Electronic Voting in Massachusetts: problems and prospects

Electronic Voting in Massachusetts: problems and prospects
A Massachusetts Conference on Electronic Technology and Voting Transparency

The 2000 Presidential elections left no one satisfied with electoral process, and there was strong sentiment for quick technical fixes. “The Caltech/MIT Voting Technology Project came into being even as the ballots from the 2000 presidential election were being counted and the Florida recount was being contested in court. The presidents of Caltech and MIT, David Baltimore and Charles Vest, convened a team of mechanical engineers, computer scientists, human factor designers, and social scientists from their respective campuses to respond to the national need for strong academic guidance at the intersection of technology and democracy ... [to solve] the problem of applying the best technology to the improvement of voting in the United States.” [1] The “Help America Vote Act” (HAVA), enacted by Congress in October 2002, gave the National Institute of Standards and Technology (NIST) a key role in helping to realize nationwide improvements in voting systems by January 2006. [2] These improvements were expected to be achieved primarily through installation of electronic voting machines, with Federal Elections Commission [3] oversight of state certification of voting machine products through Voting Systems Standards.

Since then it’s become clear that the danger of election machine fraud may be serious. The Caltech-MIT Voting Technology Project found that “manually counted paper ballots have the lowest average incidence of spoiled, uncounted, and unmarked ballots …” [4] and suggests that differences in accuracy among voting technologies are likely to be due to unspecified human factors. Concerned citizens like Bev Harris of Renton WA [5] probed ’02 “surprise upsets” in Georgia, Nebraska and elsewhere, and found a common thread of use of new “Direct Recording Electronic (DRE)” voting machines, made by companies whose CEOs are openly partisan and in one case even a candidate – whose own election was a “surprise upset” in which every vote was counted by his machines. Harris found a large body of information on Diebold Election Systems’ DRE machines on an unsecured company Internet site, open to the public, which proved to hold detailed technical information and internal correspondence indicating questionable design that might enable fraud. After dispute over access to Diebold’s proprietary software, technical details of Diebold’s DRE machines were analyzed by academic engineers and scientists, and by consultants to state election officials (Maryland [6], Ohio [7] ). Their reports about these machines “… do not inspire confidence,” as Paul Krugman wrote in a NY Times column, “Hack the Vote” (12/2/03) [8] “The details are technical, but they add up to a picture of a company that was, at the very least, extremely sloppy about security, and may have been trying to cover up product defects.” (Other vendors’ machines are not much better according to the Ohio study.) “There's nothing paranoid about suggesting that political operatives, given the opportunity, might engage in dirty tricks. Indeed, given the intensity of partisanship these days, one suspects that small dirty tricks are common.”

Worse still, technical solutions which can be understood by laypersons may be difficult or impossible to find. Many prominent experts in computer science and engineering [9] agree with David Dill [10] (Stanford) that “There are two unfixable problems with electronic voting machines: (1) No one knows how to write bug-free software. The more complex the software, the more difficult it is to find the bugs, and election software is very complex. (2) Malicious code embedded into the software could go undetected. Neither close inspection of the code nor thorough testing of the computer could ensure that malicious software has not slipped through the cracks.” Security experts Robert Rivest and Bruce Schneier agree that the vulnerabilities of voting machines may be insoluble. Others like Ted Selker (MIT Media Lab) insist that the problem is soluble, and that the real problem in the ’00 and ’02 elections was processing of voter registration rolls. The largest problem, however, is that laypersons are not able to judge the technical merits of either argument and accept the outcome of an election decided by a voting machine count, with complete confidence.

For this reason as much as the technical problems, computer scientist Rebecca Mercuri [11] recommends a “Voter Verified Balloting” scheme in which “the voting system prints a paper ballot containing the selections made on the computer. This ballot is then examined for correctness by the voter through a glass or screen, and deposited mechanically into a ballot box, eliminating the chance of accidental removal from the premises. If, for some reason, the paper does not match the intended choices on the computer, a poll worker can be shown the problem, the ballot can be voided, and another opportunity to vote provided. At the end of the election, electronic tallies produced by the machine can be used to provide preliminary results, but official certification of the election must come from the paper records.” [12] This “Mercuri Method” may reduce the role of the electronic system to a paper ballot printing device. Such a paper audit trail is required by House Bill H.R. 2239, the Voter Confidence and Increased Accessibility Act of 2003, introduced by Congressman (and former Princeton physicist) Rush Holt. It would mandate a voter-verified manual audit capacity for computerized balloting systems, prohibit the use of undisclosed software source code and wireless communication devices, and accelerate HAVA payment schedules to states.

The League of Women Voters of United States has taken the following position: “The issue of "voter-verified paper trails" and voting systems will continue to develop. The League does not have a position favoring any specific voting machine. The LWVUS does support an individual audit capacity for the purposes of recounts and authentication of elections for all voting systems, including, but not limited to, DREs. The LWVUS does not believe that an individual paper confirmation for each ballot is required to achieve those goals; in fact this is unnecessary and can be counterproductive. Federal and state standards for testing and certification of voting systems should, of course, be of the highest quality.”

This conference will examine both policy and technical issues of electronic voting and transparency of election processes, in the context of Massachusetts elections policy. It is open to the public. Conference details are incomplete at the time of writing; for complete, final information on date, location, registration, fees, program, and speakers, please visit http://www.aclu-mass.org.