Executive Summary
In October 2007, the U.S. Election Assistance Commission (EAC) began to consider a new draft of standards – the Voluntary Voting System Guidelines (VVSG) – that future voting systems will have to meet in order to earn federal certification. Once fully implemented, only voting systems that have been federally certified can be purchased using federal funds, and so the VVSG will constitute a weighty statement of principle by the federal government as to how voting systems should be designed and implemented.
The current draft is an extensive document. It contains several hundred pages of detailed requirements meant to ensure voting system usability, accessibility, and security. Many of its provisions are the product of substantial research and experience. This being said, we find the VVSG to be in need of extensive revision. In many cases it is excessively prescriptive, a characteristic we fear will stifle innovation. In other places, its provisions are based on faulty methodology, and require further testing and consideration.
Chief among our findings:
- In many instances, the VVSG prescribes design standards where performance standards would be sufficient. Excessive design standards can arrest manufacturers’ ability to innovate new, better systems.
- Although the VVSG purports to be agnostic towards election processes, several provisions do in fact bear on process. These provisions exceed the document’s mandate and could lead to burdensome requirements for administrators in the field.
- The VVSG’s “software independence” requirement is, in effect, a requirement that direct recording electronic (DRE) systems produce a paper record, a provision we fear trades tangible losses in election facility for merely perceived gains in election security.
- The methodology through which the VVSG derives standards for voting system usability is flawed, failing to incorporate standard, easily implemented procedures for research in the social sciences.
- By focusing narrowly on provisions for voting system security, the VVSG distracts attention from topics that should be of at least as much concern to election administrators and the public, such as poll worker training, voter registration practices, and chain-of-custody procedures.
A number of the concerns we raise call for extensive revision to the VVSG. While such revision may be unpalatable, we strongly believe that the passage of time will only serve to make this important task more costly and difficult.