Honest Voting

In Arizona, back in December 2007, Pima County Superior Judge Michael Miller ruled in favor of the Pima County Democratic Party, the party’s attorney Bill Risner, and AuditAZ, an election integrity group, in their suit to review the county’s electronic voting records for the 2006 primary and general elections.  This included copies of the Diebold GEMS databases.  (Reportedly this was the first ruling in the country to indicate that such data was a public record.)  The issue that originally concerned the activists was a 2006 regional transit bond vote that was behind in pre-election polls but won on election day.

In July 2008, Risner and the other activists reported that they’d received a written confession from a whistleblower who claimed that the chief computer programmer and operator for the Pima County Elections Division, Bryan Crane, confided to him in a bar conversation that he “fixed” the elections on orders from “his bosses” in the county.

Bill Risner wrote an article for Alternet a couple of weeks later that gives more details. Risner’s article includes not only a copy of a letter he wrote to Arizona Attorney General Terry Goddard, in which Risner asks the state to reopen its own investigation and recount the 2006 ballots, but a link to the whistleblower’s affidavit. I highly recommend reading this article, which appears to give copious details on how a state should not conduct an investigation into voting fraud. Risner points out that the investigation could easily have been skewed because county officials were involved in determining the scope and parameters of the study. It’s hard for me not to concur. If the records under review are sent to the chief election official being investigated, you’re doing it wrong. If the records subsequently disappear, that counts as a massive Fail.

Did you know you can make big bucks by disbelieving in vote-fraud? Just call 1-800-FAKEVOTE. Fake votes now, ask me how.

It makes me want to bang my head against the wall when I watch the entirely predictable, widely announced, even, Republican wave of claims that voter fraud is rampant across America. I’m reminded of the way they told me to write essays, way back in junior high school: first, tell us how you’re going to lie to us, then lie to us, and finally remind us of how you lied.

This article in the Wall Street Journal, written by Corey Dade and John D. McKinnon, says that the rising numbers of new registrants are giving Republicans the cold whibbies, especially in Virginia and Pennsylvania. To protect against successful Democratic registration drives voter fraud, Republicans are beginning to comb thousands of new registrations in those states for ineligible applicants.

I’m very glad to read, further down the page, that the Obama campaign and the Democratic Party planned ahead and already have people talking directly to local election officials to combat vote suppression.

My favorite quote appears near the end. Foley & Lardner lawyer Cleta Mitchell told the WSJ that she deplored efforts by Democrats and others to spike their anti-fraud campaign:

“What we’re not for is registering fake people at fake addresses, and creating barriers to trying to identify voter fraud where it exists, which is everywhere. It’s a growing problem, because of the professional vote-fraud denier industry.”

According to the Republican Party’s own voter fraud page, which doesn’t seem to have changed materially since I wrote about it on June 19th, they couldn’t find fraud in 30 U.S. states or territories, and of the news listings on that page, many of the articles are ongoing coverage of single cases. Just how big is that industry again? I don’t even think it can afford its own trade magazine.

You have to feel for the people in Florida, you really do. Very few people want a repeat of the 2000 election in 2008. West Palm Beach, Florida winces at the memory more than most cities might, and they’re bringing in brand-new equipment to make sure there’s no sequel: optical scanners.

Dr. Arthur Anderson, Palm Beach County’s Supervisor of Elections, told the local TV station WPTV that he’s “anxious for all 15 of the counties that are having to make this transition.” The Optech scanners that Palm Beach County will be using are products of Sequoia Voting Systems.

Reading things like the following makes me want to bang my head against the wall, though. I’m touched at Dr. Anderson’s faith in the new optical scanning system, especially under the circumstances:

Anderson said some activists are concerned that the countywide implementation of the new equipment won’t go so smoothly.

Most polling stations will only have one scanner, and none will have more than two, Anderson said.

“Numerous individuals have approached me with serious concern of whether the [voting] will slow down with long lines,” he said.

Voting delays will be kept at a minimum, even when the system goes countywide, said Charmaine Kelly, the chief deputy supervisor of elections.

Though precincts will have only have the one or two scanners, large groups of voters will be able to simultaneously prepare their ballots in privacy booths before depositing ballots in the scanners.

Nothing could possibly go wrong!

Voting Is Power, a massive voter registration drive that the Democratic party is running in Louisiana, has run into some potential problems because of complaints in certain parishes about poorly filled out or duplicate forms. When you look at registrars’ stories about the problems, though, the stories don’t add up. Any reader who’s done a compare-and-contrast exercise will notice that something’s off one way or the other. This writer, however, will come right out and say that it looks as though someone has done their best to foul up the registration drive from at least one end, and maybe both.

The registration drive, while its short-term goal was to register voters in time for a July 19th election, is ultimately aimed at beefing Democratic voter numbers for the November presidential vote. By Louisiana law the voter registrations have to be completely vetted by June 19th.

Shaila Dewan reported in Monday’s New York Times that the drive has raised complaints from some registrars about large numbers of duplicate, invalid or incomplete applications, and has led to an investigation by the Louisiana secretary of state, Jay Dardenne, a Republican.

The registrar in Jefferson, Dennis A. DiMarco, said that about 35 percent of the 4,000 cards his office had sorted were invalid because they had no address, the applicant was already registered or was a felon, or the signature did not match one on file at the Department of Motor Vehicles.* Another group of cards, he said, was missing information that the office hoped could be obtained by mail. DiMarco said he suspected that Voting Is Power canvassers were paid by the form. However, Brian Welsh, a spokesman for Louisiana Victory, the umbrella group coordinating local and national Democratic voter drives, said in a June 7th New Orleans Times-Picayune article that no, the canvassers are paid by the hour no matter how many forms they hand in.

Election officials have expressed concern that large numbers of people who believe they are registered will show up at the polls in November, only to find that they cannot vote because their application had been improperly submitted.

In Louisiana, the biggest complaints about the drive have come from Republican registrars in Caddo Parish, which includes Shreveport; East Baton Rouge Parish, which includes Baton Rouge; and Jefferson Parish, just outside New Orleans.

However, in Orleans Parish the registrar, Sandra Wilson, said she had received more than 19,000 Voting Is Power applications and had problems with only about 400 of them. There are 4,000 to 5,000 that have not yet been sorted. If the card is missing information but has a phone number, she said, “We immediately call that person and get what we need.”

Compare and contrast information from the two stories. DiMarco was already complaining about problematic registrations by June 7th; Wilson was overwhelmed but happy to get the registrations. A week later, DiMarco’s complaints have sparked an investigation from the secretary of state, a fellow Republican; also, DiMarco says that they’ll “try” to send mail to voters with incomplete forms. Wilson, however, noticed a far lower percentage of incomplete or false registrations, and says her office will simply telephone people to fill in the information gaps.

Either someone deliberately flooded DiMarco’s office with fake registrations, or they’re paying unusually close attention, shall we say, to the registrations coming into his office.

Meanwhile, eager to help, Secretary of State spokesman Jacques Berry cautioned people to register directly with government officials, not workers who approach them on the street. In other words, don’t do anything to help the Democratic voter registration drive. Oddly enough, this story has already started to show up in conservative PR outlets like WorldNetDaily. What a coincidence!

_______
*Uh, wait — who’s doing the signature matching? And let me get this straight, they *match signatures to your driver’s license*? Goodness, that doesn’t make me uncomfortable at all. My own signature varies depending on the type of writing instrument and the writing angle.

For what it’s worth, signature matching depresses voter turnout (check near the end of the linked article) — that is, it doesn’t catch fraudulent voting, it means people are less likely to register in the first place, especially among less educated or lower-income voters..

The Election Assistance Commission faces nearly impossible odds. Established in 2002 as part of the Help America Vote Act, this independent, bipartisan commission (two Republicans, two Democrats) was given the job of creating and improving election standards for all 50 states. But Congress forgot to give it any funding. At first the four commissioners had to meet in a Starbucks.

Since then Congress has ponied up some operating cash, at first $700,000, now $16.5 million. (In a June 15 story by Deborah Hastings of the Associated Press about the EAC, the agency’s current operating budget was erroneously reported at $115 million. The larger figure actually represents money appropriated by Congress for election reform projects in the states.)

The commission isn’t entirely toothless. In January 2007 the commission banned Ciber, Inc. from testing voting machines until it got its quality assurance procedures up to snuff. (Also, Ciber contributes to election campaigns, mostly to Republicans. Nope, no conflict of interest there!) The commission dinged but did not decertify SysTest Labs, another machine certification company, pointing out that it just made electronic voting look bad if a voting machine certifier accepted work from a winning political candidate whose election had been challenged because of possible voting machine irregularities.

But the EAC is, at best, forced to do too much with too little.  At worst there’s the possibility that the commission was designed to be brought out at parties and congressional inquiries so that the administration could say, “But we’re working on the problem, honest!” Either way the result has been weak standards, poorly enforced.

What an impressive acronym!

[Cut to visual of author running outdoors to vomit into the potted palms. Haven’t we all had enough of these? Good day to you, acronyms! …I said, Good Day!]

Sorry, I’m back now. These bursts of acronym fatigue attack you right out of the blue.

No matter how painfully engineered its acronym, ACCURATE nevertheless has noble goals. Created in 2005 with a $7.5 million award from the National Science Foundation, ACCURATE is part of NSF’s Computer and Information Science and Engineering directorate’s CyberTrust program, a multi-year initiative that seeks to make the nation’s underlying computer networks and infrastructures reliable even in the face of cyber attacks.* Their roster of principal investigators and advisory board members would be hard to equal — or confound, I hope.

On May 5th ACCURATE submitted public comment to the U.S. Election Assistance Commission on the commission’s draft Voluntary Voting System Guidelines (VVSG), urging the Commission to adopt certain key features for the draft. (The VVSG provides a national certification framework for U.S. voting systems against which 40 states require their voting systems to be certified.)

ACCURATE particularly applauds the requirement for software independence, which would require voting systems to be designed so that undetected flaws in the voting system software cannot cause changes in the vote count. Hard to argue against that in principle. However, when I think about critical software design I’m inescapably reminded of an old friend, a programmer who wrote software on contract under the usual conditions of pull-an-all-nighter, chug-Mountain-Dew-till-your-eyes-bug-out, do-you-want-it-right-or-do-you-want-it-Tuesday. One day somebody told him, “Hey, we used your compiler in our fighter plane navigation system!” He said thanks, but inwardly he couldn’t help thinking, “Wow, I hope they’re not using those floating-point routines.” In other words, when it comes to having results not affected by software flaws, well, you can ask… At this stage of the game, I’m willing to settle for just plugging the loopholes we already know about.

When you track back from the EAG guidelines to the Voting Systems Center page, by the way, you can almost see your face reflected in its ultra-clean design. Unfortunately, the voting system guidelines they recommend are still voluntary. That’s the bad news; the good news is that the government is funding the testing at all. There’s a lot more to be said about voting system testing done through the EAC, and I’ll say at least some of it down the line.

*First source for this information: an article in the Carlsbad Current-Argus.

…is a far different group than the people behind the Election Technology Library. For one thing, they know how to proofread their webpages. But also, the four members of this council manufacture the voting system platforms for 90% of the registered voters in the U.S.

These are the council members:

• Election Systems & Software (ES&S)
• Hart InterCivic
• Premier Election Solutions (better known as Diebold Election Systems, which recently and unsurprisingly changed its name when the Diebold parent company couldn’t find a buyer) - It’s worth noting that United Technologies is attempting a hostile takeover of Diebold and hopes to take Diebold’s technologies abroad.
• Sequoia Voting Systems
  

I’ll take a closer look at the Election Technology Council as a whole and at its individual members over the next several months.

Finland will try out electronic voting during its October 2008 elections. Only people entitled to vote in the Finnish cities of Karkkila, Kauniainen, and Vihti will get to use the pilot system, which will be available during the advance voting period of October 15 to 21 and at the regular election day polling stations on October 26. Traditional voting by paper ballot will also be allowed. (That four-day gap between advance voting and the official election day must make their television coverage of elections very interesting, if advance totals are released to the public before election day.) Here’s how Finland’s Ministry of Justice reports the system will work:

When voting electronically, the voter will, upon establishing proof of identity, receive a voting card on which his or her data are recorded. The voter inserts the card in the card reader in the voting terminal, presses the number of the candidate on the touch screen and approves it.

The secrecy of the election process is equally well protected in electronic voting as in traditional ballot voting. The vote and the social security number of the voter are registered in the electronic ballot box of the Ministry of Justice via the voting terminal in such a way as to make it impossible to combine the two.

The core elements of the electronic voting system are based on the Linux operative system. The voting system is supplied to the Ministry of Justice by TietoEnator and the data security solution is based on a product of the company Scytl . In the spring of 2008 the Department of Mathematics at the University of Turku will perform an independent assessment of electronic voting after which possibly necessary amendments will be made to the system.

Following up on that last sentence:

The Ministry of Justice of Finland has successfully completed the acceptance testing process for Scytl’s Pnyx, which will be implemented in the Finnish local elections in October 2008, representing the first e-voting experience in Finland. The acceptance testing was witnessed by eleven representatives from the following municipalities: Helsinki, Espoo, Karkkila, Vihti, Kauniainen and Tampere. Scytl has partnered for this project with TietoEnator, the largest system integrator in Finland.

I’d feel a lot happier about this if ES&S weren’t Scytl’s partner in the Finnish election tryouts. This is the ES&S I know.

Meanwhile, Electronic Frontier Finland (Effi) , a Finnish association for promoting digital rights, sent a request for information to the Finnish Ministry of Justice back in February . The ministry’s response was that documentation could not be provided because it would violate the vendors’ trade secrets. Doesn’t that sound familiar? [Note: there’s no permalink to this story on the home site; look for the February 7, 2008 bulletin in the Legal News column on the right.]

Effi, therefore, has only analyzed the system as far as high-level documents from the Ministry of Justice and a U.S. patent held by Scytl will permit it, according to Legislationline. Effi expects that the system will not utilize a voter-verified paper ballot system or an electronic receipt system.

In traditional Finnish elections, manual vote counting, mandatory repeat counting, and publication of the results are overseen by representatives of the competing political parties at each polling station. Ah, but electronic voting is so…so shiny!