Voting Machines Undergo Final Tests Ahead of Special Election

Touch-screen voting machines

There's a row of nondescript warehouses along V Street NE just off of Bladensburg Road where the District's electoral machinery resides. No, really -- every voting machine used in city's elections is here.

On any normal day, the place would likely be locked up and empty. But today employees from the D.C. Board of Elections and Ethics were in the process of running the final tests on the touch-screen and optical scan voting machines that will be used in the April 26 Special Election. The logic and accuracy tests ensure that each and every one of the machines that go to the 143 precincts -- each precinct gets a touch-screen and an optical scan -- operate as they're supposed to. In short, that they record the votes that District residents cast.

Paul E. Stenbjorn, the board's Director of Information services, walked me through the process, which kicked off on Wednesday and will culminate with a public testing of both types of voting machines next Saturday at the board's headquarters in Judiciary Square.

Optical scan voting machines

Each touch-screen machine is programmed to test itself, casting a vote for each and everyone of the candidates appearing on the three different ballot combinations that will be used. (There will be a standard citywide ballot for At-Large candidates, and two seperate ballots for Ward 4 and Ward 8 that include State Board of Education candidates.) Those results are compared against the paper receipt that the machine prints every time a vote is tabulated to make sure they're consistent with each other.

The process is much the same for the optical scan machines, but instead of the machine testing itself, a number of test paper ballots are fed into the machines to ensure that they're read properly. (There are enough ballots to test for each candidate getting a vote, for overvotes and for undervotes.)

After the testing is complete -- all of which is open to the public, though there hasn't been much interest this time around -- all the machines are sealed and readied to be shipped off to the precincts where they will be used.

Once voting wraps up on Election Day, each machine is closed down and the cartridges holding the vote totals are physically driven to the board, where the actual vote tabulation happens. (Stenbjorn insisted that while the September 2010 primary saw delays in reporting due to technical issues, the results from the November general were only delayed because some of the drivers transporting the votes back to the board got stuck in traffic.) Both the touch-screen and optical scan machines have secondary means to count votes, and those are stored and used for random audits after the election to make sure everything worked as expected.

I asked Stenbjorn how voters reacted to each type of machine, touch-screen and optical scan. He said he thought most voters felt more confident in the optical scan, and though the technology is older, he pointed out that it offered a distinct advantage -- multiple people could fill out paper ballots at once, while only one person is able to use the touch-screen machines at a time.

I also asked why it is that even though touch-screen machines are in use, there's no way to communicate results over a secure network for faster tabulation. He answered that while Estonia has developed a secure and reliable system on online voting, neither privacy advocates nor election specialists here are particularly fond of the idea. First off, there's the whole issue of the security of the network -- remember when the board ran a pilot program for online voting and how easily that was hacked into? Second, auditing bytes and bits is harder than electronic media and paper receipts, he said.

The machines will be loaded on trucks and delivered to precincts on April 25, ready for voters the next day.