After the U.S. presidential election on November 2, 2004, some pro-Kerry sources have made allegations that data irregularities and systematic flaws occurred during the election. The overall official result of the election is not at this time being challenged by the U.S. Senator John Kerry, the Democratic candidate, who is the only possible winner if the unofficial results change. This is however not relevant to the US electoral process in which the Electoral College not the candidate has the final say, and which may 'draft' him even unwillingly to serve unless he actually refuses, in which case they may select John Edwards as President.
At the moment any change in the result of the election highly unlikly due to that fact that Bush's margin of victory was greater than the number of alleged fraudulent votes. However, some people and groups (including the media, independent candidate Ralph Nader, Kerry's brother and legal advisor Cameron Kerry, members of the House Judiciary Committee, and many Democratic groups) are currently analyzing the available data.
Dan Hoffheimer, the statewide counsel for the Kerry campaign, has said the Kerry campaign is not trying to challenge the election. "We're not expecting to change the outcome of the election," Hoffheimer said.
No comprehensive analyses have yet been produced, but there is a large volume of both primary and secondary data, and preliminary analyses, reports and observations have been made by a variety of commentators ranging from computer scientists to voting rights organizations, and many others. One preliminary attempt to analyze the issue from Caltech concluded that "there is no evidence, based on exit polls, that electronic voting machines were used to steal the 2004 election for President Bush." However, this analysis used exit polls that had been weighted by the final vote count, thereby assuming the conclusion.
One part of the controversy are electronic and optical-scan voting machines, which were used in greater numbers than before as a result of concerns over the reliability of manual machines raised during the 2000 election. Other reported problems relate to abnormally high voter turnout (more votes in many precincts than registered voters in said precincts), discrepancies between exit poll data and actual results especially in swing states and the complications which arose due to long lines; particularly in high-population areas and in closely contested states.
In many cases there were concerns as to whether votes were fairly, reliably, and accurately recorded and reported by the electronic machines involved. The charts below demonstrate this.
In 2003 Wally O'Dell CEO of Diebold Election Systems' parent company said in a letter to Ohio Republican officials that he was committed, "to helping Ohio deliver its electoral votes to the President". [1] Diebold is the company which makes electronic touch screen voting machines used in Ohio and other states, most significantly Florida. Ohio and Florida were two of the "swing" states critical to the 2004 election.
However, it should be noted that according to one website "Diebold’s election-systems division is run by a registered Democrat" and Mark Radke (Director of Marketing for Diebold Election Systems) "has an exclusively Democratic donation history"..."including the legal limit of $2000 to John Kerry in the recent campaign".
Chuck Hagel, the previous chairman of ES&S, another major manufacturer of voting machines and still a $1m stock-holder in McCarthy & Co which owns a quarter of ES&S, became a Republican candidate. Hagel's Democratic opponent made a formal protest to the state of Nebraska over the conflict of interest.
In at least one case it appears a voting machine was hacked during a primary election in King County Washington and a warning was issued to disconnect all voting machines from the internet. But this would not prevent the effects of hacking totally [2].
Note: As with all statistics, it is very important to consider other causes of apparent anomalies, and to provide verifiable and neutral source data that can be checked in a neutral way by third parties. All the information and sources below appear prima facie to be statistically reasonable in terms of both analysis and assumptions, and to be based upon verifiable public data.
(1) An analysis of Florida counties with 80,000 - 500,000 registered voters concluded (with a few caveats of a usual kind) that machine type (E-Touch vs Op-Scan) was a "significant predictor" of vote at the p < 0.001 level (less than one chance in a thousand of this degree of anomaly happening by chance) [5][6][7]Source data and calculations [8].
(2) One thread on the "democraticunderground" website discusses Gahanna, Franklin Co. Ohio. The vote reported by the county in Gahanna precinct 1-B was 4,258 Bush, 260 Kerry, and the total votes cast in Gahanna overall were 20,736. However:
Source: [9], source data from govt website pdf
(3) An analysis reported in the New Zealand press looks at the differences between exit polls and reported voting in more detail. It identifies that in a selection of non-swing states, the exit polls and final results match. However in a large proportion of what were identified before the election as key swing states (Wisconsin, Pennysylvania, Ohio, Florida, New Hampshire, etc.), the exit polls and final votes do not match.
The error was in each case a statistically anomalous and electorally critical 4 - 15% swing (change between exit polls and electronic voting) and furthermore the anomalies were not random. In each of the above swing states, this variation between what voters said they voted and what the machines reported was in favour of Mr. Bush. Source [10], article discussing here, graphs here.
(4) An interesting article comments that:
(5) There were additional reports of significantly large data irregularities with the "optical scan" type voting machines in at least Florida. In one county using optical scan voting machines for example, election records showed 77% registered democrats but Bush received 77% of the vote.
(6) Wired Newshas examined this issue and reports that, "...according to academics, the internet pundits are reading the data out of context. Demographic figures and vote trends over several years show the numbers to be consistent with previous elections. According to University of California at Berkeley political scientist Henry Brady, the Republican vote share has been going up in Florida's rural optical-scan counties for years."
Wired further reports that, "[t]hree professors of government also examined the numbers after being pressured by many people, including lawyers for the Democratic Party, and concluded the same thing."
(1) Testimony of Dr. Aviel D. Rubin to U.S. Federal Election Assistance Commission, on Electronic Voting Systems, May 2004:
(Witness credentials: Professor of Computer Science, Technical Director of the Information Security Institute at Johns Hopkins University, served on SERVE security peer review group for Dept. of Defense, member of National Committee on Voting Integrity, and 2004 election judge in local county)
Voting fraud is also both possible and hard to prove with some versions of electronic voting machines.
See http://www.ustogether.org/election04/florida_vote_patt.htm
Exit polls have been used successfully in other countries to determine election fraud.[24] Because final published exit polls in America are matched to vote counts, they cannot be used to determine election fraud. However, in the 2004 election, pre-matched exit polls for 17 states were leaked onto the internet.
Voting locations that used electronic or other types of voting machines that did not issue a paper receipt or offer auditability correlate geographically with areas that had discrepancies in Bush's favor between exit poll numbers and actual results. Exit polling data in these areas show significantly higher support for Kerry than actual results (potentially outside the margin of error). From a statistical perspective, this may be indicative of vote rigging, because the likelihood of this happening by chance is extremely low. A study of 16 states by a former MIT mathematics professor places the likelihood at 1 in 50,000. [25]
File:2004 us popular vote2.gif
Supporting the same conclusions of the maps above, here are bar graphs indicating the differentials between Exit Polls and Machine Tallies for nine e-voting and paper ballot states. The discrepancies appear to affect the e-voting states to a significantly greater degree than they affect the Paper Ballot states.
File:Exit poll small.jpg Source and background discussion are listed here: http://www.bradblog.com/archives/00000893.htm
Source data and analysis: http://ustogether.org/Florida_Election.htm http://ustogether.org/election04/mitteldorf/Liddle.htm
Corroborating Data and Analysis: [26]
The following tables compare final exit poll data with penultimate exit poll data, note the large swing of support towards Bush, with Kerry losing votes, which is impossible if votes are only being added. National Election Pool, the consortium which conducts exit polls, has stated that the early data was inaccurate due to regulations preventing pollsters from approaching voters, barriers, (neither of which would skew the data) and the alleged perception that Democrats are more willing to answer exit polls. The consortium dismissed the possibility that their early exit poll was accurate and that vote counts were wrong, but did not provide any reasoning for this assessment. The early exit poll data was not meant to be released to the public. The data that was meant to be released to the public was intended to be weighted by the actual vote count. Exit polling companies claim this is standard procedure. Critics argue exit poll data should never be weighted by final results and have requested access to the raw data.
Direct link to screenshots and data: CNN website 12.21am CNN website 1.41am
CNN screenshot #1:
12.21 am, 1963 respondents so far
Total vote: Male 47% , Female 53% of which:
Male - Bush 47% x 49% x 1963 452 Male - Kerry 47% x 51% x 1963 471 Female - Bush 53% x 47% x 1963 489 Female - Kerry 53% x 53% x 1963 551 TOTAL - Bush 941 TOTAL - Kerry 1022 (rounding: estimates of voters in each category accurate within +/- 10)
CNN screenshot #2:
1.41 am, 2020 respondents so far (57 more than above)
Total vote: Male 47% , Female 53% of which:
Male - Bush 47% x 52% x 2020 499 Male - Kerry 47% x 47% x 2020 451 Female - Bush 53% x 50% x 2020 535 Female - Kerry 53% x 50% x 2020 535 TOTAL - Bush 1034 TOTAL - Kerry 986 (rounding: estimates of voters in each category accurate within +/- 10)
The addition of an extra 57 voters at this station was therefore reported as +93 votes for Bush by AP and CNN at least, and voters monitoring the exit polls were told authoritatively that Bush had now taken a lead from Kerry.
Note that the counts for Kerry under Male voters changed in a negative direction after additional voters were included. The net subtraction of 20 votes from the Kerry total after adding new voters seems to reflect an adjustment process.
Long lines, though seemingly benign - "a mere inconvenience" - may well be the most serious problem with the 2004 election. In many places, lines were over 6 hours long.
Prior to the election, there was much ado about each precinct getting enough ballots, but an equally serious matter that seems to have been overlooked by people trying to protect people's right to vote is whether the precints had a sufficient number of voting machines, such that the votes could be proccessed at a sufficient rate. Machine quantity as well as ballot quantity determines the saturation point of votes. Number of machines * Max. votes per hour per machine * hours poll is open = max. number of votes precinct is able to process. Every voter over this limit is effectively disenfranchised, just as if the precint had run out of ballots; the precinct runs out of voter-time-slots.
Although low population precincts had relatively plenty of voting machines and were well within the limits of processing capacity, high-population centers often did not, and sometimes had less than half the machines requested and were well outside the limits of processing capacity, effectively disenfranchisng an undetermined number of voters.
This may explain the discrepancy between expected voter turnout in high-population areas and counted voter turnout in these areas. Since high-population areas are predominantely Democratic, this would primarily effect the Democratic constituency, and appear on the surface to reflect inefficacy in the Democratic GOTV effort.
841 incidents of this type have been reported, 241 of which are from Ohio, and 106 of which are from Florida. 124 such incidents have been reported out of Cuyahoga county, Ohio.
Specific concerns were raised in the course of the election in respect of votes from key minorities, such as Blacks [27] or Cuban Hispanics.
On November 5, Ralph Nader filed a request for a recount of the votes in New Hampshire with that state's Secretary of State. Nader's request cited "irregularities in the vote reported on the AccuVote Diebold Machines in comparison to exit polls and trends in voting in New Hampshire" and added: "These irregularities favor President George W. Bush by 5 percent to 15 percent over what was expected." [28] As one of the candidates on the ballot, Nader has the right to demand a recount, but is required to pay for it (because he lost by more than 1 percent of the vote). The state Attorney General's office has responded that Nader's request was not valid because no check for the expenses was submitted by the deadline. [29]
As of yet, neither major political party has made an official response to the issue.
Several Democratic members of the House Committee on the Judiciary have written to the GAO requesting a formal investigation. Their first letter was written three days after the election, on November 5 [30], and this was followed by a second letter on November 8 listing further matters which had since come to light [31].
House Committee website for later information
Glitch gave Bush extra votes in Ohio http://www.cnn.com/2004/ALLPOLITICS/11/05/voting.problems.ap/
An article by Salon.com purporting to debunk many of the data irregularities. http://www.salon.com/tech/feature/2004/11/10/voting/index_np.html
A wired news article purporting to debunk many of the data irregularities. http://www.wired.com/news/evote/0,2645,65665,00.html
This article written after the election summarises media viewpoints from many countries round the world.