Electronic Voting, The Basics
Among the election fallout stories, a few of us in the newsroom have been speculating about the pratfalls of electronic voting, or “direct recording electronic” (DRE) voting machines. Irregularities about the national election have been trickling though through news reports and the blogosphere. These irregularities and problems, which have been detailed elsewhere in the press and on the web, will be detailed in this space in the coming weeks.
But since we’ve published some good coverage on the issue by reporter Van Smith, we’re bumping that info here.
On Dec. 26, 2003, Smith wrote that “a year ago–indeed, 10 or 15 years ago–the only people making a hue and cry about the inherent frailties of computer voting were a handful of geeky academics obsessed with computer security. Something happened in the aftermath of the November 2002 elections, though, when computer terminals recorded more votes than ever before. The public suddenly began to understand what these experts had been saying all along: that without an independent paper record of each voter’s choices, computer-voting systems are shockingly vulnerable to undetected insider manipulation of election results.”
On Oct. 30, 2002, Smith reported on the burgeoning anti-electronic voting critics, including Bryn Mawr College professor Rebecca Mercuri and Blackboxvoting.org’s Bev Harris in The Vote Counters: Computerized Ballot-Counting Systems Under Fire. The skinny: “Unlike punch-card or optical-scan systems, in which ballots can be traced to the voters who cast them, DRE systems do not provide a full paper trail as a way to determine that the votes counted are the same as the votes intended.” If you voted in Baltimore City, you probably used a voting machine made by Sequoia Pacific Systems, which figured in a 1999 bribery scheme in Louisiana.
Future Vote: Computerized Balloting is Taking Over Elections In Maryland–But Can We Trust the Results?, 12/11/2002, emphasizes the point, raised in the previous story, that electronic voting machines’ weakness is that they leave no paper trail. Smith explores the possibility of tampering with election outcomes, recalling the way a Baltimore computer programmer employed by a electronic-wagering company tampered with horse-racing stakes.
In that article, Maryland Secretary of State John Willis defended the state’s increasing reliance on electronic voting:
“It was the intent of the commission and of the governor and state legislature to capture as much voter intent as possible as efficiently as possible. And with Diebold’s system, we are capturing more voter intent–assuming we believe the codes are right and nobody’s manipulating them.”
How does Willis know that’s not happening? “Well, one, you have to have confidence in the vendor,” he says. “And then you have testing before, during, and after the election, and the source codes are kept in escrow so they can’t be changed. And then if something looks odd in the outcome, there are experienced people around who are going to notice aberrations from historical voting patterns at the precinct level. And if it could be empirically demonstrated that something there does not make sense, then a court could order a look at the code and the outcome would be challenged with technical expertise.
“So if somebody tried to rig a computer election here, it would require a good deal of sophistication. I think we’d be able to detect it,” Willis continues. “But if they can beat all of that, how would we detect it and what would be do about it? Well, the academics are right–some of those questions are still unaddressed. We were aware of the risks, and I have no objection to raising theoretical concerns, and I can’t say that those risks aren’t there, but they were outweighed by the advantages of capturing as much voter intent as possible.”
Detailing something “odd in the outcome,” the story describes the discrepancies between primary and general elections in Georgia counties, resulting in Max Cleland’s unseating in the 2002 midterm elections:
60 percent of Georgia’s electorate live in counties which dramatically shifted partisan loyalties between the primary and general elections. Yet the final tally didn’t follow those shifts. In addition to his stable power base, Cleland won those anomalous Southern counties that shifted toward the GOP in the primary by an 18 percent margin. Unfortunately for him, Chambliss won the Northern counties that veered towards the Dems in the primary by a whopping 29 percent of the vote.
So are the huge discrepancies between exit polls in the 2004 election and the final tallies indication of fraud, faulty polling techniques or complicated shifts in voters’ opinions and loyalties? We’ll survey the developing story as the weeks go by.
A few more articles:
Ballot Check: Computerized Voting Comes Under Fire in Georgia and California, 2/19/2003. “Reputable computer-security experts have joined forces in California to insist that the systems have inherent, though simply solved, security flaws. And allegations have surfaced in Georgia, where computers were used by all voters for the first time last November, that sensitive election-software files were on a publicly accessible Internet file server–a situation, computer experts say, that would present an opportunity for code tampering to manipulate election outcomes.”
On July 30, 2003, Smith wrote about a group of Hopkins researchers who determined that the Diebold systems were full of security holes. They wrote: “Voters can trivially cast multiple ballots with no built-in traceability, administrative functions can be performed by regular voters, and the threats posed by insiders such as poll workers, software developers, and even janitors, is even greater.” See Tamper Proof.
One Dollar, One Vote, published Sept. 3, 2003, reports on the push by lobbying group Information Technology Association of America to convince municipalities around the country to switch to electronic voting.
Maryland’s nascent voting technology protest groups are featured in Nov. 26, 2003′s Protest Vote. The article notes a 1994 New Hampshire law requiring computerized voting machines have a “printed audit.”
Attempts to impose similar safeguards in Maryland were derailed. See April 4, 2004′s Paper Chase: Bill Mandating Reform for the State’s Computer Voting System Stalls in Annapolis.
And finally, the Nose debunks “myths and facts” about electronic voting the Maryland State Board of Elections provided Maryland residents in Fact-Flouting Mission July 14, 2004.