Election Security and Technology

Matt Blaze, president of the Election Integrity Foundation, opens the Voting Village's morning session with a broad introduction to election security. He describes a tension that is hard to resolve: serious technical weaknesses exist in many parts of US election equipment, yet no credible evidence has shown that any of them has changed a US election outcome. He splits the discussion into three problems, the technical one, cynical distrust fed by unfounded claims, and unwarranted optimism about internet voting, and says he is most hopeful about the first.

He reduces the technical problem to a simple question: does the certified result match the votes cast? That is hard in practice because US ballots are complex, errors cannot easily be redone, no neutral third party can be trusted, and the secret ballot must be preserved. Blaze separates visible failures, such as registration chaos on election day, from silent tally errors. He credits two ideas with making flawed software tolerable: Ron Rivest's software independence and Philip Stark's risk-limiting audits. Paper optical-scan ballots make these audits possible, while touchscreen machines do not.

He then turns to denialism that rarely specifies how an alteration would work, and to the appeal of voting from a smartphone. Voting is irreversible in a way banking is not, and remote voting breaks software independence. His takeaways for technologists are to explain clearly what elections can and cannot guarantee, and to work as poll workers.