[Issued to the press and handed out on the Capitol steps in Salt Lake City UT, JUL 13th, 2004]
Four years ago, America learned there were some serious flaws in our voting system. Personally, I was embarrassed that my country--a country that prides itself as a champion of democracy and a champion of technology--was having so much trouble with such a basic thing as counting the votes.
I was one of those that jumped into the debate, determined to see the problems corrected. Four years later, we've made some progress but we still have a very messy situation on our hands. Mistakes, conflicts of interest, entrenched interests, incompetence, and outright corruption are a few of the factors that are turning the issue of voting modernization into a national scandal.
At a time when budgets are constrained everywhere, billions of dollars have been allocated to improve the voting system. Already, hundreds of millions have been squandered on technology that should never have seen the light of day.
One of the ideas for a sweeping solution to the voting problem involves using Direct Record Electronic voting machines, also known as "DREs." The voter makes their selections on a computer screen--often a touchscreen--and then selects a button that says, "Cast my ballot." But where is the ballot? It's invisible! How does it work? "You don't need to know," say proponents. "Trust us," they say.
To the election administrators, DREs seem to solve every problem. They're expensive. But they accomplish everything. They eliminate the need to print paper ballots; they can accommodate voters with special needs; they can easily handle multiple languages; they can tabulate the vote very fast. If they seem too good to be true, it's because they are too good to be true. They are absolutely untrustworthy! The concept of invisible ballots created with secret software is fundamentally flawed.
Regarding DREs, the consensus opinion of computer scientists and engineers has been well known for a long time. The Association for Computing Machinery, the oldest and largest group focused on computer issues, currently has a poll that is running 95% against DREs. Back in 2000 and 2001, almost every computer scientist asked about paperless DRE voting machines said it's a bad idea. But were they organized to stop them? No, they weren't. Why should they have to organize to stop people from buying these machines? Wasn't it obvious?
Now the computer scientists and engineers are organized against DREs, thanks to David Dill and many others. The days are numbered for the voting machines with the invisible ballots and secret software. In California, as of 2006, paperless DREs will no longer be permitted in public elections. We expect this trend to be nationwide, and we are looking to have a better solution ready to go in the not-too-distant future.
My organization, the Open Voting Consortium, is advocating public software for public elections, and we're building this software now. You may not see OVC software in use in 2004, but you are likely to see it in some jurisdictions in 2005 or 2006. You will see a system with all the advantages of the DREs, except that the voter will print out their finished ballot on the spot. The software will be free and open for public inspection and testing. Companies and governments who use Open Voting Consortium software will utilize off-the-shelf PCs and printers along with our free software so it will be very inexpensive. No need to warehouse expensive dedicated components! No secrets! A visible ballot!
I am very impressed with the decision-makers here in Utah. You have not jumped into buying voting technology that just isn't ready. I have noticed the thoughtfulness here. Your government is one of the first to join the Government Open Code Collaborative. I believe there are now eight states that have joined, and Utah is one of them. This organization of state and local governments, in collaboration with academic institutions, intends to facilitate the development and use of open source software in governments. This is a great idea and I commend your CIO Val Oveson and his staff for taking this step.
Last Friday, the Utah State Government issued a Request for Proposal, or "RFP," to consider purchasing a VOTING SYSTEM SOLUTION FOR THE STATE OF UTAH ELECTIONS OFFICE. Given the fluidity in the voting technology world, this was a great challenge. It's going to be a difficult process. Your officials took into consideration a great many variables. And yet, this RFP raises many questions.
Utah has taken the extraordinary step of centralizing the purchase decision. This move has its advantages, but it also adds gravity to the decision. Most states allow their counties to make their own voting system purchasing decisions. Your Director of Elections, Amy Naccarato, Val Oveson, and others involved in evaluating responses to the RFP have their work cut out for them. Since computerized voting systems will be included in the bids, I recommend at a minimum that they consult with computer science faculty from leading Utah research universities before drawing any conclusions.
These decisions are coming under closer scrutiny. Every day, we see more articles about what's being done to modernize the voting system--including mistakes that have been made along the way. There is likely to be more coverage on television. Don't try to hide anything! There is no excuse for hiding any part of the public process of voting, including the software code. And, whatever you do, don't hide the ballots!
Alan Dechert has been a software test engineer and application developer for the past 15 years. In 2001, Alan co-authored a voting modernization proposal for California designed as an in-depth study of the voting system, including development of reference open source voting software. In 2003, along with Dr. Douglas W. Jones (Univ of Iowa) and Dr. Arthur Keller (UC Santa Cruz), he founded the Open Voting Consortium (OVC). He currently serves as President and CEO of the OVC.