University of Colorado Boulder Evidence Based Elections and Cybersecurity Paper

Elections pose unique and fascinating challenges for technology policy. Conventional cybersecurity approaches are unworkable, because elections demand high security, transparency and auditability, yet ballots must not be linkable to individual voter identities. These requirements led to a central guiding principle: the “software independence” of voting systems, combined with robust auditing of all phases of elections, from voter registration, to authentication, to tabulation of the results. The ultimate goal is “evidence-based elections” which manage to convince the losers that they lost, fair and square. We can also address current topics like the enormous challenges of accessibility, internet voting, ballot marking devices, the irrelevance of blockchains to this task, and the opportunities and pitfalls of improved voting methods (like approval voting, or multi-winner contests yielding proportional representation).

Guide:Neal McBurnett is an independent consultant in election integrity, security and data science. Since 2002, he has pioneered election auditing, working with election administrators, legislators and secretaries of state. He was a major contributor to “Principles and Best Practices for Post-Election Audits” (2008 and 2018). He conducted the first risk-limiting audit outside California, and consulted on Colorado’s Risk-Limiting Audit project, which achieved new levels of efficiency and scale.