Sunday, November 14, 2004
Ballot Fraud?
On Tuesday afternoon, 2004 November 2, I looked at some exit polls on the site http://www.electoral-vote.com . They clearly showed a trend towards Kerry, including in Pennsylvania, Ohio, and Florida. I watched that evening as the votes were being tallied. Everything went as predicted (except for unusually high, in my opinion, margins for Bush in Indiana and in the entire country). Then CNN's anchors expressed concern about the returns in Florida. Not only was central Florida going to Bush, but the counties in central and northern Florida were going more for Bush than they were in 2000. That meant that the state was leaning to Bush. The same sorts of things occurred elsewhere, and it ended with Bush taking Ohio by a 3-point margin and the popular vote nationwide by 3 percent. So the exit polls were wrong. They polled too many women or something.
But then came the articles on this Internet. They kept saying ballot fraud all over the place. Votes were being thrown out, machines were erring for Bush and so forth. Sounds like conspiracy theories to me. Except for one thing. Consider this site: http://ustogether.org/Florida_Election.htm . It shows registrations of voters as Republicans and Democrats, and it compares how the voters voted for President versus their party affiliation. Now I know that Democrats vote for Bush. In fact, in some northern Florida counties, most of the residents are Democrats but most vote for Bush and have voted Republican for years. What concerns me about the table in this site is that it says that whether voters vote more for Bush then their party affiliations would indicate is a function not of sex, not of sexual orientation, not of race, not of party affiliation, not of the area of Florida they live in, or in any other sociological or demographic variable, but on what type of machine was used to record the vote. One machine produced far more Bushward electoral votes than the other. This suggests a bias in that machine. Of course we are assuming a lot of things; among them, the authenticity of the data on the http://ustogether.org/Florida_Election.htm website. But if the data is official, then something has to explain this dependence on the type of machine. I think it should be investigated. I don't know if it will affect the outcome, though, but I still think we should look at it and see if there is anything to it.
But then came the articles on this Internet. They kept saying ballot fraud all over the place. Votes were being thrown out, machines were erring for Bush and so forth. Sounds like conspiracy theories to me. Except for one thing. Consider this site: http://ustogether.org/Florida_Election.htm . It shows registrations of voters as Republicans and Democrats, and it compares how the voters voted for President versus their party affiliation. Now I know that Democrats vote for Bush. In fact, in some northern Florida counties, most of the residents are Democrats but most vote for Bush and have voted Republican for years. What concerns me about the table in this site is that it says that whether voters vote more for Bush then their party affiliations would indicate is a function not of sex, not of sexual orientation, not of race, not of party affiliation, not of the area of Florida they live in, or in any other sociological or demographic variable, but on what type of machine was used to record the vote. One machine produced far more Bushward electoral votes than the other. This suggests a bias in that machine. Of course we are assuming a lot of things; among them, the authenticity of the data on the http://ustogether.org/Florida_Election.htm website. But if the data is official, then something has to explain this dependence on the type of machine. I think it should be investigated. I don't know if it will affect the outcome, though, but I still think we should look at it and see if there is anything to it.
