I await the debunking — this is too cute to be easy to believe:
Clear Evidence 2006 Congressional Elections Hacked: A major undercount of Democratic votes and an overcount of Republican votes in U.S. House and Senate races across the country is indicated by an analysis of national exit polling data, by the Election Defense Alliance (EDA), a national election integrity organization.
These findings have led EDA to issue an urgent call for further investigation into the 2006 election results and a moratorium on deployment of all electronic election equipment.
“We see evidence of pervasive fraud, but apparently calibrated to political conditions existing before recent developments shifted the political landscape,” said attorney Jonathan Simon, co-founder of Election Defense Alliance, “so 'the fix' turned out not to be sufficient for the actual circumstances.” Explained Simon, “When you set out to rig an election, you want to do just enough to win. The greater the shift from expectations, (from exit polling, pre-election polling, demographics) the greater the risk of exposure—of provoking investigation. What was plenty to win on October 1 fell short on November 7.
Also, I've never heard of the Election Defense Alliance, nor of its leaders (although some of the bios are intriguing). Perhaps someone could evaluate their report?
The point of an election is to enable the peaceful transfer of power by convincing the losers that they lost.
The contents of the report are frankly irrelevant, the fact that the perception is widespread that electronic elections are rigged should condemn the system.
I read the report, and the results are pretty disturbing…. so I did a little of my own research — I picked a state at random (New Jersey) and looked at the numbers. And while its “possible” to explain away the “lower percentage of Kerry voters came out” stuff, there are other aspects that are just not credible…
For instance, supposedly this year 4% of the voters were people who voted for “other candidates (neither Kerry nor Bush) in 2004 …and 80% of them voted for either Menendez or Kean this year.
This year, there were 2,187,989 votes cast for the senate in New Jersey.
In 2004, a total of 30255 Jerseyans voted for “other candidates” — only 0.837% of voters.
4% of 2,187,989 is 87519… or 290% of the actual number of voters who voted for “other candidates” in 2004.
That’s not even close to possible….. when you add in the fact that 50,083 people (2.29% of votes cast) voted for “other candidates” this year, well, its extremely difficult to believe that while 80% of the people who were voting for the “socialist worker” or “libertarian”, or various other fringe parties in 2004 decided to vote for “mainstream candidates” this year, 44,032 Jerseyans who voted for Bush or Kerry went to a fringe candidate this year.
******
I’ll pick another state….Virginia.
According to the exit polls, In Virginia, 3% of the votes cast in the senate race this year voted for “other candidates” (not Bush or Kerry) in 2004. In 2004, there were 26,666 votes for “other candidates”.
In 2006, there were 2,370,445 votes cast in the Senate race. 3% of that is 71,113 — or 267% of the number of voters who voted for “other candidates” in 2004.
*********
Montana
According to the exit polls, In Montana, 6% of the votes cast in the senate race this year voted for “other candidates” (not Bush or Kerry) in 2004. In 2004, there were 10,661 votes for “other candidates”.
In 2006, there were 405,025 votes cast in the Senate race. 6% of that is 24,301 — or 228% of the number of voters who voted for “other candidates” in 2004.
(and according to the exit polls, 7% of the people who voted in the senate race this year did not vote for President in 2004 — that’s an AMAZINGLY high number….)
****************
In sum, the “adjusted” exit poll numbers aren’t just “highly unlikely” to reflect what actually happened, they’re completely impossible.
(I tried looking at “non-competitive” senate races to see if the trend was the same…and of the fiveI looked at at random (California, Massachusetts, Maine, Nevada, and Michigan) none of them had data on “who did you vote for in 2004?”)
Paul Lukasiak: I read the original report and your comment. Two questions:
First, these people saying they voted for “other candidates” in 2004 — did the exit poll question really ask only for “other candidates” or was this a broader bucket for “other candidate/did not vote in 2004/decline to state who they voted for in 2004”? If the latter,then perhaps the numbers aren’t so out of whack.
Second, how possible is it that some people voting in an election in 2006 will lie to a pollster about who they voted for in 2004? Certainly some Bush voters have to be embarrassed by their vote, given subsequent developments. But perhaps some Kerry voters as well.
Could it be that the discrepancies you are seeing can be explained by the two items above?
As a follow on to my last comment: let me be clear that I consider the likelihood very, very high that there was some hacking of the election machines this year and in the past two Federal elections. But I also want to make sure that any purported evidence of such hacking stand up to thorough peer review.
My rationale is simple: As someone who spent many years in the computer security industry in the 1980s and 90s I am in agreement with the experts about just how vulnerable these machines are to undetected vote fraud. So, technically this would simply not be much of a challenge. Then, consider the extensive electoral cheating for which we have confirmation, which shows that there are people who are willing to break the rules to influence the result. Put these two facts together and frankly it would be a shock if there wasn’t large scale tampering of the computer voting machines.
Goose… there was a separate category for “did not vote in the last election”. There was no “declined to answer category.”
On the question of “declined to answer”…. I just checked, and it appears that all the percentages in the “who did you vote for last election” add up to 100, so there is no telling where “declined to answer” wound up — that being said, I doubt that they would include “declined to answer” in the “other” category; my guess is that they just excluded those who “declined to answer” from the results (“declined to answer” appears in no other question as an option, yet the percentages all up to 100 ..or close enough that “rounding error” would explain why they add up to 99 or 101.)
And while its true that people might lie… the assumption here is that they lied about voting for Bush…. which would mean that the turnout of “Bush voters” would be even higher (and thus even more at odds with the “raw data” and what anecdotal evidence and common sense told us about voter behavior this year.
I looked at all the exit poll data, and the table below shows the results of all the state where the question was asked. the first column is the state, the second is the exit poll percentage for “other” voters, the third is the number of actual “other” voters, the fourth is the number of votes for Senator this year (data from state election web sites), the fifth is the number of “other” voters that would have had to vote this year for the 2nd column to be correct, and the final one is column 5 divided by column 3, expressed as a percentage.
Ohio 2% 26973 3826829 76537 284%
Minnesota 3% 36678 2202772 66083 180%
Missouri 3% 16480 2114954 63449 385%
Pennsylva 3% 33822 3997568 119927 355%
Tennessee 3% 16467 1826310 54789 333%
Virginia 3% 26666 2370445 71113 267%
New Jersey4% 30255 2187989 87520 289%
Montana 6% 10661 405025 24302 228%
(apologies for the formatting, which is practically impossible to do in this comment section)
needless to say, these numbers need SOME explaining….
PL: thanks for the answer. So, what we’re left with then is that the only possible explanations are:
1) The exit polls were flawed in some way such that the voter responses overstated Bush’s vote in 2004 while understating the Republican vote in 2006.
2) The actual vote counts were wrong.
3) Some combination of the above two.
Correct?
goose…
you forgot the most obvious answer…. the elections were hacked.
If I read the study cited by Michael correctly, the discrepancies between the exit polls and the actual results occurred only in the close races — THAT strongly suggests to me that hacking did occur. The data that I posted is basically evidence of how extreme the “forcing” of the exit poll data had to be in order to make it “fit” with the results in those close races.
Does anyone find it odd how reluctant Republicans are to challenge vote totals in close elections where they lost? This seems so out of character that I wonder if they are afraid of closer scrutiny. The Allen-Webb race is an example. Nothing in the previous behavior of George Allen, the state Republican party, or the national party would suggest that they would not challenge those results. And yet all we hear now is that they don’t want to create any more divisiveness. Hello?
you forgot the most obvious answer…. the elections were hacked.
That’s what I meant by #2 — the vote counts were wrong. And you’re right, if the discrepancies only happened in races expected to be close, that adds weight to the hypothesis.
The report you link to says the following:
“For the purposes of this analysis our primary attention is directed to the exit poll in which respondents were asked for whom they cast their vote for the House of Representatives. 2 Although only a few House races were polled as individual races, an additional nationwide sample of more than 10,000 voters was drawn,3 the results representing the aggregate vote for the House in E2006. The sample was weighted according to a variety of demographics and had a margin of error of +/- 1%.”
The last sentance, “weighted according to a variety of demographics” is not footnoted. The reader cannot know what this means, nor how the weighting might have influenced the data.
Given that our nation is basically divided both between and within demographic groups, it is probably incorrect to weight demographics. But even if it is, then how was it done? I am consistently annoyed when such studies do not publish the confidence interval belonging to the 1%…was it 95% or 90%? Given that the report seems to call for revolution over a 3% discrepency, its important to know how sure they are.
Much as I agree voting machines are insecure and much as I’d like to see more secure elections, I’m not confident that the discrepancy to the 2004 Bush/Kerry results is relevant here. My impression is that it is conventional wisdom that the electorate tends to skew more Republican in midterm years – older voters (who skew Republican) are more likely to vote than younger voters (who skew Democratic). Now 2006 may be an exception to this, but without additional proof (that I didn’t see in the article), I don’t see why a slightly larger margin for Bush in the 2006 electorate than in the 2004 results is cause for alarm.
Expecting exit polls to be accurate at the 4% level is not realistic. I don’t think it is at all unlikely that voters in New Jersey would vote Libertarian rather than Republican. Given the attack adds claiming that Menedez is a gangster a sudden surge in the independent/other vote is to be expected.
You don’t rig an election by rigging the count, you rig an election the way Katherine Harris rigged the Florida election in 2000, you stop legitimate voters from voting. The purging of the electoral rolls cleverly designed to ensure that blacks were disenfranchised disproportionately was standard voter suppression tactics.
“Expecting exit polls to be accurate at the 4% level is not realistic. I don’t think it is at all unlikely that voters in New Jersey would vote Libertarian rather than Republican. Given the attack adds claiming that Menedez is a gangster a sudden surge in the independent/other vote is to be expected.”
first of all, exit polls do tend to be accurate….when you are talking about samples of 10,000 people, its unsurprising that you get a margin of error of +/-1 and a confidence level of 95. People have been doing exit polling for years, and have it down to a science. IF an exit poll is off, there has to be an explanation (exit polls were “off” in Florida in 2000 because there was an was a bizarrely high number of undervotes and overvotes in areas where Gore was expected to do well. Exit polls reflect voter intent — tens of thousands more Florida voters intended to vote for Bush than Gore, but never had their votes counted. In other words, the Florida 2000 exit polls were a far better indicator of voter intent than the outcome was.)
and the data I put together had nothing to do with the number of people who voted “libertarian” in NJ this year — its about the number of people who — according to the exit poll data, claim to have voted “libertarian” (and other fringe parties) in 2004. It compares the actual number of people who DID vote for fringe parties in 2004, and compares it to the number of people who, according to the exit poll, voted for fringe party candidates in 2004.
One of the interesting things about the data is that the discrepancies seem to get bigger the larger the claimed “other” vote is, which argues against the idea that “other” includes “refused to answer ” or “did not recall” (RTA/DNR). In other words, if RTA/DNR are included as part of “other”, then in Connecticut, where 2% were listed as “other”, 1.53% of the voters would have been RTA/DNR. In the five states where 3% were listed as “other”, 2.20% of the voters would have been RAT/DNR. In New Jersey, where 4% were “others”, 3.16% would have to be RTA/DNR. And in Montana, where 6% were RTA, 3.66% would have to be RTA/DNR. (all of these numbers assume 100% turnout of the “other” voters from 2004 in 2006.)
This would suggest that there is a correlation between unco-operative and forgetful voters, and voters who vote for “other” candidates. There is no reason to suspect that there is any correlation in the general population between forgetfulness and voting for “other candidates” — and (based purely on “anecdotal” evidence, people who vote for “other” candidates are not especially shy about the fact that they aren’t responsible for the current mess in Washington.
In other words, there is no explanation for these numbers….other than CNN “forcing” the data to fit the results, regardless of whether their numbers actually added up.
So is this the same Lukasiak who was defending the Dan Rather forgeries to the last gasping breath?