Kerry lost? ("Op-Scan Voting Machines Bring out the Republican Vote" Edition)

Continuing our Kerry Lost? feature:

Check out this table.

In Florida's Calhoun county, four times as many Republicans voted as might be expected from the number registered. Over three times as many Republicans voted as might be expected in Dixie county.

In fact, there are a large number of Florida counties which recorded a larger than expected number of Republican voters and a lower than expected number of Democratic voters. And, all of those counties are using the Op-Scan voting machines.

However, if you look at the counties that use E-touchscreen machines, there aren't the huge number of voting fluctuations.

I don't claim to completely understand what exactly the column headers mean, where they're getting their numbers from, and whether they're accurate or not, but look at the table. Someone needs to look into this.

UPDATE: See my analysis of the "Reagan Difference." Basically, voting for Republican presidential candidates appears to be the norm for those in the counties mentioned above. And, Bush II's support is down markedly from the support most Florida counties gave to the Republican presidential candidates during the Reagan era.

Comments

Rather than using the registration numbers to estimate a number of Bush and Kerry voters, I would use the Bush-Gore counts. This seems like a better baseline to compare against than registrations. Registrations stack up and are not pruned well. Actual vote counts should predict future vote counts better. A quick scan there doesn't show the problem you are seeing. For example, consider the expectation for Baker county was 2,400 Republican votes and 6,800 Dem ones. Really? A rural Florida county expected to vote Democratic? Is that how they voted in 2000? No, check http://www.sbgo.com/election.htm and you see that Baker was 5,600 Bush to 2,400 Gore. So 7,700 Bush to 2,100 Kerry should be a dissapointment to Kerry, but not a shock. The 2,400 Bush to 6,800 Kerry expactation is what's wrong.