Guest editorial: special issue on electronic voting

T HIS special issue provides representative samples of the excellent research being carried out in electronic voting today—in areas as diverse as statistics, usability, cryptography, formal methods in security and experimental computer security analysis. It is perhaps the first broadly defined special issue of a technical journal in this rapidly growing and important research area. The last few years have witnessed the large-scale deployment of electronic voting systems worldwide: both optical scan voting systems and direct recording electronic (DRE) systems. The latter have drawn the most attention because of the detection , by voting researchers, of a wide variety of flaws that could be used to manipulate election outcomes. At the same time, several new voting systems, that would not have similar flaws, have been proposed, prototyped, and used in binding elections. The possibility of having an impact on deployed systems has transformed the field of voting technology research to one that is now growing rapidly, attracting researchers of various backgrounds. In this multidisciplinary special issue, we bring together several of the most interesting new results in the area. II. THE PAPERS We begin the special issue with two papers illustrating two different approaches to improving the integrity of elections that use optical scan equipment. Antonyan et al. describe new audit procedures, and their implementation for elections in the state of Connecticut, U.S., on the request of the Office of the Secretary of the State. Chaum et al. describe Scantegrity II, which will be used by Takoma Park, MD, for its municipal election in November 2009. The paper describes the use of confirmation codes for cryptographic audits of the election outcome, which can be performed by voters and election observers, and are not restricted to privileged individuals. Fink, Sherman, and Carback describe the use of trusted platform modules (TPMs) to reduce the trusted computing base of DREs. This approach enables the early detection of threats to election integrity. Gardner, Garera, and Rubin present a new approach by which a human, such as a poll worker, can determine whether the software on an electronic voting machine has been modified, using the time taken by the computer to respond to a challenge issued by the human. If the software is changed, differences in main and cache memory access times, or CPU clock cycle times, cause the response times to be greater. Villafiorita, Weldemariam, and Tiella describe ProVotE, an electronic voting system with …