analysis of touchscreen fraud in California 2004

[garybeck]

email received form David Bayer:

There were 210,000 less votes for Kerry than Boxer in the 2004 Nov Election which suggests that the electronic machines played a role:

I have not finished with the California data but it would seem that the six Touch Screen counties accounted for the 210,000 excess Bush votes relative to the Senatorial candidates: Kerry got 210,000 less votes than Democratic Senator Barbara Boxer. These votes were most likely transferred to Bush who got 954,000 more votes than Republican Senatorial candidate, Bill Jones. We call this "spread analysis". (Bayer email 25 December 2004).

The goal of the vote transfer by electronic machines was to inflate the popular vote for Bush.

See the attached table and comments by Kris.


A spread of 50,000... 100,000 but 210,000? Something was done to those machines!

A spread of 50,000... 100,000 but 954,000? The difference is too great!

[Chi]

1. Would be interesting to see these numbers compared to...

non touchscreen county numbers.

Mandate my ass 8)

[OnTheOtherHand]

2. shouldn't we consider some context?

Via RealClearPolitics (AFAIK they report the poll results
right, at least):

SurveyUSA   10/23-25  Boxer +15  Kerry  +9  difference  +6
Field Poll  10/21-27  Boxer +19  Kerry  +7  difference +12
SurveyUSA   10/29-31  Boxer +21  Kerry +11  difference +10

official return       Boxer +20  Kerry +10  difference +10

polls from
http://www.realclearpolitics.com/Presidential_04/ca_polls.html

That doesn't rule out touch screen fraud, but it sure doesn't
give me a strong suspicion that votes were stolen from Kerry
relative to Boxer.

(EDIT: used plain text)

[garybeck]

3. can you explain those numbers?

what does 10/23-25 mean?

thanks

[OnTheOtherHand]

4. oh, sorry, those are the field dates of the surveys

I've summarized the last few polls in California, at least according to RealClearPolitics. (By the way, there is a tradeoff there; generally, a survey conducted over more days _may_ give the pollster a chance to do more callbacks and reduce possible sampling bias, but if opinion is shifting over the course of the survey, the pollster may miss it. That's just deep background, it shouldn't matter here -- but it's one of the reasons why people are in the habit of reporting field dates.)

You can follow the link and look further back -- I think there were some older polls that showed Kerry further ahead.

CA wasn't polled as heavily as most battleground states were.