by Alan Dechert and David Mertz, Ph.D.
We citizens often see billion-dollar government contracts turn on a single vote in a city council. Lives, careers, and large amounts of money are routinely at stake in local elections. At State and Federal levels, the stakes are even higher. The future of our country and democracy itself is continually at risk. With so much at stake, how likely is it that cheating will occur?
Throughout American history, voting systems have evolved  as measures were taken to thwart cheaters. This has been an uneven process, sometimes utilizing inappropriate technologies. Punch cards, for example, were an aging holdover of an earlier computer age when they were widely used for voting in 2000.
The 2000 election fiasco spurred large-scale Federal efforts to modernize the voting system. The Help America Vote Act of 2002 (HAVA) made billions in Federal money available to states for voting modernization. But will new voting systems be less susceptible to failures and fraud?
The measures being taken now are not aimed at making our votes more secure. The billion-dollar word in voting technology is "accessibility."
Despite their stunning pitfalls, Direct Record Electronic (DRE) voting machines can improve accessibility for people with disabilities. Computers can improve voting rights for people with disabilities by means of various assistive devices (headphones, for example). Since DREs are basically re-packaged PCs, many people who would otherwise need assistance -- and thereby sacrifice privacy -- can vote privately and unassisted on these machines.
Disabled rights activists have won this argument. They have demanded accessible voting machines, and they will--and should--get them. Like wheelchair ramps at the county courthouses, it may be a small percentage of visitors who need the ramp, but it will be a permanent fixture.
Why not make these custom accessible systems available to people who need them, but then let everyone else vote on less expensive systems? The problem with such a dual system is that it is not compatible with the need for a fully secret ballot. What if I am the only voter in a precinct to require one of these machines, or one of just a few? Anyone with the ability to see the votes cast on the accessible voting machine would know how I voted, or at least be able to make an unfairly accurate guess.
What we need is an accessible voting machine that everyone can use. DRE makers, and their shared-interest supporters, have claimed that paperless voting is the way to go. They argue that adding any voter verifiable record of vote would only complicate matters unnecessarily. Secondarily, proponents claim verifiability would discriminate against reading-impaired voters since they cannot verify a vote on paper.
DREs can have some real advantages, such as eliminating the need for a pre-printed ballot and making it easier to handle multiple languages. However, DREs are expensive, contrary to the goals of an open and transparent election process, and practically an embossed invitation for fraud and tampering. Invisible ballots created with secret software may appeal to election officials and vendors, but should not appeal to well-informed voters. Computer scientists and engineers overwhelmingly agree that DREs present unacceptable security risks.
In October of 2003, the Open Voting Consortium (OVC) demonstrated technology to allow reading impaired and blind voters to confidentially verify their paper ballot. A barcode printed on the edge of a ballot enables voters to listen to vote selections read back at a vote verification station outfitted with a barcode scanner. A privacy folder--simply an envelope cut so as to display only the barcode--is provided to voters to keep their choices secret. In April of 2004, OVC demonstrated  our prototype voting system to public acclaim .
The OVC voting machine (technically a ballot marking device) is similar to DREs in that selections can be made from a computer screen (or by listening to voice prompts through headphones), and has all the other advantages of DREs. The main difference is the printer--one for each voting booth. Unlike a DRE, the printout is the authentic ballot: Electronic records are used as an audit trail.
In prior instances where printers have been added to DREs, the paper is treated as only part of the audit trail. The vote is still electronic. The value of a paper trail depends on procedures surrounding its use. In Nevada, printers were added to the DREs for the 2004 general election but the printout was never used, but was stored "just in case." This is better than nothing, but we deserve better still.
OVC is seeking to establish a new paradigm for voting systems. We have updated and expanded on the definition of open voting systems put forth by New York University mathematics professor Irwin Mann in 1993 .
We consider an open voting system to be "a voting system in which every component is available to the general public under non-restrictive licensing terms or is in the public domain. For hardware components an Open Voting System relies on commercial off-the-shelf (COTS) devices." This means that every machine used in the voting process is programmed with code that anyone can freely download from the Internet, and every voter can examine how his or her vote will be counted. Furthermore, in the words of professor Mann, "there are institutionalized protocols for public monitoring of all components and [of] the electoral process, sufficient to find any hypothetical discrepancy from the intended design, if it should happen to exist."
We propose that OVC should maintain standards for open voting systems and become a voluntary membership-based consortium of businesses that deliver, service, and provide training and technical support for jurisdictions using open voting systems. Once established, the voting system will require minimal government oversight, while enabling maximal public scrutiny of the whole process.
The San Jose Mercury News wrote, "Open Voting Consortium appears to have what it takes to inspire faith in electronic voting. Its system can't come to market soon enough." If we want a corporate controlled voting system run in secrecy, we don't need to lift a finger: That's what we have, and that's what we'll get by default. If we want an open voting system, we have to do something. In order for open voting to become a reality, we need your support and the support of the general public to gain the necessary institutional support. Visit www.OpenVoting.org to find out how you can help bring open voting to your state or county.
 From Professor Douglas W. Jones, OVC co-founder,
 See http://www.dalelane.co.uk/cache/ovc_news4.htm
 See http://www.kentucky.com/mld/mercurynews/news/opinion/8383100.htm
 See http://www.oreillynet.com/pub/a/policy/2004/04/26/ovc.html