US Election Assistance Commission - Voluntary Voting System Guidelines Vote
EAC Home
Introduction
View Guidelines
View Comments
Glossary

View Comments

Section CommentsGeneral CommentsGlossary Comments
 
Name :   Rachel Garner
Organization :   N/A
Post Date :   9/30/2005

General Comments
Comment :  (Quoted by Rachel Garner, who submitted this comment.)

CAN VOTING MACHINES BE TAMPERED WITH THROUGH ACCESS TO
PROGRAMMERS?

The Washington Post characterized Hagel's election in 1996 as the  
biggest upset of the election season. At the time, voters did not  
know that he owned and had held key positions with the company that  
counted his votes.
But is it improper for a candidate to have ties  
with voting machine companies?

Harris examines the issue of tampering security in the upcoming  
"Black Box Voting" book. One of her sources, Dan Spillane, a former  
Senior Test Engineer for a voting machine company, believes that the  
computerized
voting machine industry is riddled with system integrity  
flaws.

"The problems are systemic," Spillane says, and he contends that the  
certification process itself cannot be trusted. Despite industry  
characterizations that the code is checked line by line, this does  
not appear to be the case. Spillane points to frequent, critical  
errors that occur in actual elections and identifies omissions in the  
testing procedures themselves. His own experience as a voting machine test  
engineer led
him to address his concerns about integrity flaws with the owner of the  
voting machine
company, who then suggested that he resign. He did not, but shortly before  
a General
Accounting Office audit, Spillane was fired, and so was his supervisor,  
who had also
expressed concerns about system integrity.

Election Technology Labs quit certifying voting machines in 1992. Its  
founder, Arnold B. Urken, says that the manufacturers, specifically  
ES&S (then AIS), refused to allow the detailed examination of code  
needed to ensure system integrity. Wyle Labs refused to test voting  
machine software after 1996; testing then went to Nichols Research,  
and then passed to PSINet, and then to Metamor, and most recently to  
Ciber.

But even if certification becomes adequate, nothing guarantees that  
machines used in actual elections use the same programming code that was  
certified.
Machines with adjusted code can be loaded onto  
delivery trucks with inside involvement of only ONE person. To make  
matters worse, "program patches" and substitutions are made in  
vote-counting programs without examination of the new codes, and  
manufacturers often e-mail technicians uncertified program "updates" which  
they install
on machines immediately before and on Election Day.

Both Sequoia touch screen machines and Diebold Accuvote machines appear to  
have
"back door" mechanisms which may allow reprogramming after votes have been  
cast.
Diebold's Accuvote machines were developed by a company founded by Bob  
Urosevich,
a CEO of Diebold Election Systems and Global Election Systems, which  
Diebold
acquired. Together with his brother Todd, he also founded ES&S, where Todd  
Urosevich
still works. ES&S and Sequoia use identical software and in their optical  
scan machines.
All three companies' machines have miscounted recent elections, sometimes  
electing the
wrong candidates in races that were not particularly close...(Reference  deleted for privacy.)