Walter Mebane (Cornell University) has updated his research paper on the Franklin County (Ohio) voting machine allocation debate. His updated paper is available from his website.
Walter’s updated analysis includes some newly received data from Franklin County. Here is a snip of the email that Walter sent to the electionlaw listserv regarding his updated database and his new analysis of that data:
Late in the afternoon on February 8, 2006, Matt Damschroder, Director of the Board of Elections in Franklin County, Ohio, responded via email to several questions I had asked Franklin County officials about data they had sent me pertaining to their voting machine allocation decisions for the 2004 general electon (to be precise, he responded to a message sent by my research assistant). We had asked the Franklin County Board of Elections to supply the count of voters in each precinct that they used to make the machine allocations. In response they had sent us several files on January 23, 2006. We then asked, among other questions, which of several alternative formulas matched the measure used to make the allocation decisions. Damschroder’s message was in response to those questions.
In light of Damschroder’s detailed responses, I now believe that the data provided by Franklin County do not allow us to recover the information or the decision rules Franklin County officials used to allocate voting machines to precincts for the 2004 general election.
I have accordingly updated my paper posted at
The description I included in my previous message on this subject (sent at 8 Feb 2006 02:22:11), regarding the relationship between the racial composition of precincts and the number of voters per machine, is not meaningful. Damschroder’s responses make it clear that the variables I used to try to measure the number of active voters in June
cannot be used to do that. See the paper for more details.
Thanks to Walter for continuing his research on this question and for providing this updated study for public distribution.