<x-flowed>
Walter Mebane writes:
In a message timestamped Monday, July 25, 2005 4:27 PM, Michael
McDonald raises some questions about my analysis of the allocation of
voting machines to precincts and the consequences of that allocation
in Franklin County, Ohio, for the 2004 general election. In a piece
published by tompaine.com
(http://www.tompaine.com/print/timing_and_turnout_in_ohio.php), I
wrote an op-ed response to the DOJ letter that reported the
termination of their investigation of the county. The detailed
analysis of data from Franklin County that I drew on to write that
op-ed is presented in a longer paper available from
http://macht.arts.cornell.edu/wrm1/franklin2.pdf. McDonald seems to
be responding to the tompaine.com piece but not to be aware of the
franklin2.pdf paper.
In particular, McDonald takes issue with analysis conducted using the
number of registered voters as the denominator when computing voter
turnout. In fact I also considered turnout using the number of
"active voters" (a concept defined and measured by Franklin County).
I find that "the typical reduction in voter turnout---when comparing a
precinct that has a machines per active voter ratio near the third
quartile to a precinct having a ratio near the first quartile---is
more than five times larger in the set of precincts with high
proportions African American than it is in the set of precincts with
low proportions" (Mebane 2005, 8).
At one point McDonald suggests that my analysis confirms the DOJ
findings. He states, "Furthermore, from the DOJ report, which Mebane
confirms as being true: 'the allocation of voting machines actually
favored black voters because more white voters were voting on each
voting machine than black voters.'" In fact I dispute many claims
made in the DOJ report. I specifically do not agree that "the
allocation of voting machines actually favored black voters." I find
that there were more ballots cast per voting machine in the precincts
that had the smallest proportions African American than in the
precincts that had the largest proportions African American: "In the
set of precincts with low proportions African American there were
typically 170 ballots cast per voting machine while in the set of
precincts with high proportions African American there were typically
164 ballots cast per voting machine" (Mebane 2005, 10). But I go on
to show that "Far from being a measure of the fairness of the voting
machine allocation, however, the relationship between the number of
voting machines and the number of ballots cast is primarily an
indication of the length of time it took for each voter who did manage
to vote to complete the ballot" (Mebane 2005, 10).
There are other claims stated in the DOJ letter that, in the
franklin2.pdf paper, I find are not true. For instance, the DOJ
report also claims that there is a "tendency in Franklin County for
white voters to cast ballots in the morning (i.e., before work), and
for black voters to cast ballots in the afternoon (i.e., after work)"
(Tanner 2005, 3). Survey data gathered for the DNC Ohio 2004 report
shows that not to be true, at least when considering a random sample
of voters from across Ohio: "during the late afternoon and early
evening period there is no disparity in the proportion of each group
arriving at the polls: 27 percent of white voters and 28 percent of
African American voters arrived at the polls to vote after 3 p.m."
(Mebane 2005, 2). See the franklin2.pdf paper for other details of
the DOJ letter I dispute.
McDonald says that I argue "that Franklin County purposefully
discriminated because they used the April 1 active registration count
rather than the Nov. 4 active voter registration count when allocating
voting machines to the precincts." I do not argue that Franklin
County officials "purposefully discriminated." The data available to
me do not on their own support any inferences about the reasons why
Franklin County officials acted as they did. Conjecturing that
Franklin officials used the spring information about active voters to
make the machine allocations, I wrote the following:
"If the allocation of voting machines is compared to information
about the size of the active electorate that was available to
Franklin County election officials at the end of April, 2004, then
the allocation of machines is not biased against voters who were
active at that time in precincts having high proportions of African
Americans. But if election officials did use that information to
make their allocation plans, then they made plans that involved
using a total number of machines that was nearly 45 percent too
small. Even using the April measure of the size of the active
electorate, 5,023 working voting machines were needed, not 2,800
machines as data supplied by the county indicate were actually
deployed on election day. And sticking with plans possibly made
using the information from April meant that the officials ignored
information during the summer and fall that showed that the November
electorate would be substantially larger. Between April and
November, voter registration in the county increased by 15 percent.
If nothing else, the surge of new registrants should have been a
clear indicator that plans made based on the April information would
prove woefully insufficient." (Mebane 2005, 12).
References:
Mebane, Walter R. Jr. 2005. "Voting Machine Allocation in Franklin
County, Ohio, 2004: Response to U.S. Department of Justice Letter of
June 29, 2005." July 7, 2005.
http://macht.arts.cornell.edu/wrm1/franklin2.pdf
Tanner, John. 2005. Letter to Nick A. Soulas, Jr., Assistant
Prosecuting Attorney, Franklin County, Ohio. June 29, 2005.
http://www.usdoj.gov/crt/voting/misc/franklin_oh.pdf (accessed July 2,
2005).
Walter Mebane
</x-flowed>