|
|
![]() |

2004 Florida and Pennsylvania Registration and Voting
Recently there has been a discussion of the trends in voter registration and voting in Florida counties at: http://www.ustogether.org/Florida_Election.htm
The USTogether discussion and associated links present voting and registration data in a variety of ways. JOHO adds to that discussion by presenting a single comprehensive view that includes all of the counties, voting machine types, and county size for both Florida and Pennsylvania in one easily interpreted image shown above.
This JOHO image shows the reversal of voting percentages from party registrations in the smaller Florida counties where Diebold optical scan devices were used in the 2004 presidential election. It shows there was no reversal in larger op-scan counties, E-Touchscreen counties, or any county in the whole state of Pennsylvania.
There is a very large amount of information in this image. Patience and careful study will help you to discover the trends discussed below. There is no statistical analysis, curve fitting, or error analysis required here. The image is made just from raw numerical counts. Your eyes can see the trends discussed at USTogether without relying on expert interpretation and analysis.
The Data
The left panel of this JOHO image shows the raw data from Florida contained in a table at:
http://ustogether.org/election04/FloridaDataStats.htm
This large table shows the number of registered voters in the Republican and Democratic parties, and the number of votes cast for the party candidate for each county in Florida that used Diebold optical scan or Diebold E-Touchscreen voting machines. We have calculated percentages from the counts and totals in that table. The table includes information about the type of Diebold machine used. The Pennsylvania data is from a similar table at http://ustogether.org/election04/PA_vote_patt.htm.
JOHO image description
The Florida panel on the left shows data for Florida counties that used Diebold’s voting machines. The top section shows counties that used E-Touchscreen voting machines. The bottom section shows counties that used Diebold’s optical scanning voting machines.
Each row in the panel represents one Florida county, as labeled on the left. The rows in each section are sorted so that the county with the smallest number of total votes is at the top and the largest county is at the bottom. Total vote labels are on the right. There were about 4.1 million E-Touch votes and 3.5 million Op-Scan votes in Florida.
The first column shows the percentage of the total registered voters that are registered as Republicans. The second column shows the percentage of total voters that voted Republican in the 2004 presidential race. The third and fourth columns show the same pair of numbers for the Democratic Party.
The color in a cell represents this percentage. Use the legend on the right to translate color into the corresponding percentage. Percentages under 50% are green or blue. Those over 50% are yellow, red, or purple.
The Pennsylvania panel on the right has the same layout as the Florida panel. There are four sections because four different types of voting machine were used.
If voting percentages followed party registration percentages the columns in each pair would have the same color.
Florida Panel
The Florida panel shows that, in Florida, the registration and voting patterns for the two voting machines are different. In the top E-Touchscreen section there is a slight increase from the registration percentage to the vote percentage. The effect is larger for the Republican columns (average Republican Reg:Vote = 43:56, average Democrat Reg:Vote = 36:44).
This effect appears as a tendency to move to a redder (higher) color when going from the registration column to the vote column. It is not unexpected because voters not registered for either party (but counted in the total registrations used to calculate the percentages) may have voted for one of the party candidates. The much discussed GOP GOTV efforts in Florida and the gradual swing from Democrat to Republican voting without corresponding changes in party registration may also have contributed to the larger effect for Republican voters.
In the smaller (Liberty-Jackson) optical scan counties this tendency is greatly exaggerated. The image shows that these small counties are predominantly Democrat in registration, but voted overwhelmingly Republican. Both characteristics are worthy of some explanation. The Reg:Vote ratio increased dramatically for Republicans (Reg:Vote = 22:64), and dropped significantly for Democrats (Reg:Vote = 70:35). Gadsden is the smallest op-scan county where the registrations roughly match the votes like they do in the E-Touch counties.
Gadsden and larger op-scan counties, where the registration percentages are more nearly equal, show a rough similarity between registrations and votes like the E-Touchscreen counties (average Republican Reg:Vote = 40:58, average Democrat Reg:Vote = 41:41).
Pennsylvania Panel
The Pennsylvania panel on the right has the same layout as the Florida panel. There are 4 kinds of voting machines used, including the optical scan machines in the third section. This panel does not show the huge flip-flop of voters in optical scan counties. There is the same general tendency for the percentage of Republican votes to be slightly higher than the percentage of registered Republicans, and the reverse for Democrats.
About JOHO
JOHO is discipline-independent data visualization software that can display up to one million numerical values on a PC screen in one easily interpreted view. We produced this JOHO image by pasting the tables from the links given above into Excel, calculating the percentage columns, sorting the rows by vote total, and importing the resulting spreadsheet into JOHO with a couple of mouse clicks.
JOHO makes it easy to inspect and absorb the data in the table visually. The JOHO image has about 540 numbers. Other graphics programs have difficulty displaying this amount of information in an easily interpreted form.
To see more examples of JOHO images please visit the rest of our web site using the links below.
| |Home| |Financial Images| |DSI Service| |Traffic Images| |Communication Images| |Medical Images| |Election Results| |
| |Joho Software| |About Us| |FAQ| |Contact Us| |Request Info| |