A detailed forensic analysis and recommendations for Rhode Island\u27s present and future voting systems

In 1997 Rhode Island moved from the mechanical lever machine to a more technically advanced optical scan precinct count voting system. In 2006 after several close races, Rhode Island\u27s Supreme Court ruled that Rhode Island elections rely too heavily on voting technology. It is no secret that concerns about election systems are on the rise around the country. Each election, voters rely on machines from proprietary vendors to carry out democracy and officials worry about Another Florida. We cast our vote and walk away with no evidence that our vote has become part of the official tally. In the event of an election failure, the only extant recourse is a total recount of paper ballots if they exist. Recounts are not only discouraged but by law are only conducted in a race with a very slim margin. This dissertation represents the first technical initiative to address Rhode Island\u27s need for further technical understanding of the voting equipment in which we entrust our elections. We seek to increase the awareness of forensic techniques within our election technology community and to show how their inclusion in the election process can improve voter confidence as well as security and reliability. This work reviews and critiques the emerging election technology in the forefront today and analyzes its ability to work here in Rhode Island. We provide an analysis of our current system and offer technical advice on the future technologies we may consider. We introduce a new election algorithm, the WAVERI algorithm (Watch, Audit, Verify Elections for Rhode Island). Using set theory and forensic techniques, we prove it is possible to add an audit trail to our current system with little impact to the way our citizens vote. Working closely with the Rhode Island Board of Elections, we borrowed two election machines to hold a mock election. We then compared that election to an election on the WAVERI prototype to show how an audit trail can be realized. Finally, a new metric, the Election Forensic Metric, is introduced to measure how well an election process protects itself and allows for comprehensive audits pre- and post-election