at WA -- what a mess). They are susceptible to fraud as well.

One method would be biometric registration at the polls using fingerprint scanners. The voting information could be encoded with the fingerprint as a watermark and encrypted with PGP. Thus the voter could be able to securely verify his vote in the case of a recount.

This would also help with registration problems.

The essential elements of information:
1. Identification
2. Vote
3. Privacy Control
4. Verifiability
5. Authentication
are all here.

Posted by: Paul Deignan | Jan 31, 2005 9:02:49 PM

Posted by: pedro

Now that, Paul D., sounds like a good idea.

Posted by: pedro | Jan 31, 2005 9:05:07 PM

Posted by: Paul Deignan

Tom,

It is fine that the members of the organization are partisans. Partisan checking is a built-in feature of our system. What matters is the validity of the criticism.

(I am assuming that the partisan composition of the panel in this report was disclosed as partisan??)

Posted by: Paul Deignan | Jan 31, 2005 9:06:32 PM

Posted by: Paul Deignan

Pedro,

I am a bit shocked that we allow the system that we have. It is a disgrace and defeats one of the very important functions of voting -- verifiable trust.

Whether one is on the winning or losing side -- this is an abominable situation and is worth a couple billion to correct. ( some of which should find its way into the support of state of the art information theory :) )

Posted by: Paul Deignan | Jan 31, 2005 9:11:25 PM

Posted by: No Labels Please

I have a background in math, physics and statistics and can tell you that the US Count Votes paper has elementary problems of analysis. I would not trust any of the results based on their misunderstanding of interpreting the implications of the response rate by percent support of precinct for Bush as I outlined above. It is simply flat out scientifically wrong to assume it menas what they assum it means.

Again - I ask Dr Velleman to comment?

Secondly, why were the exit polls wrong? - who knows? But the exit polls are not the election - the election is the election. The natural assumption, absent other evidence, is that the exit polls are wrong - not the election. It is not the responsibility of the US government to establish it's innoncence in the face of a botched exit poll. And it isn't plausible to assume partisan malfeasance absent ANY evidence of tampering, the fact that the results corresponded closely to most pre-election polls, the fact that BOTH democrats and republicans run and monitor these elections, on and on and on...

Dr Velleman?

Thirdly, there are

Posted by: No Labels Please | Jan 31, 2005 9:14:06 PM

Posted by: oliver

D.A. Ridgely is being outrageous in sticking to his guns about flukes and in poo-pooing peer-reviewed findings by professors of statistics. We are at a stage of civilization at which we are weighing elementary particles with an accuracy out to nine decimal places or better. It's only thanks to science's and the world's faith in the mathematical methods of statistics that anybody believes there's proof for the standard model of physics, general relativity, dark energy and a zillion other hard to measure things. If 6% is dismissible prima facie, then so are rulings of the US Supreme Court.

Posted by: oliver | Jan 31, 2005 9:19:44 PM

Posted by: Tom

Paul Deignan,

Well, I have to agree with you that the validity of the criticism is what is important. But I must confess to being skeptical about the US Count Votes analysis for several reasons.

1) I trust that the candidates have better, more accurate polls than what is provided by E/M in election day exit polls. In no case did I hear anyone from the Kerry campaign say that their internal polls matched those of the exit polls. In fact, they seemed giddy that the exit polls had them in better shape than they expected. On the other hand, the exit polls didn't come close to matching the internal polls of the Bush campaign - at least according to Rove and VP Cheney.

2) As has been raised in the last 70 posts, there are plenty of other variables to look at that neither E/M or US Count Votes has yet to comment on.

3) If US Count Votes is another George Soros type organization trying to put an academic cover on political spin - why should they have any more credibility than, say, CBS News?

4) There are lies, damn lies, and statistics. Who knows what US Count Votes left out of their analysis? Who knows what E/M hasn't factored into theirs?

Posted by: Tom | Jan 31, 2005 9:23:19 PM

Posted by: oliver

Also, note there's no "model" or "theory" to this analysis and the polsters themselves consider the 6% bias a rock-solid fact that they feel obliged to explain. The problem is their explanation doesn't hold water.

Posted by: oliver | Jan 31, 2005 9:34:06 PM

Posted by: Tom

Paul D.,

You ask, "(I am assuming that the partisan composition of the panel in this report was disclosed as partisan??)

Well, here is the disclaimer at the end of the US Count Votes press release:

USCountVotes, a nonprofit, non-partisan Utah corporation was founded in December 2004. Its mission is to create and analyze a database containing precinct-level election results for the entire United States; to do a thorough mathematical analysis of the 2004 election results; and to fully investigate the 2004 Presidential election results. USCountVotes actively seeks volunteers and accepts donations to help make this unprecedented civic project a reality – visit www.uscountvotes.org for further information.

I question the non-partisan label that they have attached to themselves.

Posted by: Tom | Jan 31, 2005 9:34:42 PM

Posted by: Ben

Question--how did the E/M account for location? I'd assume they'd have to weight results from different precinct--how else could they get the right ratio of voters in Cleveland to voters in Akron to voters-everywhere-else?

To do that, I'd think, they'd have to guess the turnout in the various places. Which seems like a likely source of error. So is there some better way of doing it?

And Tom P.--If folks on your side do things the same way as on mine, you were only calling Bush supporters on election day. Most people vote in the afternoon, whatever their political stripe--if you'd called some Democrats, you would have gotten the same pattern. I've seen the same thing, working the phones for Democrats in NH.

Thanks, Paul, fo the post!

Posted by: Ben | Jan 31, 2005 9:34:44 PM

Posted by: D.A. Ridgely

Pedro is quite correct and I should have made it clear previously that I find the apparent correlation between type of voting technology used and the discrepancies to be the most troubling assertion. (I believe AlanC9 made the same point earlier.)

Whatever biases there may be on the part of those raising these questions, bias affects only the credibility of the speakers, not the truth of the matters asserted. Again, Pedro is on the money: we can spin theories forever but for now we appear to lack sufficient information on a number of issues.

My self-appointed nemesis, Oliver, chides me for being insufficiently deferential to peer-reviewed professors. I can’t quite fathom what Supreme Court decisions or, for that matter, quantum physics have to do with the case at hand, but I plead guilty as charged. He seems perhaps too easily outraged, however, as all I have done is admit ignorance and raise questions that, for all I know, might easily be answered. “Because I say so” isn’t much of an answer, though. “Because they say so,” isn't any better.

Posted by: D.A. Ridgely | Jan 31, 2005 10:30:37 PM

Posted by: Jim Hu

As Doug pointed out earlier in this thread, there is a lots of analysis on this at Mystery Pollster (Mark Blumenthal). Here are two that caught my eye:
on the Reluctant Bush Voter theory
on the inexperienced interviewers

Caveats. I haven't read the report because I really should be prepping for my class, and if I read 77 pages of anything tonight it will be about DNA replication, not exit polling. I am also pretty sure that I wouldn't have the expertise to get more out of it that what others are picking out above. While some of my work relies on measuring the masses of molecules to several decimal points, I consider myself to be only slightly above average at best in understanding statistics.

That said, the claim that sampling error cannot account for the discrepancy seems quite plausible to me. Having read what's out of Robert Caro's biographies of LBJ, I don't completely discount fraud, but I'd expect it to go both ways based on recent history...I would rank inaccuracy/fraud as only slightly more plausible than sampling error. Part of my tendency to discount fraud is that at least for the relatively early numbers, my recollection of election day is that the exit polls were favoring Kerry in states that not only went strongly for Bush in the end, but were also expected to go strongly for Bush.

I also think that selection bias is a plausible explanation without suggesting that there was bias on the part of the exit pollers themselves, or at least very much. Several of Mark Blumenthal's observations in the posts above, which have been suggested by others here as well, seem worth looking at.

1) Republican distaste for the pollsters, who were marked with logos for the MSM. But this should show up in the refusal rate, shouldn't it? Perhaps not.

2) Indirect avoidance effects. Blumenthal's Professor M reports that one of the exit pollsters he interviewed was constrained by election rules to stand near a MoveOn demonstration. I'm wondering how frequent this kind of activity was in precincts with exit polling.

3) Inaccurate/fraudulent reports from exit pollsters. The form of the inaccuracy Blumenthal suggests is not kids making up Kerry votes...it's inaccurate reporting of their refusal and miss rates, and/or not admitting to counting people who volunteered to be counted. This would affect detection of #1.

4) Attempts by the Kerry supporters to create a bandwagon.
a) Note that if someone wanted to inflate the Kerry responses and knew the sampling rules (e.g. ask the 3rd person), a group could simulate exiting from the polls in a large enough group to ensure that one of them would be counted. If the pollsters were enriched for first-time college kids, they might have inadvertantly leaked this info to their Kerry supporter friends.
b) Moles in the exit poll workforce could have doped the data toward Kerry.
c) "One student reported at least one instance of a person simply taking a survey from her supplies (which were out in the open at her table) filling it out, and dropping it into the survey box."

I'm not sure why the Kerry folks would do #4, and it might not have been coordinated. But remember that some Dems made a concerted efforts to tilt the MSM post-debate internet polls (I hate these, btw...why do we care about the opinions of Perl scripts?), and had coordinated letter to the editor spam attacks worthy of Brent Bosell's group complaining about TV indecency to the FCC. I suspect that the likelihood of these things happening in November but not so much in earlier elections is related to the heatedness and extra-crispy partisanship we saw this year. The sense among Dems that they wuz robbed in 2000 may have led some of them to rationalize these behaviors, along with the tire slashing, sign defacing, and window breaking in the run up to election day.

Posted by: Jim Hu | Jan 31, 2005 10:47:38 PM

Posted by: oliver

D.A. Ridgely writes: "My self-appointed nemesis, Oliver, chides me for being insufficiently deferential to peer-reviewed professors."

Not quite. The deference I considered to be due was to basic principles of statistics as applied under the oversight of peers by a team of practitioners with the highest credentials achievable in this mathematical field. The peer review and surfeit of loftily credentialed participants in this study are an extreme instance of lily gilding, to my mind, which I see as an attempt to avoid easy dismissal by anyone with any regard for or sense of the intellectual institutions of this country. I may have made a mistake in assuming that Mr. Ridgely would read the report, as I did, as representing a straightforward application of relatively basic statistical methods, in contrast to a speculative attempt to advance the scientific cutting edge. To me it seems one would have to assume bad faith on the part of the authors to have read it as anything other than a basic exercise, given how the authors represent what they have done. So if "sufficient deference" to a person or group means taking at least some pains to avoid concluding they are willfully misleading you, then I am indeed tempted to conclude that Mr. Ridgely is being insufficiently deferent here. Out of my own deference to him (after all, he's my personal nemesis), I will not jump to this conclusion and admit the possibility that he's simply being foolhardy.

Posted by: oliver | Jan 31, 2005 11:57:51 PM

Posted by: CDC

D.A. Ridgely: "...I find the apparent correlation between type of voting technology used and the discrepancies to be the most troubling assertion..."

If I understand the analysis correctly (and I probably do) neither the primary data nor the analysis (in the form of an Anova table and the pairwise comparisons) are available. We don't know if the differences between the classes of voting proceedures are significant. Until we know that, all the rest of this is so much thrashing.

Pedro (to Paul Deignan): "I find it difficult to follow you sometimes."

Perhaps this link will help:
http://www.ottosell.de/pynchon/corey.htm

Posted by: CDC | Feb 1, 2005 12:11:26 AM

Posted by: DBCooper

I find a few things fishy with the US Count Votes analysis of the E/M data report. The primary problem I see with the study is the stated belief of the authors in the historical reliability of exit polls. They make the statement “Exit polling is a well-developed science, informed by half a century of experience and continually improving methodology3.” The footnote reference leads to Steven F. Freeman’s “The Unexplained Exit Poll Discrepancy” study. Freeman is also an under signor of the E/M analysis. Talk about a circular reference.

In Freeman’s study, he provides the most threadbare evidence to support the historical reliability of exit polls. He relies on German and Utah exit poll data to prove his assertion, (two rather distinct monolithic groups, I would say) rather than making the more obvious and relevant comparison against exit polls in previous nationwide elections in a heterogeneous society. In fact, if exit polls have been so reliable, why was VNS disbanded after the 2000 election?

Also in the US Count Votes analysis the following claim is made. “The Edison/Mitofsky report confirms there were large differences between their exit polls and the official results of the 2004 presidential election –