E-Voting: It's Security, Stupid
By Ben Rothke  in eWeek
August 23, 2004

Secure e-voting is possible but not without a sizable effort.
Last month, Harris Miller, president of the Information Technology Association of America, reportedly stated that the open-source movement is using the issue of e-voting security to wage a "religious war" that pits open-source software against proprietary software. The only thing more absurd would be for Miller to blame the woes of e-voting on a vast right-wing conspiracy. As a citizen and voter, Miller should applaud, not disparage, the whistle-blowers who have demonstrated the security flaws of e-voting systems.

Leading security professionals say they are against e-voting because of its intrinsic security weaknesses. Indeed, if e-voting were a drug, the FDA would never let it out of the lab.

Miller also stated that a recent ITAA survey showed that 77 percent of registered voters are unconcerned about the security of e-voting systems. Such a figure demonstrates how little those polled know about information security. Could any of that 77 percent find insecure code in the voting software or explain, for example, how blind signature voting systems are supposed to work? Democracy is magnificent, but the optimism of 77 percent of the populace cannot make insecure and buggy code workable. If the 77 percent truly understood the many security problems, their enthusiasm for e-voting would be quickly extinguished.

In fact, e-voting creates the largest and most unique set of challenges to information security today—greater than the security challenges of e-commerce or electronic tax filing systems. When the ITAA points a finger at the open-source movement, it is only in a futile attempt to deflect criticism of the inherent security flaws of e-voting systems.

Paper voting is not without its problems, as the presidential election in 2000 made clear. What traditional voting systems offer, however, are audit trails. Although far from perfect, paper audit trails are the best we have. While fraudulent repeat voting has long been a problem, the most fraudulent votes a single person could place in a single day might be 20. Move that election online, and there's theoretically no limit to the number of hacked votes that could be placed.

Click here to read about the growing opposition to e-voting.

Electronic audit trails, if implemented effectively, could provide an ideal solution. The only problem is that there are no vendors that are developing auditability levels that would permit secure e-voting.

Secure e-voting is not beyond our reach. It could be developed, but it might amount to the application equivalent of the Manhattan Project. Until such a commitment is made, e-voting is one technology that we would be well-advised to do without.