Reply #62: Using Kerry/Gore discrepancy to indicate counties to analyze further...

[IndyOp]

Time for Change & EOTE's original analysis compared Triad and non-Triad counties. Interesting result.

The Kerry/Gore discrepancy could be used in other ways: (1) To determine whether Arnebeck is fingering all of the counties in which fraud most likely occurred, and (2) to determine which counties should be analyzed further at a precinct-by-precinct level.
====================================

Does Arnebeck's Bush v Moss include all/most counties in which the Kerry/Gore discrepancy is largest?

Did Kerry do relatively worse than Gore in the counties implicated in fraud by Arnebeck's lawsuit, but as well as or better than Gore in counties not implicated in fraud by Arnebeck's team? Why do this analysis? To see if Arnebeck is focusing on all/most of the counties in which the Kerry/Gore discrepancy suggests there was fraud.

The 36 counties implicated in switching votes from Kerry to Bush in Arnebeck's suit are a mix of Triad (18), ES&S (16), and Fidlar (2) counties (named in footnote). I believe that the implicated counties are those in which John Kerry received 1,000 (or more) votes fewer than C. Ellen Connally (the Democratic candidate running for Chief Justice of the Ohio Supreme Court). The fact that the Dem candidate for Pres received fewer votes than someone lower on the ticket suggests that votes were switched from Kerry to Bush in these counties. Using Time for Change's data, the average Kerry/Gore discrepancy in these 36 counties is -2.51% (range: -12.08 to 6.51).

The 16 counties implicated in switching votes from Connally to Moyer in Arnebeck's original suit are also a mix - ES&S (8); Triad (6); Danaher (1); Diebold (1) - (named in footnote). It is clear that Kerry got more votes in these counties than Connally, but I don't know how they calculated the number of votes they think were switched from Connally to Moyer in Bush v. Moss. Using Time for Change's data, the average Kerry/Gore discrepancy in these 16 counties is 1.94% (range: -2.4 to 12.84).

The 36 Ohio counties not implicated in fraud by the Arnebeck suit have an average Kerry/Gore discrepancy of 0.42 (range: -5.77 to 7.19).

It makes sense that in the 36 counties that Arnebeck thinks votes were obviously moved from Kerry to Bush, Kerry in 2004 would average 2.51% fewer votes than Gore in 2000. In the 16 counties in which vote skimming from Connally is more obvious, Kerry average 1.94% more votes than Gore. The difference isn't large enough, IMO, to think that the Kerry/Connally discrepancy is the best way to sort out which counties had fraud and which did not.
====================================

Using Kerry/Gore discrepancy to indicate counties to analyze further...

I think that the Kerry/Gore discrepancy is a pretty good indicator of the counties in which fraud probably occurred, and, if this is so, then Arnebeck is not fingering all of the counties in which fraud occurred. The 24 counties with the largest Kerry/Gore discrepancy include 15 that *were* listed in Bush v Moss (Auglaize, Champaign, Clermont, Clinton, Darke, Defiance, Highland, Lawrence, Logan, Madison, Mercer, Miami, Preble, Shelby, Van wert) and 9 that *were not* listed in Bush v. Moss (Belmont, Clark, Hardin, Harrison, Jackson, Marion, Paulding, Ross, Wyandot). The average discrepancy for all 24 counties is -5.12; the 15 implicated in Bush v Moss = -5.98; the 9 that are not named in Bush v. Moss = -4.34.

Conclusion: Anyone who is motivated to do precinct-by-precinct analyses might do well to start with the 24 counties with the largest Kerry/Gore discrepancy.
=======================================

Full list of counties implicated in fraud in Bush v. Moss:

The 36 counties implicated in switching votes from Kerry to Bush in Arnebeck's suit are a mix of Triad (18), ES&S (16), and Fidlar (2) counties. The Triad counties are Brown, Champaign, Clinton, Darke, Defiance, Greene, Highland, Holmes, Logan, Madison, Mercer, Miami, Pickaway, Preble, Putnam, Union, Van Wert and Warren. ES&S Counties are Adams, Allen, Auglaize, Butler, Clermont, Crawford, Delaware, Fairfield, Geauga, Hamilton, Hancock, Lawrence, Licking, Richland, Shelby, Williams. Fidlar: Medina & Morrow.

The 16 counties implicated in switching votes from Connally to Moyer in Arnebeck's original suit are also a mix ES&S (8); Triad (6); Danaher (1); Diebold (1). The six Triad counties are Ashtabula, Jefferson, Lorain, Montgomery, Portage, and Summit. ES&S is in Athens, Cuyahoga, Erie, Mahoning, Ottawa, Sandusky, Stark, and Trumbull. Danaher is in Franklin; Diebold is in Lucas. Huron is also listed in the lawsuit, but I can't tell from VerifiedVoting.org what type of voting system they have.