Case Against Computerized Voting Broadens

"Software flaws stunning" says researcher

by Rodger Herbst

A number of respected computer scientists, as well as attorneys, political scientists and elected officials, have opposed fully computerized voting systems (see the list of endorsers on the "Resolution on Electronic Voting" at http://verify.stanford.edu/evote.html). Their objection, as David Dill, professor of computer science at Stanford notes on his website, is that the machines "pose an unacceptable risk that errors or deliberate election-rigging will go undetected, since they do not provide a way for the voters to verify independently that the machine correctly records and counts the votes they have cast."

A close connection between the Republican party and US election systems companies has also been noted. There are potential conflicts of interest with regard to stock ownership at the company Election Systems and Software (ES&S) reported at the website www.talion.com/election-machines.html#disclosure, which also lists stockholder information for similar companies Diebold, Accenture, and Sequoia. The New Zealand Scoop website lists dozens of contributions of the Diebold Board of Directors to the Republican Party (www.scoop.co.nz/mason/stories/HL0211/S00081.htm). Bob Urosevich, CEO of Diebold Election Systems is also the founder of ES&S. Together, these two companies are responsible for tallying around 80 percent of votes cast in the US.

Still, most critics have been optimistic that it's possible to preserve the advantages of computerized voting--while still reducing the risk of accidental or malicious miscounting--by choosing optical or touch-screen systems that also produce a "voter-verifiable audit trail."

Recent discoveries by Seattle based activist and researcher Bev Harris however, have begun to cast doubt on that optimism. According to Harris, the software used by election systems is critical. Here's the process. Votes are cast at the precinct, via optical scanner or touch screen. Poll workers transmit these votes to the county office by modem. At the county office, there is a "host computer" with a program called GEMS. GEMS receives the incoming votes and stores them in a vote ledger.

Harris found that the GEMS software discovered at an unprotected Diebold FTP site makes additional copies of the original vote ledger: "On the programs we tested, the election summary (totals, county wide) come from vote ledger two instead of vote ledger one, and votes can be added and subtracted from vote ledger two, so that it may or may not match vote ledger one." The official voter report comes from vote ledger two. The detailed vote report, by precinct, comes from vote ledger one. Therefore, if the correct votes are kept in vote ledger one, a spot check of detailed precincts, even when compared with voter-verified paper ballots, will always be correct. The elections supervisor sees only the reports that are generated by the software: election summary (totals, county wide), or a detail report (totals for each precinct).

The original Diebold website was taken down on January 29th 2003, but not before copies were made by several investigative websites. Scoop has encouraged file sharing and further study by the public (www.scoop.co.nz/mason/stories/HL0307/S00065.htm).

According to the New York Times, a team lead by Aviel D. Rubin, technical director of the Information Security Institute at Johns Hopkins University analyzed the same software Harris discovered at the Diebold FTP site. Rubin noted, "We found some stunning, stunning flaws." A long list of those flaws appears in the team's final paper (avirubin.com/vote.pdf).

The study confirmed Harris's finding that ballot counts could be altered by anyone with access to the software. The researchers also found that even without direct access to the software, election results could be compromised. "The systems, in which voters are given computer-chip-bearing smart cards to operate the machines, could be tricked by anyone with $100 worth of computer equipment... practically anyone... could produce these smart cards that could allow someone to vote as many times as they like," said Adam Stubblefield, co-author of the Rubin paper.

Significantly, a Diebold spokesman, Joe Richardson, did not voice an objection to the expert team's analysis, though he did note the code was proprietary. In the past, examination of voting software has been blocked by the courts. "We're always open to anything that can improve our systems." Richardson said.

Richardson said the company could not comment on the flaws until it had seen the full report. He said that the software on the site was "about a year old" and that the code could have since been corrected. Another co-author of the Rubin paper, Tadayoshi Kohno, said it was unlikely that the company had plugged all of the holes they discovered. "There is no easy fix," Mr. Kohno said.

Douglas W. Jones, an associate professor of computer science at the University of Iowa, was shocked to discover flaws cited in the Rubin paper that he had mentioned to the system's developers about five years ago as a state elections official. "To find that such flaws have not been corrected in half a decade is awful," Professor Jones said.

Stubblefield said the election software scrutiny standards must be as high as the stakes. "This isn't the code for a vending machine," he said. "This is the code that protects our democracy." (http://nytimes.com/2003/07/24/technology/24VOTE.html)

According to verifiedvoting.org, Representative Rush Holt and over 30 co-sponsors have introduced bill HR 2239, the Voter Confidence and Increased Accessibility Act. If enacted into federal law, it would require all states to use election equipment that provides a voter-verifiable paper audit trail. A paper audit trail wouldmake it possible to reconstruct the election results from the original voter-verified records, without having to trust the election equipment.

Currently this bill is still in committee, needs more co-sponsors, and must be passed before Congress recesses in October. Washington state 3rd district Representative Brian Baird is a co-sponsor, and 6th district Norm Dicks is said to support the bill. The remaining state representatives have not taken a position. Please contact your representative and urge co-sponsorship. If congress does not pass this bill, actions at the state and county level will be necessary. Please see www.verifiedvoting.org/index.asp for more details.

What if the states and counties also fail us? One suggestion, which does not depend on any concessions from government: enter the name of the candidate of choice as a "write-in" on the electronic keyboard. This appears to be possible on Sequoia voting systems, which are used in Washington state. However, write-in candidates may have to file declarations of write-in status.

WWW Washington Free Press