I share Stephen Ansolabehere's belief that voter autonomy, vote equality, and decentralized authority ought to be paramount values as we consider new election technologies. I share, as well, his enthusiasm for a modular approach to the voting process, and his admiration for Brazil's revamped election system. However, Ansolabehere is more sanguine than I am about using the Internet for voting. Like the Federal Election Commission, which recently solicited public comment on a proposed framework to guide online voting product development,1 Ansolabehere assumes the Internet "is the next step." I don't think we should make that assumption until questions about the Internet's implications for vote security, voter privacy, and democratic community have been adequately answered.
Ansolabehere acknowledges the security problem, but does not bring it fully into view. Online voting poses a much greater security challenge than financial transactions: elections occur in a limited time period, an express set of individuals must be allowed one and only one vote, and the ambient stakes and passions run high. Even today, with online voting confined to experiments, the risks of electoral disruption are variegated, serious, interwoven, and rising. That's because so much election day information, from news to get-out-the-vote (GOTV) operations to governmental administration, has already migrated to the Internet. Technological threats to the integrity of an election could come from anywhere on earth, corrupt any one of these lines of communication, spread quickly across jurisdictions, and go undetected for an indeterminate time. Therefore, in addition to the security standards and tests Ansolabehere recommends, we need to improve the post-election process by which results may be questioned, investigated, and corrected. Breakdowns and deficiencies in this process hurt the nation far more than the precipitating irregularities of the statistical tie in the Florida vote-count.
By proposing to restrict online voting to kiosks and precincts, Ansolabehere eliminates many, but not all, privacy concerns. My biggest worry about the system his project points toward has to do with the storage of voting data after tabulation. Today's "voter files," a precious commodity in the campaign marketplace, contain individualized records of registration and turnout, but no record of how someone votes. Allowing online voting would increase the chance that an organization or individual could find a way to tap into the lockboxes containing voter data. While everyone would denounce the theft and sharing of this virtual political gold mine, who would know about it? Who would prevent or stop it? In other words, if voting moves online, how can individuals be assured that their ballot is, and will remain, secret?
Similarly, Ansolabehere's proscription against online voting from home or work would retard, but not halt, the degradation of elections as the supreme ritual of democratic life. Election days are the only days of the year that some people physically mingle with neighbors and encounter political activists. Going to a polling place symbolically affirms both our political independence and our civic interdependence. Once online voting begins, its speed, economy, convenience, and coolness will lend impetus to calls for more direct democracy. The vote will become even less distinguishable from an opinion poll than it already is. If instant elections are judged a success, pressure will mount to submit more official decisions to digital plebiscites. The representative and deliberative components of our political system will suffer by technical comparison (already, some deem them "undemocratic"). Alternately, if security and privacy breaches beset the online vote, democracy itself will be tarnished. I am not sure how to retrofit the election day ritual for the digital age. But that ought to be an important part of any electoral upgrade.
Lest I come across as a techno-curmudgeon, I would like to mention two aspects of the electoral process for which the Internet holds great promise: voter education and election accountability.
Voter autonomy means more than the absence of coercion and corruption. Autonomous voters are informed about the choices before them, and have had the opportunity to discuss those choices in advance with others. Here the Internet offers wonderful possibilities. It can make available simulated ballots, with links to information on each candidate or initiative option, and to relevant discussion groups. In California in 2000, calvoter.org and ballotmaker.com pioneered these online voter education services, respectively.2 Another promising voter education initiative was youth-e-vote.net, a mock election open to K-12 schools across the nation. Indeed, the notorious Gore-Nader (and, in the UK, Labour-Liberal Democrat) "vote-swapping" operations should be regarded as pre-election deliberation and bargaining enterprises, instead of actual vote-trading centers. If political elites can talk and deal in advance of a vote on legislation, there's no reason why voters can't too—so long as the actual vote remains secret, secure, accurate, and non-fungible.
Democracy is also served when citizens who suspect chicanery or error have the timely opportunity to voice their suspicions to those with the authority to correct the situation. In 2000, Florida voters posted messages about the problems they experienced before the polls closed. Again, we need a better system to handle election disputes—and here, the Internet could play an important role. Future election controversies in America will likely attract worldwide attention and the resolution process must be technologically sophisticated enough to meet the demands of both voters and international observers alike.
The early years of a new technology are exciting times, because no one knows exactly what the right (and profitable) uses are. We will progress as a democracy by recognizing that while the Internet may not be suitable for voting per se, it can enhance the electoral process before and after the main act.
Michael Cornfield is research director of the Democracy Online Project, and professor at the George Washington University Graduate School of Political Management.
Return to the forum on machine politics, with Stephen Ansolabehere and respondents.
1 For more information, go to:
2 I describe these sites in "Why California Voters Have It Best," Campaign on the Net, September 29, 2000, available at: www.wwnorton.com/e-2000/mc092900.htm