Bill:
Good points.
1. One way to deal with large jurisdictions is to break them up into smaller
constituencies. If 50,000 polling stations each count 100 ballots, for
example, you get a quick count. Spend money to train and equip talliers and
have party observers sign the local tallies at the local level. If the
polls close at 6, this can be done by 8. The results can then be shipped to
the next level up for a confirmation.
2. Yes, humans make mistakes counting things manually. Training and the
right number of polling stations reduce this problem. As Florida 2000
taught us, automated systems also fail. Ideally, we could use some
combination of both so that we maximize speed as well as transparency.
Fabrice Lehoucq
Division of Political Studies
Centro de Investigaci—n y Docencia Econ—mica (CIDE)
Carret. Mexico-Toluca 3655
Lomas de Santa Fˇ, Mexico City, DF, CP 01210
Tel. 52/5727-9800, ext. 2215 (voice) & -9871 or 9873 (fax)
E-mail: Fabrice.Lehoucq@cide.edu
-----Original Message-----
From: owner-election-law_gl@majordomo.lls.edu
[mailto:owner-election-law_gl@majordomo.lls.edu]On Behalf Of Huennekens,
Bill
Sent: Martes, 24 de Septiembre de 2002 12:59 p.m.
To: election-law@majordomo.lls.edu
Subject: RE: paper ballots
Terms must be carefully defined. Yes the Caltech/MIT study found that paper
ballots had the lowest residual vote rate, but they are talking about paper
ballots that are hand tabulated. It is administratively inconceivable that
we could expect large jurisdictions like King County, WA, Los Angeles
County, CA or Cook County IL to hand count paper ballots. Especially if you
want anything close to timely results.
I would argue that in large jurisdictions the actual tally would not be
accurate either. If an individual counts a ream of paper 3 time I bet you
will get three numbers, a machine will get it correct every time.
Bill Huennekens
Policy Analyst
Office of the Secretary of State
Washington State
Elections Division
-----Original Message-----
From: Fabrice Lehoucq [mailto:fabrice.lehoucq@cide.edu]
Sent: Tuesday, September 24, 2002 10:53 AM
To: 'Larry Levine'; 'David Schultz'; T.Round@mailbox.gu.edu.au;
smulroy@memphis.edu
Cc: election-law@majordomo.lls.edu
Subject: RE: paper ballots
Other arguments in favor of paper ballots:
1. The MIT-Caltech study showed that paper ballots had the lowest error rate
in 2000, if memory serves me.
2. Most dictators lose elections with paper ballots. It was pretty easy to
tell that Milosevic stole the election in Serbia with paper ballots. The
Mexican PRI lost with paper ballots.
Sure, some of this is experience -- e.g., people are familiar with this
technology, so they make fewer mistakes. So, arguably, with enough training
and voter education, an electronic system might work. Two countries now use
semi-electronic and computer systems: Costa Rica and Brazil. At least for
CR, I know that citizens are automatically registered and the Supreme
Tribunal of Elections spends a time and money on educating citizens
(conditions completely absent in the US).
3. A lot of the preference for computer systems stems from aggressive
companies. Yet, computer systems like so many electronic databases
emphasize speed and accesibility. They don't emphasize information
retrieval, which is key for transparency. In Bolivia last year, I asked
local officials interested in a non paper system how they could guarantee
transparency, and they had few answers.
Fabrice Lehoucq
Division of Political Studies
Centro de Investigaci—n y Docencia Econ—mica (CIDE)
Carret. Mexico-Toluca 3655
Lomas de Santa Fˇ, Mexico City, DF, CP 01210
Tel. 52/5727-9800, ext. 2215 (voice) & -9871 or 9873 (fax)
E-mail: Fabrice.Lehoucq@cide.edu