WORKSHOP ON DEVELOPING A RESEARCH AGENDA FOR ELECTRONIC
VOTING TECHNOLOGIES
September 17-18, 2004
Convened by the American Association for the Advancement of Science
Main | Program | Participants | Synopses
Dan Wallach
STATEMENT ON RESEARCH CHALLENGES
Without exaggeration, voting systems are one of the pillars of our
democracy. Voting systems allow the electorate to determine the course
taken by our nation. As a result, voting systems face a wide variety of
requirements and constraints. Voting systems must be secure against
tampering, yet they must be easy to use for all voters. They must
satisfy a variety of state and national standards, yet they must be
affordable to purchase and maintain. They must help voters to correctly
indicate their voting intent, even when the voter intends not to cast a
vote! They must preserve a voter's privacy and anonymity, to reduce
risks of voter coercion and bribery, yet they must be sufficiently
auditable and transparent to allow for mistakes and errors to be
identified and reconciled. They must be robust against corruption and
malice among system developers and the officials who run the election,
yet the systems must be safe enough to leave unattended in a school
cafeteria overnight.
Engineering voting systems to satisfy these often contradictory
constraints is a difficult problem, requiring research into the full
gamut of the problem, from the software and hardware design through the
careful consideration of legal and administrative procedures. Human
factors issues must be considered to make the voting systems accessible
to all eligible voters, regardless of disability. Likewise, the system
must be comprehensible to poll workers and transparent to election
observers. Ultimately, the election system is responsible not for
naming the winner of a race, but for convincing the loser that he or
she, indeed, lost the election. We must investigate software
architectures, tamper-resistant hardware, and cryptographic protocols.
We must look at the role paper should play in electronic voting
systems. We must examine system usability and study how public policy
and administrative procedure can better safeguard the system. Only by
considering all possible aspects of these systems can we have any
assurance, at the end of the day, that our elections will be fair and
that the will of the electorate will be correctly reported.
Our most basic question is: how can we responsibly employ computer
systems for tasks that require high levels of trustworthiness, when we
know that those systems will not be totally reliable, bug-free, or
totally secure, particularly when every human participant from the
system designers to the end users is a potential adversary and when
human errors are commonplace? Solving this problem requires thinking
about the end-to-end behavior of a whole system, including software,
hardware, procedures, law, and people. Perhaps more importantly, the
research problem requires people from different areas of computer
science, public policy, and human factors to combine their efforts in
new and innovative ways.