I find this especially egregious given that our foreign policy currently is to try to take democracy to other countries, and it appears we've still a lot of bugs to work out in our own system. We previously reported on the inaccuracies of electronic voting machines, and these have definitely come through.

First, we go to Ohio, where an electronic voting machine (made by Diebold, interestingly enough) gave Bush more than 4,258 votes, in a county where only 638 votes were cast1.

Then, in North Carolina, a computer lost 4,500 votes because election officials believed that each storage unit could hold 10,500 votes (they were so told by the manufacturer), when in reality it could only hold 3,0052.

In Miami, electronic voting machines lost votes on a state constitutional amendment, because of a glitch. Basically, the machine was in charge of tabulating totals from 5 different voting machines. Since none of the counties in question have more than 32,000 voters, once the machine gets to that number of votes, it starts counting back down. This glitch has been known to the company that made the machine, ES&S, since 20023.

None of these mistakes affected the outcome of the election. That being said, they alone show we still have a lot of work to do in terms of making sure every vote is counted accurately. Again, these would be bad enough, but there's more.

If the 6pm exit polls are to believed, more people voted for Kerry than for Bush. This is where things get interesting:

  • Florida: CNN's exit polling had Bush leading by 5355 votes4, with the 6pm exit polls at Daily Kos giving Kerry a 2% lead5. Bush won the state by more than 300,000 votes.
  • Minnesota: The aforementioned exit poll gives Kerry a 10% lead, but he only won by 3%6.
  • Ohio: Kerry led by 2% in the exit polls, with Bush taking the state by 2% (a shift of more than 272,000 votes)7.
  • Nevada: Kerry led exit polls by 1.2%, with Bush taking the state by 3%.

Now, some variance from exit polls is to be expected (the Blue Lemur article mentions more examples), and that's fine. What's alarming, however, is that exit polls were much less accurate in states with electronic voting machines that do not leave a paper trail. In those states, there is a consistent shift towards Bush between exit polls and final results. If one looks at this image, it shows the discrepancies between exit polls and actual results in 9 states. Those that use paper ballots are either dead-on or see a shift of a couple percentage points. Those that use electronic ballots, on the other hand, see much more drastic changes. Considering the accuracy of exit polls in other states, this is highly suspect.

BlackBoxVoting has issued the largest Freedom of Information Act request in U.S. history, and is trying to sort out what actually happened. Anyone interested in aiding the effort should check out their site and pitch in.

It may be awhile before we figure out the actual results (if ever). Nothing conclusive has come out yet, but I encourage everyone to keep a close eye on this, and if you really want to help, join the BlackBoxVoting campaign mentioned above. One thing is clear, though: we have a long way to go.

UPDATE: Another blogger has found that there are 13 counties in Florida that list higher numbers for Bush than they list for turn-out. For example, one county listed voter turnout as 7,460,556, and this same county listed the number of votes for Bush as 7,586,995; that's 126,439 more votes for Bush than voted. The story is here.

ANOTHER UPDATE: Another issue in Florida has arisen. It larger Florida counties (regardless of voting method), percentages of registered Democrats and percentages who voted for Kerry are basically the same. In smaller counties, however, it was basically the reverse. In Baker County, for example, 70% of voters are registered Democrat and 24.3% are Republican. The final totals, however, were 2,180 votes for Kerry and 7,738 for Bush. In Dixie County, 77.5% are registered Democrat and 15% are registered Republican, yet the vote totals were 4,433 for Bush and only 1,959 for Kerry (note: the article misquotes the number of total registered voters; according to the usatogether.org chart, the county has 9,676 registered voters). ( source)

Further updates as events warrant.

Sources:

  1. E-Vote Glitch Inflates Bush Total -- Wired News.
  2. N.C. Computer Looses More than 4,500 Votes -- Associated Press.
  3. Defective Software "Lost" Votes -- Miami Herald.
  4. Exit Polls and "Actual" Votes Don't Match -- The Blue Lemur.
  5. 6pm Exit Polls -- Daily Kos.
  6. Election Results - Minnesota -- CNN.
  7. Election Results - Ohio -- CNN.
vote FOR this article vote AGAINST this article flag as spam/abuse
Log in if you are DEMachina for more options. 
Find similar articles