2004 Florida Voter Reg. Info vs. votes for Bush

This is a pretty busy graph, so it may require some explanation:

• The Florida counties are listed across the bottom (X axis), ordered left to right, from the most to the least number of registered voters.
• The red area is the percentage of all of the registered voters who were registered Republican, by county.
• The white area is the percentage of registered voters who registered something other than Republican or Democrat, or were registered unaffiliated.
• The blue area is the percentage of registered voters who registered Democrat.
• The yellowish line with diamonds is the percentage of the registered voters who said they were "white."
• The red line with squares is the percentage of the total vote that was cast for Bush.

I laid the chart out this way so you can kind of see how much of the non-Dem-Rep vote Bush received (or not).

For example, all the way on the left, Miami-Dade county has the most registered voters. 45.72% of the votes cast for the office of President were cast for Bush.

Voter Registration stats for Miami-Dade county:

• 34.79% Republican
• 22.37% Other Party (or none)
• 42.84% Democrat
• 26.22% said they were white

By contrast, all the way on the right, Liberty county has the least number of registered voters of any county in Florida. 63.79% of the votes were cast for Bush.

Voter Registration stats for Liberty county:

• 7.85% Republican
• 3.88% Other Party (or none)
• 88.27% Democrat
• 88.96% said they were white

Stuff i've noticed

• In many counties, Bush's percentage of votes tends to hang about 15% lower than the percentage of white voters.
• In the counties with fewer registered voters (right half of the chart) Bush's votes very consistently track 10% behind the percentage of white voters.
• Gadsden, about 25th from the right, stands out as a deep dip of blue. Look at the high percentage of registered Democrats, and how poorly Bush did. Compare Gadsden with Lafayette and Liberty counties (far right). There were similar percentages of registered Democrats, and yet Bush received over 60% of the vote in both Lafayette and Liberty compared with just 30% in Gadsden. The obvious difference is the percentage of the registered voters who said they were white: ~40% in Gadsden compared with about 90% in Lafayette and Liberty. However it turns out there's more to the story about Gadsden county. Gasden has a very large African-American population, and almost 2,000 votes were disqualified and not counted in the 2000, perhaps because of poor-quality machines. (Some claimed discrimination.) So perhaps there may have been a large percentage of the heavily African-American Democrats in Gadsden who were intent on having their votes counted in 2004.
• Something remarkable definitely happened in the lower population counties, on the right side of the chart. Unless fraud or errors resulted in incorrect counts, Bush must have gotten large numbers of registered Democrat voters from these counties to vote for him. There's just not enough non-Democrat-registered voters to add up to what Bush received. It's really quite amazing. In many of these counties, Bush was scoring consistently 30 points higher than the voter registration numbers would suggest.
  But before you decide those counties gave Bush the win, take a look at this chart of votes for Bush by county. There's just not that many registered voters in those counties. If you add up all the votes Bush received in those counties, and subtract what Kerry received in the same counties, you're talking about 50,000 votes for Bush, net. If you exclude a few of those counties that are less than 65% white, then the net gain for Bush is about 70,000. A substantial number, but still less than 25% of Bush's overall 370,000+ margin of victory in Florida. Bush won the election in Florida because of the votes he received in Florida's other counties.

A PDF version of this chart

A simpler version of this chart

$Id: reg-vs-bush-votes.html,v 1.4 2004/11/09 09:25:48 johan Exp $