Punchcard Performance and the October 7, 2003 Recall Election

Henry E. Brady

 

The following documents on punch card performance were prepared by me regarding the October 7, 2003 Recall Election: 

  1. Initial Declaration of Henry E. Brady for Southwest Voter Registration Education Project v. Shelley — August 11, 2003
  2. Supplemental Declaration of Henry E. Brady for Southwest Voter Registration Education Project v. Shelley —August 16, 2003
  3. Memorandum to Secretary of State Shelley on Improving Voting System Performance — September 4, 2003
  4. Post-Election Memorandum on Punch card Performance — October 9, 2003
  5. Letter to Ms. Conny McCormack, Registrar of Voters for Los Angeles County, on the usefulness of recounts —October 10, 2003
  6. Detailed analysis of punch card performance in the twenty largest California counties in 1996, 2000, and 2003 —October 17, 2003

 Documents1-2 constitute my declarations for SVREP v. Shelley which eventually ended up in Federal Ninth Circuit.

Document 3 was an (unsuccessful) attempt by me to suggest a method for improving the performance of punch cards.  I was told by the Secretary of State’s staff that it was too late to make the required changes in punch cards because they had already been prepared.  I replied via e-mail that I thought it would be worth the cost to implement my proposal. 

Document 4 analyzed the performance of punch cards after the October 7, 2003 election and came to the conclusion that 176,000 voters were lost by punch card systems on the recall question. 

Document 5 responded to Ms. Conny McCormack’s invitation for me to appear at Los Angeles County’s 1% recount.  I could not do so on such short notice, and, as I explained in the letter, I did not think that a recount would tell us much about whether or not votes were properly recorded by punch card systems.  In effect, recounts amount to counting your pocket change after your wallet has been stolen.  Presumably you will obtain the same result again and again, but the fundamental problem is that your wallet is gone. 

 Document 6 provides what amounts to a “difference-in-difference” analysis of punch card performance in California’s twenty largest counties over three elections:  1996, 2000, and 2003.  This paper redoes some of the analyses I did for my declaration by including the 2003 data.  The results are stunning, clear, and hard to ignore:  punch card voting systems lose votes at a significant and very worrisome rate. 

 

http://ucdata.berkeley.edu/new_web/recall.html
Last revision: 10/21/2003
Copyright 1995 - 2003, UC DATA