INTERNET VOTING IS NO "MAGIC BALLOT," DISTINGUISHED COMMITTEE REPORTS
Panel calls for further study of security and societal issues

Trials should proceed in which Internet terminals are used at traditional polling places, but remote voting from home or the workplace is not viable in the near future. So says a new report, commissioned by the National Science Foundation (NSF), in which a committee of experts calls for further research into complex security and reliability obstacles that for now impede the Internet's use in public elections.

In December 1999, the White House directed NSF to lead a study of Internet voting. With a grant from NSF, the Internet Policy Institute (IPI, a nonprofit, nonpartisan institute) and the University of Maryland organized an October 2000 workshop whose results are summarized in the report available at http://www.internetpolicy.org/.

Internet voting systems fall into three categories: poll site voting (in which traditional election locations are augmented with Internet technology), kiosk voting (in which Internet terminals would be placed for convenience at non-traditional sites such as malls), and remote voting (in which citizens could vote from virtually any Internet terminal, including at home or work). Remote voting holds the greatest promise of convenience and universal access, but it also poses substantial security issues in addition to other risks, according to the report.

"E-voting requires a much greater level of security than e commerce -- it's not like buying a book over the Internet," said University of Maryland president C.D. Mote, Jr., who chaired the committee. "Remote Internet voting technology will not be able to meet this standard for years to come."

Prior to November 7, interest in online elections centered on the potential convenience of voting at home, according to the report, but public interest now tends to focus on reliability. The authors -- a diverse group of political scientists, computer scientists, election officials, industry experts and others -note that the 2000 elections demonstrated the "critical importance of ensuring confidence in the integrity and fairness of election systems." The report makes clear that Internet voting is not a cure-all for problems with currently used voting technology.

With federal, state and local officials considering new technology to overcome shortcomings exposed by the 2000 elections, the report urges them to resist pressures to embrace remote Internet voting systems as the technological cure. "The security problems that could arise might well undermine the legitimacy of the electoral process," said David Cheney of IPI. "We must dispel the myths associated with Internet voting and educate public officials to avoid this scenario."

The committee's main findings about feasibility are:

• Poll site Internet voting systems offer some benefits and could be responsibly deployed within the next several election cycles. Poll site voting could add convenience and efficiency in the short term, the report says, while adding speed and certainty to the tallying process. Election officials would control both the voting platform and the physical environment, making security more manageable than with the other two methods. Because these issues could likely be solved with existing technology, the committee recommends poll site experiments "to gain valuable experience prior to full-scale implementation."
• The next step beyond poll site voting would be to deploy kiosk voting terminals in non-traditional public voting sites. According to the report, many issues related to kiosk voting still need to be resolved, such as authenticating a voter's identity and preventing on-site coercion of voters. The committee states that, although kiosk voting would be more challenging than poll site systems, "most of the challenges could, at least in principle, be resolved with extensions of current technology."
• Remote Internet voting systems pose significant risk and should not be used in public elections until substantial technical and social science issues are addressed. Although remote Internet voting could maximize convenience - including better access for people with disabilities -- the security problems cannot be resolved using even the most sophisticated technology today, according to the report. The authors state that public officials must educate themselves about the dangers and ramifications of remote Internet voting.
• Internet-based voter registration poses significant risk to the integrity of the voting process, and should not be implemented for the foreseeable future. The report says online registration would have to rely on unique biometric input (fingerprint, retinal scan, etc.) to verify a voter's identity and avoid the "high risk for automated fraud (i.e. the potential undetected registration of large numbers of phony voters)." The voter registration process is already one of the weakest links in our electoral process, according to the committee, so attempts to implement Internet-based registration "without first addressing the considerable flaws in our current system would only serve to greatly exacerbate the risks."

The report cites three broad areas for further research, which NSF will help fund through its existing Digital Government program:

• the economics, design, certification and policies of poll site Internet voting
• the technical factors of security, encryption and authentication of using kiosks and remote voting
• the political science issues of how poll site and remote Internet voting would affect participation, the character of elections and democracy itself