I’ve been meaning for some time to post a link to these reports produced by the University of Alaska-Anchorage: “State of Alaska Election Security Project Phase 1 Report;” and the “State of Alaska Election Security Project Phase 2 Report.” The Phase 1 report was published in late December 2007, and I was a peer reviewer of that report. The Phase 2 report was published in May 2008.
The reason I say that this work is a must-read is that it is more than just the typical security review of voting technologies; the Alaska-Anchorage team took a broader perspective, and argued strongly for studying voting systems in their policy and procedural context. Here is a long quote from the Phase 1 report:
The report is an overview-level evaluation of recent studies and a determination of their relevance to Alaska’s systems, technologies, and procedures. It is the first part of a multi-phase project to evaluate the security of Alaska’s election system.(footnote deleted). The research includes a detailed study of the evaluation reports from California, Florida, Maryland, Ohio and Connecticut, a review of the response to the California study by equipment provider Premier Election Solutions (formerly Diebold) , and an overview of Alaska’s election laws and procedures. The team conducted interviews with members of the University of California and Florida State University evaluation teams, with officials from the State of Alaska Division of Elections, and with election officials from California, Florida and Connecticut. As a result of this Phase 1 work, the research team identified areas within the election system and equipment used that warrant further analysis necessary to develop more definitive conclusions.
California, Florida, Connecticut, Maryland and Ohio conducted their studies using equipment that is also used in Alaska. These evaluations found serious technical vulnerabilities in the systems studied. Most reports also point out that procedures have the potential to either mitigate or exacerbate vulnerabilities
reported at the equipment level. Many of these items have been proactively flagged by the officials in the Division of Elections. As appropriate, they have implemented measures and identified possible approaches to address some of these vulnerabilities.
Each state in the US can adopt processes and procedures to meet their unique requirements. They can also select from a range of vendors and equipment provided federal certification standards have been met. Given the wide range of implementations, it is critically important that all system and procedural issues be investigated carefully in the context of policies and procedures in place in the state in which they are evaluated. In Alaska, this approach is essential to determine what, if any, impacts these issues have on the security of elections in Alaska. The use of paper ballots as the primary record of votes, the procedures for hand recounts and the uniform practices across the state may reduce vulnerabilities in Alaska elections.
The Phase 2 report implements this approach, and issues a long list of recommendations that are well-worth reading.