Paper trail for voting not a cure
Editorial   Athens Banner-Herald   13 March 2005

Georgians may be getting something more than an "I'm A Georgia Voter" sticker when they walk away from their voting machines, if proposed legislation making its way through the General Assembly earns approval of the state's lawmakers.

 Sen. Vincent Fort, D-Atlanta, and Rep. Tim Bearden, R-Villa Rica, have introduced bills in their respective chambers that would require outfitting the state's 24,000 electronic voting machines with equipment that would produce a voter-verified paper audit trail. What that would do is provide voters with a printer-generated copy of the ions they made on the machine's touch screen. Voters wouldn't be allowed to take the paper verification of their ballot choices with them; it would into a locked box.

For some of the state's citizens, who clearly don't believe the electronic voting machines are accurately counting their votes, the move to establish a paper audit trail likely will come as welcome news.

According to the latest Peach State Poll, a telephone survey of 800 people conducted from Jan. 31-Feb. 10 by the University of Georgia's Carl Vinson Institute of Government, non-white respondents were not particularly confident their votes are being counted accurately. Only 39 percent of non-white respondents were "very confident" their votes had been accurately counted, with 35 percent only "somewhat confident," 10 percent "not very confident" and 13 percent "not at all confident" of accuracy.

By contrast, 76 percent of white respondents were "very confident" of accuracy, 20 percent were "somewhat confident," only 2 percent were "not very confident," and just 1 percent were "not at all confident" of accuracy.

Whatever the reason for the wide variance in opinion across racial lines on the accuracy of electronic voting, the perception among such a significant percentage of people in the state that their votes aren't being tallied correctly clearly provides some rationale for the proposed legislation.

The perception of inaccurate counting also makes it incumbent on Secretary of State Cathy Cox, who turned back a similar legislative effort last year, to move away from outright rejection of a paper trail, particularly in light of the fact that persistent concerns remain about whether electronic election systems are tamper-proof. Earlier this year, the nonpartisan, nonprofit group Black Box Voting demonstrated its ability to hack into one such system.

But by the same token, a paper trail should not be viewed as the last word in assuring the accuracy of electronic voting. Even a brief conversation with technical support personnel in the average business office will reveal that a printed record of a transaction, even if it accurately reflects the keystrokes or machine touches made in that transaction, is not necessarily a guarantee the electronic machinery itself has recorded that transaction as intended.

In terms of casting a ballot electronically, that means that just because a voter presses the voting screen next to candidate John Doe's name, and a printout confirms that John Doe's name was ed on the ballot, neither of those things absolutely guarantees the voting machine did, in fact, record the screen touch as a vote for John Doe. There is a possibility, however remote it might be, the screen touch was recorded as a vote for another candidate or was not recorded at all.

Therefore, the only way to have absolute assurance that votes were cast as all voters intended would be to compare the actual vote count with the voter-verified paper trail.

That would obviously be an unwieldy process, but the mere fact it could be done might serve as a safeguard against the possibility of election tampering.