NOTE: Site was entirely re-written and updated on 11/14/04. Subsequent updates appear in RED. Most recent update: 11/19/04.

Did Bush Lose... Again?

Why did Florida voters with one type of machine vote one way,
and voters with a different machine vote another?

Why were the exit polls telling a different story?

Why do all the reports of problems favor the same candidate?

On this page, we try to clearly summarize some of the suspicious circumstances of this election. We also try to show both sides of the issues, and point you to where you can find more information. We do not take the position that fraud has taken place; only that there seems to be sufficient cause for further investigation.

voter fraud Ohio voting machine

Why did Florida voters with one type of machine vote one way,
and voters with a different machine vote another?
Can this really be explained simply by the "Dixiecrat" phenomenon?

There are a series of arguments and counter-arguments for this one, so let's start at the beginning...

In Florida, the figures show that, in counties with one type of voting machine, voters with no Democrat or Republican party affiliation appeared to split their votes roughly 50/50 between Bush and Kerry, which was to be expected; yet in counties with another type of voting machine, unaffiliated voters seemed to vote nearly 100% for Bush!

Click HERE for a table that shows, county by county, how many registered Republicans there were, how many registered Democrats there were, and what percentage of registered voters turned out to vote. From those figures, they arrive at an "expected" minimum vote for each candidate. However, since many more people vote besides those who are registered Democrats or Republicans (i.e. Independents, unaffiliated, or members of some other party), the actual vote for each candidate is naturally going to be higher than the "expected" figure which is a projection based only on registered Democrats and Republicans. The table also then shows the actual vote, as of 98.6% of the vote tallied.

Some Florida counties used an "E-Touch" voting machine, and some used an "optical scan" machine. Let's look at these two groups:

"E-Touch" Voters


Approx. 3.86 million total voters
in these counties

Kerry's Base: about 1.57 million votes*

Bush's Base: about 1.44 million votes*

Kerry's final tally: about 1.98 million votes
26.5% more than his given base

Bush's final tally: about 1.85 million votes
28.6% more than his given base

Close race, as expected,
unaffiliated voters appear nearly evenly split
between the two candidates

* - based on the number of registered Democrats or Republicans,
adjusted for turnout

"Optical Scan" Voters


Approx. 3.42 million total voters
in these counties

Kerry's Base: about 1.43 million votes*

Bush's Base: about 1.34 million votes*

Kerry's final tally: about 1.45 million votes
Less than 1% more than his given base

Bush's final tally: about 1.95 million votes
45.8% more than his given base

Virtually every unaffilated voter
appears to have gone for Bush

* - based on the number of registered Democrats or Republicans,
adjusted for turnout

These two Florida populations who seem to have voted so differently are roughly equivalent in size and relative party strength. Yet the counties with the "optical scan" machines went disproportionately, overwhelmingly, and seemingly illogically for Bush. What could account for the difference in results? Did the machine itself make the difference, or is there another explanation?


• Geographic distribution of machines
E-Vote machines tended to be used in larger/urban ares whereas optical scan machines tended to be used in smaller/rural areas, so without a random distribution of the type of machine used, the discrepancy could be based on the different perspectives of those two populations rather than on the machine itself. And that makes sense, considering the "Dixiecrat" phenomenon which we'll discuss shortly. However, it is also worth noting that further analysis HERE showed that, in mid-size counties (some of which used each type of machine), there remained a significant correlation between unusual strength for Bush and the use of optical scan machines even when comparing counties of similar population size.

• Historical trends
It has also been pointed out that the rural counties have been trending more heavily Republican for some time, out of proportion to what would be expected from party affiliation. That data from the last few presidential elections has been posted HERE. While past results do provide additional support for believing this year's numbers, there continue to be arguments on the other side, discussed below. In addition, if there is, by some chance, a flaw in these machines leading to a higher result for the Republican candidate, can we be sure that this same flaw did not exist in prior recent elections?

• Dixiecrats
Regardless of whether or not you accept the exact numbers from the historical data discussed in the previous paragraph, it does seem clear that there is a regional rural phenomenon that many registered Democrats do indeed tend to vote Republican. And that certainly provides some explanation, as the estimated base figures discussed above were based on an assumption that, while there are crossover voters in each party, they would probably balance themselves out in about equal numbers. So does this mean we have a fully suitable explanation? Not quite. According to a CNN poll, while 14% of Florida Democrats did indeed vote for Bush, 7% of Florida Republicans in turn voted for Kerry. The net difference is not nearly enough to account for the discrepancy. It wouldn't even have been enough if there were no Republican cross voters, or if all the cross-voters were in counties with one kind of machine, and none in counties with the other. However, on the other side, one reader suggests that, since the CNN data is based on an exit poll, perhaps there were registered Democrats who voted for Bush who did not identify themselves as such when polled. However, while exit polls may be flawed for a variety of reasons (including reasons like that), these polls would have to have been off by a tremendous amount to yield the results necessary to use the Dixiecrat phenomenon as a full explanation for the final results. And even bad exit polls aren't that bad.

• More counter-argument for the Dixiecrat explanation, and a nod toward Ohio
One might assume that counties that went seemingly disproportionately for Bush (where registered Democrats arguably tend to vote Republican as a matter of course) would show, not just surprising strength for Bush, but also comparable support for the Republican Senate candidate. However, there were numerous counties where Bush's support seemed substantially greater than that of the Republican Senate candidate--and, as it happens, that phenomenon also correlated with use of optical scan voting. That information is HERE. That site also has some interesting charts, breaking down the vote by specific voting machine model, and showing the trend according to the size of the population of the county. Also, the same site shows that in Ohio, Bush also did best where optical scan machines were used. There is not enough data available from Ohio to see what other variables may correspond with the use of optical scan machines there.

• Machines from different manufacturers
The table this section is based on also shows the manufacturer of the voting machine used in each county. The machines in the suspect counties were not all made by the same company. Mostly they were made either by Diebold or ES&S. The fact that the results tended to skew regardless of which company's machine was used argues against an inherent bias related to the machines themselves. However, as it happens the two companies share some "corporate DNA," so to speak. You can read about that HERE and HERE (scroll down about halfway, to the section titles "Incestuous relationships"). There was one county that used machines from a third vendor, Sequoia. One individual county is not enough to support or refute the presence of a pattern, but for what it's worth, that county also did exhibit an unusually strong swing toward Bush.

• Turnout of the base
The estimated base figures in the shaded areas above assumed that about the same percentage of each party's members would actually show up and vote. Were Republicans far more effective in getting out their base than were Democrats? In fact, it looks like that's true. CNN reported that Republicans made up 41% of the voters in this election, Democrats 37%. Since the table referenced earlier shows that Florida actually has more registered Democrats than Republicans, it is apparent that a greater percentage of Republican voters showed up at the polls than did registered Democrats. If members of the two parties had shown up in about the proportion of their party enrollment, you would expect the Democrats' percentage of the total to be about 4 points higher than that of the Republicans, rather than 4 points lower. However, this does not seem sufficient to explain the figures above, even if all the higher turnout was concentrated in the counties with the optical scan machine, and even when combined with the percentage of Democrat crossover votes also reported.

Other Facts About Florida That May Make One Wonder...

• Officials are trying to hide something
The people from have gone into Florida to inspect records they were legally entitled to inspect. What the ran into is startling, and too much to explain here, but you can read all about it HERE or check their site and look for the entry marked "TUESDAY NOV 16 2004: Volusia County on lockdown" for a blow by blow.

Another study shows that the E-Vote machines may be flawed as well
In an interesting turnabout, another study showed potential problems with the purely electronic (e-vote) machines! As described above, many people dispute the validity of the optical scan Bush bias by saying that the votes in those counties are historically consistent. Well, a study at Berkeley found that the number of votes received in the
other counties (while stronger for Kerry than were the optical scan counties) were, themselves, far stronger for Bush than could be justified by historical voting patterns. So those who dismiss the theoretical optical scan problem based on historical data, may find themselves troubled by these findings, as the same data applied to these counties appears to argue the opposite point.

The Berkeley study concluded "Irregularities associated with electronic voting machines may have awarded 130,000 excess votes or more to President George W. Bush in Florida.. Compared to counties with paper ballots, counties with electronic voting machines were significantly more likely to show increases in support for President Bush between 2000 and 2004. This effect cannot be explained by differences between counties in income, number of voters, change in voter turnout, or size of Hispanic/Latino population.We can be 99.9% sure that these effects are not attributable to chance." They dismiss the Dixiecrat factor by pointing out that some of the biggest, most suspicious gains were in decidely non-Dixiecrat territory.

As an added point, they also mention that, while they statistically predict that there were 130k excess Bush votes, they do not know where they came from. That is, they may not be "extra" votes, but rather they may be "switched" votes. Which means it is conceivable that, not only should Bush have 130k fewer votes, but that Kerry should have 130k more, meaning a total 260k swing between them. (If it happens that both voting systems were compromised, you could further augment that 260k with whatever excess votes may have been generated by the optical scan machines, for a still larger number.) However, as my understanding is that these machines do not have a paper trail, no manual recount is possible, and I don't know that there is any way to confirm or refute the results of these machines. So this argument is probably one that can never be empirically proven either way, and will probably remain the purview of people who debate one statistical scenario over another. You can find more coverage HERE and all the documentation HERE.

• Exit polls
As happened in other states (and discussed in more detail below, with references/links), the exit polls indicated that Kerry was ahead. CNN was posting that information on their web site for many hours after the Florida polls had closed (though that data was later adjusted, apparently to make the polls reflect the reported results).

• Earlier polls
In addition to beating the exit polls, Bush also handily beat all the polls that had been taken immediately before the election (and in fact, almost all the polls that had been taken since the Summer). You can see that data HERE (aggregate on top, scroll down to see individual polls; you can also click to see similar information about other states). While polls are snapshots and imperfect, the fact that so many polls were all off and in the same direction and over a long stretch of time is interesting.

• More votes than voters?
Early readers of this site may remember that preliminary results from Florida showed more people had voted for President than had actually cast ballots in the state. Those numbers have since been corrected (though with no explanation as to the source of the error). However, that wasn't the end of it, as
this story shows that even after the corrections, at least one county had been reporting about 46,000 excess votes, about 30% more than the number of people who had voted. While none of this seems indicative of intentional attempts to manipulate the vote, it does indicate the imperfection (to put it kindly) of the systems in use, and errors are errors, regardless of intent.

So What Should We Believe?

The problem is, everyone who wants to believe one thing or the other can find reason to do so. Each side has data that cannot be easily explained away by the other side. Side A sees some unusual looking results and says "my gut says this looks wrong." Side B says, "here's some data to explain it." Side A says, "okay, but here is some data that says that that's not enough to explain everything we're seeing." Side B says, "my gut says that that data is also explainable, even though I can't (yet) produce the numbers."

I would like both sides to see past their gut. Just recognize the possibility that something may--or may not--be amiss, and so generate some will to actually have some of the ballots manually recounted, to be sure, either way. While all these arguments may sway someone one way or the other, they will never prove anything. Instead of trading theoretical arguments about statistics, correlations, historical voting patterns, whether people are honest to exit pollsters, whether results of past years can be counted on to be accurate, and whatever other theories people are coming up with on both sides, the question could be settled if people with authority and access would randomly pick, say, three optical scan counties and do a manual count of the ballot cards and see if they matched. Forget all this hypothetical back and forth, the answer is there to be had, if someone would just do the count. As one reader wrote to me, "Why put up all the fuss about making sure of a verifiable paper trail unless we are prepared to use it?"

NOTE: I have come up with a scenario described HERE where the voter turn-out and cross-voter exit poll data, combined with the geographic distribution, is consistent with the final reported numbers. It doesn't address the other concerns, and coming up with one possible explanation doesn't mean other possible explanations were wrong, but it does show that it is at least possible to make that data "work" with the reported results.

And what about the exit polls, in Florida and elsewhere?

Why were exit polls in crucial states wrong?

Why were they all wrong the same direction, saying that Kerry was getting more votes?

Why were they wrong only on the Presidential race, and right on other races?

Exit polls aren't perfect, but they are usually pretty darn good. The odds of an exit poll being off beyond the margin of error is small. Even within the margin of error, the odds of many such polls all being off in the same direction is also small. But that's what happened in this election.

While the exit polls were indeed quite accurate in many states, the polling in more crucial states tended to be wrong. And by what seems like a strange series of coincidences, they all erred in the same direction, showing Kerry ahead in states that ultimately went for Bush. If there were problems with the sampling in numerous, different states, wouldn't it have been more likely that some would be off in one direction, and some in the other? (Click HERE for more.)

Further, these "flawed" polls were apparently still surprisingly accurate in other races. If there were problems with the sampling, shouldn't the results for, say, the Senate race have been comparably off? That wasn't the case, according to this site, which reports, "Significantly, polls also showed Republicans carrying the bulk of the tight Senate races. However, when the official results were tallied, the exit polls for president proved wrong while the Senate polls proved right." If poor polling had skewed the results so as to favor the Democrats, why would it show up in only one race? How could the same sample be both sound and unsound, depending on which race you're looking at? Sure, that could happen by chance now and then, that a poll calls one race correctly and another one incorrectly, but it is something else when the same phenomenon is happening in multiple polls in multiple states simultaneously, and all in the same direction. (I would like to find more detailed references on this. If anyone reading this has some, please let me know. Specifically, in which states did this happen? And I would like to see the exit poll data for the Senate races in these states as it was reported on the night of the election, which may or may not be the same as what is currently available on the CNN site. We know that the posted poll data for the presidential race was subsequently modified to match the reported results, I don't want to assume that the Senate results were not.)

There is a report that the flawed vs. accurate results further corresponded to whether the state used electronic or paper balloting (see HERE). However, according to this site, the difference between the exit poll accuracy in states with electronic balloting versus paper is not statistically significant. But before you take solace in that, that same site reports that the odds of Bush having, overall, the gains he did relative to what the exit polls anticipated, was 1 in 50,000. And if that's not enough to raise an eyebrow, according to this site, the odds of, in particular, the combination of Ohio, Florida, and Pennsylvania shifting as much as they did between the exit poll and the final reported results is a whopping 250 million to one. (Yes, Pennsylvania went to Kerry, but by a much smaller margin than the exit polls had indicated, so that exit poll was also considered significantly flawed.) The most comprehensive treatment of this is in a PDF file that you can download HERE.

This site provides some further data and analysis concerning the apparent difference in exit poll accuracy when comparing battleground and non-battleground states. And according to this site, more specifically, many states shifted toward Bush beyond the margin of error, states with no paper trail to permit manual verification of the vote.

More detailed exit poll information has now been posted on this site.

Statistics is a science, not voodoo. At a certain point, statistical analysis can indeed tell you with near certainty whether or not certain conditions exist. Whether any of the analysis I've referenced here crosses that threshold, I will leave to the experts. But at the least, there are issues that should be examined.

And some more...

In one Ohio precinct, Bush was found to receive 4,258 votes to Kerry's 260. What made it particularly surprising is that only 638 people had voted. You can see the CNN story here.

There were scattered reports of touchscreens that asked the voter to confirm a choice for Bush after they had selected Kerry. You can see this CNN story here.

Isolated cases or tip of the iceberg? I don't know. And sure, they could be honest errors without malicious intent. But it is curious that all these errors always seem to favor the same candidate...

Q & A

Q: Do you really believe there was voting machine fraud in this election?

A: I am still agnostic on the issue. I am by no means convinced that there was fraud in the count. However, I believe that there are sufficient reasons to be suspicious, which warrant further investigation.


Q: What does Diebold have to do with all this?

A: Diebold is the Ohio company that manufactured many of the electronic voting systems used throughout the country, including new touch screen models (often criticized for the lack of a paper trail), as well as the devices that read voters' punchcard-style ballots in many areas, and the Windows PC-based system which computes the final tally. The CEO of the company is a vocal, staunch Bush supporter (infamously saying in 2003 that he was "committed to helping Ohio deliver its electoral votes to the president”), and so eyebrows have been raised by the fact that many of our votes are electronically tallied by the products of a company that has a vested interest in the outcome.


Q: How can opti-scan results be manipulated? Don't they use real, physical, countable ballots?

A: While the voter uses a physical card ballot, the cards are read by a machine. That card reading machine transfers its data to a PC. If we at least assume that people aren't playing games with the cards themselves, there are still two possible points of manipulation: the device that reads the cards and transmits the data to the PC could be somehow compromised, or the software on the PC could be compromised. In addition, once the data is on the PC it can conceivably be hacked to alter the results (look HERE for an overview, then here for a more detailed look).


Q: Is anyone doing anything about this?

A: The people at are currently engaged in a massive campaign to acquire documentation of many aspects of the election via the Freedom of Information Act. Ralph Nader has issued a press release urging that these matters be investigated, and is attempting to initiate a hand recount in New Hampshire. (Yes, Kerry won New Hampshire, but there was still suspicion of irregularities that favored Bush. Finding a flaw in any state will strengthen the case for looking at other states.) More information can be found here. Some of the machines being looked at are the same kind of optical scan machines as used in Florida. Finding problems with them would arguably increase suspicion of those results in Florida (and finding no problem may do the reverse). According to this press release, the Green Party and the Libertarian Party are mounting a challenge to the results in Ohio. I am uncertain about to what extent it will be a manual count, especially in light of the presence of some machines that provide no paper trail!


Q: Are there any more examples of electronic voting systems generating errors?

A: Click HERE for a site with numerous examples, and assorted other information you will probably find interesting. In addition, click here to read about a recent glitch just discovered in Indiana. And here's one about Utah county in Utah, where they discovered that 33,000 ballots (about 20%) hadn't been counted.


Q: I don't get it it... Which is the problem: the touchscreen machines or the card reading machines?

A: Both are electronic systems of a sort. In general, there has been more conern about the touchscreens. Statistical anomolies have raised eyebrows about the card reading machines in Florida in particular. The problem may be with one, the other, both, or neither. This site doesn't claim to know the answer, only that further investigation seems warranted. Some good information about the potential pitfalls of these systems can be found HERE.


Q: What's the story with North Carolina?

A: An interesting anomaly is described HERE, a very wide swing in North Carolina between the third of voters who voted early (or via absentee ballots) vs. the two thirds who voted with machines on election day. What makes it particularly noteworthy is that most of the results were pretty consistent between the two sets of voters, with the notable exceptions being a marked swing for the Republican Senate candidate and an enormous jump for Bush among the voters who used the machines on election day. In addition, there are thousands of missing votes, which you can read about here.


Q: Does it matter? Kerry already conceded.

A: The President is not elected based on the opponent's concession. The election is not final until the electoral college convenes to cast their votes on December 13 (A comment by Randi Rhodes on her program on AirAmerica said that the election is not final until those results are certified in early January, but my understanding is that that is a technicality, as I don't believe there is any mechanism by which a state's electors could alter their votes in the interim). If between now and then, it is determined that a state's electoral votes should go the other way... anything can happen. Is it likely? Not at all. For one thing, time is a factor. For another, even if there are errors in the counts--whether honest glitches, intentional fraud, or some combination--they would have to add up to a pretty big number to swing one of the crucial states from Bush to Kerry. No, it's not impossible. But regardless of that, IF there was fraud, even if by too few votes votes to alter the outcome of the election, it is worth pursuing to (a) help prevent it from happening again, and (b) make people liable for any criminal action that may have been involved. And if further investigation shows that there was apparently NO fraud, well, wouldn't that be nice to know, too?


Q: What about other kinds of election manipulation, besides the machines themselves?

A: Many other interesting events have been reported... Democratic voters in Ohio being canvassed by telephone with incorrect information about where they should go to vote... Flyers that said Republicans should vote on Tuesday, Democrats on Wednesday... Plenty of voting machines and short lines in areas expected to vote Republican, fewer voting machines and long lines in areas that would favor Kerry, which would discourage voting... etc. It is currently beyond the scope of this page to research and report on all of that, but do some searches, you'll find it.

Though Bob Mills, a pollwatcher in Hillsborough county, FL, wrote to me with this story about a local polling site: "In Bealsville, the line was consistently 1-1/2 hours long during my stint (7 a.m.-1 p.m.). A number of people complained that only five machines were on hand, yet a couple of months earlier, IN A PRIMARY ELECTION, the same polling place had ten machines available! Don't know what this means, except that somebody assumed the turnout would go down in a presidential election versus a local primary." Hmmm.


Q: I checked the exit polls on CNN's web site. They are right in line with the election results, they do not show Kerry ahead in states that Bush won. So what exit polls are you talking about?

A: Indeed, the exit polls were changed after the fact. In the evening on November 2, after the polls closed, the exit polls in Ohio and Florida (the ones I was checking personally) showed Kerry ahead. Some time in the middle of the night, the exit poll figures were altered to match the subsequently reported actual tally, with no explanation, or even an indication that a change had been made. The numbers appear to have simply been changed using some formula to bring their polls in line with "reality." I emailed CNN for an explanation, but have received no response so far. It would seem to me to have been much more honest to leave the exit polls as they were and say "look, sometimes exit polls can be wrong," rather than alter them after the fact to make them appear as if they were right all along. Luckily, there is still a web site where you can see what the exit polls were saying that day. You can find it HERE. I have comfirmed with someone at that site that the numbers presented at the top of that article were the final exit poll numbers reported, at "6-ish" Eastern time. In addition, some of the sites referenced in the exit poll section of this page have the original exit poll numbers. Another reference can be found here.


Q: Are you partisan?

A: I am decidedly not a fan of Bush. But I don't generally buy conspiracy theories. My skeptic streak cuts in both directions. I simply feel that the things discussed here make the issue of electronic vote manipulation worth looking into.

Comments? Send email to: [email protected]


Listen to "The Squanderer"
(parody of the old hit, "The Wanderer")