After January 1, 2006, most jurisdictions across the United States adopted the use of voting systems that enabled people with disabilities the opportunity to vote without requiring assistance. Voting accessibility is a key element of the Help America Vote Act of 2002, which requires that each polling place in a federal election must have at least one accessible system available for use.
For practical reasons (not the least of which are standards of prudent legal compliance with the guarantees of full civic participation by people with disabilities in the Americans with Disabilities Act), local government entities have an incentive to provide accessible voting whether those entities have elections coinciding with federal elections or not. Federal grant money administered by the Elections Advisory Commission helped ease the transition to accessible voting, and as a consequence, a handful of private voting equipment manufacturers enjoyed a brief windfall of government purchase agreements and maintenance contracts.
Many jurisdictions purchased a type of voting system referred to as a “direct recording electronic” system. With some variation from one manufacturer to another, these voting systems tend to be freestanding plastic cabinets on tubular aluminum legs, with a lighted display screen and a few buttons. Disposable “puff and sip” tubes allow people with motion limitations to select ballot items hands-free; audio headphones read ballot selections and ballot instructional text to people with limited sight, and the systems also provide an opportunity to review and edit ballot selections prior to voting. The votes themselves are either stored in removable solid-state memory cards or internal hard drives, and are tabulated by means of encrypted proprietary software.
This is all old news. Slightly less old news (and an issue of growing concern) is that these machines aren’t built to last forever, and the software and hardware is in need of updating and replacement from time to time, whether any additional federal funds are forthcoming or not.
A lot of people distrust electronic voting systems. In some cases, this distrust is bred of unfamiliarity – electronic voting systems don’t necessarily produce a paper ballot that the voter can see and touch. In other cases, this distrust is bred of experiences of technical errors or problems associated with tabulating votes cast on these systems. Some people fear that electronic voting is somehow by its nature more vulnerable to intentional tampering, although the actual risk of intentional fraud is no greater than with any form of voting. (If one posits a scenario involving illicit access to the mechanisms of voting, then one has already established a priori that the votes may be tampered with – this fear can be reduced to nothing more than an attitude of distrust towards the election workers themselves, independent of whatever method of voting may be used.)
As electronic voting systems were being rolled out, a group of voting rights activists and computer scientists in California formed a non-profit organization called the Open Voting Consortium. Among other things, the Open Voting Consortium advocated for transparency in the choice of software and hardware for electronic voting systems, and for the adoption of voting systems that would provide (in the parlance of voting enthusiasts) a “voter-verified paper audit trail” (“VVPAT”).
The Open Voting Consortium was influential enough to compel legislative inquiries that resulted in a the California Secretary of State study of the possibility of using open-source software and non-proprietary hardware systems for voting. The final determination was that open-source software voting systems were not ready for prime-time. Concerns about security, proof of robustness of design, funding, and an inability of non-profit groups to show any proven capacity for full-scale manufacture of voting systems all combined to shut down the idea that the State of California would mandate the use of open-source software.
The thing is, … the proprietary manufacture and sale of voting systems has been shown over and over again to be a terrible business model. It would make sense for governments to build voting systems “in-house,” with software and hardware available for public inspection and review. Voting systems, like other elements of civic infrastructure, are accoutrements of the government’s natural monopoly, and as such, should be manufactured and programmed by the government.
(Of course, the phrase”open source software” doesn’t in any way imply “non-proprietary,” and isn’t any sort of technical endorsement of the quality of the software itself. “Open source software” just means that the source code for that software is more or less available for public viewing, whether the software is used in a proprietary system or not. The advantages of open source software are generally described in terms of that software being subject to critical scrutiny and recommendations for improvement from a wider audience than is available for software where the source code is secret and not available for general viewing. )
So … what happened to the Open Voting Consortium? That group had gone so far as to enlist software engineers, had designed a mock-up voting system, had gone so far as to build demonstrations of various aspects of a voting software suite, and had at least some financial backing. Was the movement to create an open-source voting system killed by lack of sufficient capitalization and interest, or is anyone out there still working on a non-proprietary accessible device? I see that the organization touts recent legislation easing the path to system certification for use in elections, and that another group (the California Association of Voting Officials) endorses the idea of an open-source voting system.