Revisiting the 2004 election, part 7: Florida

2004 election revisited, part 7: Florida further

Following up previous posts

Obama at SOTU

President Obama’s State of the Union address Tuesday night demonstrated again the greater quality of this president over the loon rhetoric coming from our current GOP debates. The quality gap is huge, and growing larger.

Mitt Romney

As written before, the guys coming out of the GOP field are going to need all the help they can get at election time, and previous elections involve some warning signs. Forewarned is forearmed.

In the 2000 election, 25 Florida counties used the maligned punch-card ballots, 41 used op-scan, and one county used paper ballots counted manually. Seventeen counties switched to optical scanners for the 2004 election.


Counties that switched from punch-card to op-scan did not necessarily show the biggest swings to Bush, but there was a distinct difference—in Bush’s favor–between op-scan counties and other counties. The picture over-all:

  • Total registered voters in the 15 counties using touch-screens:  5,576,264
  • Total registered voters in the 52 counties using op-scan ballots:  4,725,026
  • Outcome in the touch-screen counties:  Kerry 1,983,210 to Bush 1,845,876
  • Outcome in the op-scan counties:  Bush 2,110,414 to Kerry 1,591,790

Statistics come from election results and from election researcher Kathy Dopp’s analysis of touch-screen machines versus optical-scan paper voting in 2004.

There were further differences between op-scan voting results and touch-screen results, Florida 2004:

  • 15 Florida counties used touch-screen voting machines. Only 3 of these counties showed a 100+ percent jump in Republican votes, and 3 others showed a 100+ percent jump in Democratic votes. Neither party jumped more than 120 percent.
  • 52 counties used op-scan ballots. Of these, 43 showed a triple-digit jump for Republicans, two for Democrats. One had a whopping 602 percent jump in votes for Repubs, one over 400 percent, and two over 300 percent. Another ten op-scan counties had an over 200 percent jump in votes for Repubs. Nothing equivalent for Dems.

Leaving the eye-blearing numbers aside, think about the larger context of the 2004 election. Was John Kerry two or three times less popular in Florida in 2004 than Al Gore in 2000? Was GWBush—after the escape of Osama bin Laden, the invasion of Iraq, the war profiteering, the tax cuts for the rich—two or three times more appealing?

Back to those voting machines

Only three counties using op-scan machines had larger jumps for Democrats than for Republicans. The percentages below are vote gains, 2000 to 2004:

 (Dem Senate candidate Betty Kastor won Flagler County over Mel Martinez.)

Only four counties using op-scan had balanced percentage jumps in party voting:

Note that where the jump in voting by party was balanced, the outcomes were also reasonably divided. No one party or candidate (Bush) won all the time.

Turnout always matters.

Ten Florida counties with biggest turnout by number, in 2004:

  • Miami-Dade:  772,743
  • Broward:  707,202
  • Palm Beach:  547,340
  • Hillsborough:  464,253
  • Pinellas:  457,426
  • Orange:  388,095
  • Duval:  379,257
  • Brevard:  265,764
  • Lee:  242,434
  • Volusia:  229,098

Of these, Brevard, Duval, Orange, and Volusia used op-scan voting.

Ten Florida counties with highest turnout by percentage of registered voters, in 2004:

  • Flagler:  81.9% –jump gap only 102 to 103, Bush won, but close
  • Sarasota:  81.6%
  • Jefferson:  80.7% –jump gap 171 to 61, Kerry won
  • Leon:  79.8% –jump gap 105 to 81, Kerry won
  • Lee:  79.5%
  • Nassau:  79.2%
  • St Johns:  79% –jump gap 100 to 84, Bush won
  • Sumter:  79%
  • Brevard:  78.6% –jump gap 101 to 89, Bush won
  • Gadsden:  78.5% –jump gap 207 to 66, Kerry won

Of these, Brevard, Flagler, Gadsden, Jefferson, Leon and St Johns used op-scan voting. Nobody flipped the elections in Gadsden or Jefferson and Leon counties.

Ten Florida counties with lowest turnout by percentage of registered voters:

  • Hendry 57.2%
  • Osceola 63.3%
  • DeSoto 64% –jump gap 146 to 44, Bush won
  • Okeechobee 65.7% –jump gap 126 to 47, Bush won
  • Union 66.7% –jump gap 263 to 23, Bush won
  • Broward 66.8%
  • Dixie 66.9% –jump gap 305 to 26, Bush won
  • Hamilton 67.1% –jump gap 244 to 37, Bush won
  • Putnam 68.5% –jump gap 144 to 47, Bush won
  • Highlands 69.6% –jump gap 77 to 44, Bush won

Of these, DeSoto, Dixie, Hamilton, Highlands, Okeechobee, Putnam and Union all used op-scan voting. Low turnout, discouraged voters, unappealing voting technology—Bush swept.

The pattern held in the ten counties with lowest voter turnout by number, in 2004. Numbers are local turnout as reported:

  • Liberty 3,051 –Bush won
  • Lafayette 3,352 –Bush won
  • Glades 4,204 –Bush won
  • Union 4,714 –Bush won
  • Hamilton 5,131 –Bush won
  • Franklin 5,973 –Bush won
  • Calhoun 6,006 –Bush won
  • Dixie 6,472 –Bush won
  • Gilchrist 7,047 –Bush won
  • Hardee 7,281 –Bush won

Of these, all used op-scanned paper ballots, and all went for Bush. We have a winner.  Were they Florida’s ten smallest counties? Not quite: A list of the state’s ten smallest counties would bump Hardee, Gilchrist, Dixie, and Union. Baker, Bradford, Gulf, Holmes, Jefferson, Madison, Taylor, Wakulla and Washington all had populations in the twenty-something thousand range. The smallest counties also all used op-scan ballots, so we have another winner: Of these, only Jefferson went for Kerry, and Madison was close.

On a brighter note, in all these counties but Hardee, about half the total population (2000 census) turned out to vote–not too shabby for the lowest numerical turnout in the state. It was a high-interest election.

The middle

Continuing the test of brain-strain, let’s look at some middling turnout. Since the highest turnout ran 82% and the lowest percentage about 57%, midpoint would be about 70%.

Ten Florida counties with mid-range turnout, 2004:

  • Charlotte:  70.4%
  • Glades:  70.5% –jump gap 134 to 37, Bush won
  • Hardee:  70% –jump gap 182 to 32, Bush won
  • Okaloosa:  70.5% –jump gap 95 to 61, Bush crushed
  • Calhoun:  71.9% –jump gap 381 to 30, Bush won
  • Highlands:  69.6% –jump gap 77 to 54, Bush won
  • Pasco:  72.1%
  • Santa Rosa:  69.9% –jump gap 97 to 54, Bush won
  • Putnam:  68.5% –jump gap 144 to 47, Bush won
  • Polk:  71.5% –jump gap 107 to 68, Bush won

Of these, all but Charlotte and Pasco counties used op-scan. Charlotte and Pasco had predictable results, in line with voter registration and with voter turnout, with some crossover (Dem for Charlotte, Repub for Pasco).

I wrote on this material at the time.

A similar analysis can be found here. The author further pursues similar anomalies in the rural areas of other southern states. The author sums up:

“I started this page when I first saw the breakdown by voting machine of the results in Florida. Since the way people vote shouldn’t depend on the machines they use, it seemed to signal cheating. A careful examination of the voting patterns in Florida in 2000 and in Louisiana in 2000, however, has led me to conclude that the difference was due to the lopsided use of the opscan machines in rural northern Florida counties where there is a preponderance of “Dixiecrats”. A statistical analysis of the data from the 1996 election also supports this conclusion. The remainder of this page contains my analysis and reasoning in detail.”


This entry was posted in Blog, Featured and tagged , , , , , , , . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *