Policy Mechanisms for Increasing Transparency in Electronic Voting
In the early years of the American republic, only white male landowners could vote, and then typically by expressing their preferences in a public setting, for all to witness. Our electoral system has changed drastically since that time; now almost all Americans cast votes with the assistance of computerized equipment. While much good stems from the use of computerized equipment in elections—notably increased efficiency, enfranchisement and flexibility—unintended consequences of this mechanization have left us with complicated, insecure and opaque voting systems.
My PhD thesis focuses on the issue of transparency in evoting; that is, what public policy mechanisms can serve to make our voting systems less opaque? After exploring what we mean by “electoral transparency”, I examine the question of evoting transparency on three fronts. I analyze the role of disclosed and open source software in election systems and conclude that, while fully disclosed source code is a valid goal, limited disclosure to experts serves many of the same goals in the short term while preserving vendor trade secrecy. I investigate how contractual provisions between local election jurisdictions and voting system vendors serve to frustrate transparency and find that election officials need to be more careful in these negotiations. Finally, I turn to the question of auditing black box elections systems; that is, since we cannot know how these systems work in the full disclosure (“white box”) case, possibly because of contractual provisions that limit investigation, what methods and procedures can we use for “checking the math” behind our elections?