Wanted:precinct level election data for comparison studies

[Iceburg]

I am looking for datasets from the periods 1996 to 2004 that meet the following criteria:

Source: from official published canvas data, any state, any county (except Cuyahoga, Ohio ... I have that one and it is my reference point of comparison)

Granularity:precinct level

Voting method: paper, punched card, or other (must be known)

Ballot order arrangement: prefer datasets where the ballot order is known (may be fixed or rotating, if rotating I will require the specific ballot orders for each precinct)

Contests: must include presidential contest and at least one other party-based contest such as US senator

Import-ability: easily imported in db format

Variables Required:
Must have as a minimum:
- candidate data
- spoiled (under votes, overvotes)
- number of registered voters
- voting method
- ballot order (see requirement above)

Nice to Have:
Number of voting machines deployed at each precinct
----------------------------------------------------------
Any assistance in finding such datasets would be GREATLY appreciated!!!

[SheilaT]

1. Not sure how you're going to get this data.

How long is such information retained? I suspect you'd need to contact the election boards in places you're interested in.

I can tell you that here in Johnson County, Kansas we use electronic voting machines on election day, and optical scan paper ballots for advance voting. Since I ran for office (Kansas State House, lost) I had observers at each of my 16 precincts on election day to tally my votes as they were taken off the machines. And at least a third of them did not work, meaning a print-out could not be gotten at the polling place, and was presumably obtained when the computer chip was brought to the county election board and the official voting results were done. A third. in a state where Bush was going to win legitimately. I don't think this unable-to-print-the-results is indicative of fraud, but it makes you wonder about the quality of the machinery.

[Iceburg]

2. Many of the counties have published their data in text format

which is reasonably easy to clean and import into a database. The machine data, such as number of machines deployed at each precinct is much harder to come by and thankfully is not necessary for my study. With respect to the ballot order data, while I would love to have this in my hands it would be sufficient to know simply whether the ballot order was fixed/constant in each precinct or rotated.

I agree that "unable -to-print-the results" is NOT indicative of fraud and as such am not looking for results that were published in the precinct but rather the final tallies for each precinct as published by the election authorities for a given county.

Sorry about your loss. Don't give up your hopes and dreams with that set-back.

Regarding the quality of the machinery ... after months of analysis I am regrettfully a member of the camp that believes the quality of the equipment is by design. I say regrettfully because I would much rather be resting comfortably in the "I-trust-the-machinery-and-the people-who-manage-it" camp.

[berniew1]

3. I'm interested in getting Broward County Florida 2004 precinct votes

and polling place addresses

[Iceburg]

4. Berniew, do you have any datasets at the precinct level?

As you are probably aware, the distribution of "uncorrected" votes in Cuyahoga for all presidential candidates does not follow a normal distribution (in the statistical sense). Once the outliers are removed/corrected (close to 200 using rigorous criteria) the distribution appears to satisfy the criteria for a normal distribution. Further, using regression techniques, the pockets in which the theft took place become very apparent. My analysis indicates that most of the theft took place at:
- the machine level; and
- in bulk (multiples of avg votes per machine at a voting address)and not random x-precinct voting; and
- vote addresses with cascading ballot orders

The primary reason I am seeking the datasets is to determine whether or not one can expect a normal vote distribution in an untainted election.

[DoYouEverWonder]

7. Here's the Broward Co 2004 precinct votes

http://search.netscape.com/ns/boomframe.jsp?query=fl+dept+of+state&page=1&offset=0&result_url=redir%3Fsrc%3Dwebsearch%26requestId%3D1147247876e71d2d%26clickedItemRank%3D2%26userQuery%3Dfl%2Bdept%2Bof%2Bstate%26clickedItemURN%3Dhttp%253A%252F%252Felection.dos.state.fl.us%252F%26invocationType%3D-%26fromPage%3DnsBrowserRoll%26amp%3BampTest%3D1&remove_url=http%3A%2F%2Felection.dos.state.fl.us%2F

This link might have the data in other formats. The one above sucks.

http://search.netscape.com/ns/boomframe.jsp?query=fl+dept+of+state&page=1&offset=0&result_url=redir%3Fsrc%3Dwebsearch%26requestId%3D1147247876e71d2d%26clickedItemRank%3D2%26userQuery%3Dfl%2Bdept%2Bof%2Bstate%26clickedItemURN%3Dhttp%253A%252F%252Felection.dos.state.fl.us%252F%26invocationType%3D-%26fromPage%3DnsBrowserRoll%26amp%3BampTest%3D1&remove_url=http%3A%2F%2Felection.dos.state.fl.us%2F

[Wilms]

5. Are these links a help?

Atlas of US Elections
http://www.uselectionatlas.org/

USCVprogs
http://uscvprogs.sourceforge.net/

Election Data Services
http://www.electiondataservices.com/home.htm

LA County Department of Registrar- Recorder/County Clerk
http://rrcc.co.la.ca.us/elect/downrslt.html-ssi

[Iceburg]

6. Thanks Wilms, I will check each and every one of the sites

out.

I did manage to get all of the Michigan and New Hampshire data (for all precincts)from their respective SOS's and since these two states use a variety of voting methods the datasets will provide a powerful point of comparison.

[Wilms]

8. Some more...

Approved Voting Equipment:

Source: http://www.nased.org/index.htm

http://www.nased.org/ITA%20Information/NASEDQualifiedVotingSystems12-03-12-04.pdf

http://www.nased.org/ITA%20Information/NASEDApprovedSystems1.03.pdf



Voter Registration and Turnout Statistics +

Source: http://www.eac.gov/

http://www.eac.gov/election_resources.asp?format=none