''Most computer experts conclude that touch screen voting machines, without paper audit capability, have no future, due to security vulnerabilities.''

The campaigning and voting are over and we're poring over the primary election results. But, let's admit it: When it comes right down to it, voting is a science. A voter votes. The vote is counted. The success or failure of an election is based on whether the voter's vote is counted accurately. The statistical gold standard is that 100 percent of the votes cast are tallied correctly.

But, how can we know that all of the votes are recorded accurately? Good question. We must have some way to go back and re-examine the ballots to make sure. For example: Candidate A may have received 2,300 votes and candidate B 2,700 votes, but if the tally reflects that A received 2,501 and B received 2,549, A wins, B loses and goes home. So, if we suspect a discrepancy, how can we confirm that the candidate recorded having the most votes is the winner?

Unfortunately, with our current touchscreen DRE (direct recording electronic) voting machines, we cannot confirm the accuracy of the vote. The computers, or rather the computer programmers, not the voters, have the ultimate control. Votes are captured in a complex matrix of silicon and electrons. If you want a re-count, you can ask the computer for a re-count, but it will simply regurgitate the same, possibly faulty, information. In the example above, A wins and B loses.

A ridiculous, hypothetical situation, you say? In fact, this problem is not hypothetical. Discrepancies have occurred on a regular basis with DREs. In Sarasota, Fla., an 18,000 vote ''undercount'' caused the front-runner in Florida's 13th Congressional District to lose the election in 2006. In other words, 18,000 voters voted for other offices, but didn't vote for the most important race on the ballot. Which is more likely, that 18,000 people went to the polls to vote but did not cast a vote for congressman, or that the machines made a huge error? I think it was the machines.

I also think it's possible that Northampton County voting machines may record and report votes incorrectly. In case some of you may have missed the news, the voting machines we used in the last two elections, including November 2006, were uncertified. Even if these machines were to be certified, there would remain doubts about their accuracy because of the absence of meaningful auditing capability. Unlike paper, which cannot be easily destroyed or altered, electronic voting machines store their voting data in electronic bits, which are easily altered or lost. Once lost, there is no way to retrieve this information. Once a voter has entered his selection, there is no way to verify that the voting machine has recorded a vote for the selected candidate. The vote may be unrecorded, or worse yet, recorded for an opponent.

I am a plaintiff in a lawsuit against the Pennsylvania Department of State that requests that these unverifiable machines be re-examined. The secretary of state's response was to move to have our suit dismissed. One month ago, Commonwealth Court upheld our right to move forward with our request. However, voters still are waiting to get these machines re-examined. This could take years.

I met with Northampton County Executive John Stoffa, who is looking for answers to this serious and unprecedented problem. I suggested to Mr. Stoffa that continued efforts to improve our current DRE system are a waste of time and money. I think that DREs, because of their inherent corruptibility, will someday be declared illegal. Most computer experts conclude that touch screen voting machines, without paper audit capability, have no future, due to security vulnerabilities. There is legislation pending in Congress to eliminate DREs. Florida has already banned them.

What are our options? For the near term, optical scanning machines seem to hold the most hope, since paper ballots must be filled out and then fed into the machines. In this way, the paper ballot may be used as the official record and scanning technology is merely used to speed the process of counting the votes. Currently-available optical scanning systems are by no means fool-proof and will require major improvements in verifiability, before voters can feel confident of the results.

Computer scientists and political scientists stand on the sidelines ready to help. Let's use our best science to ensure that elections, the lifeblood of our democracy, run with the utmost accuracy, auditing capability and confidentiality that our technology will allow.

Alan S. Brau, M.D., of Hanover Township, Northampton County is a member of the Coalition for Voting Integrity in Bucks County. He was a write-in candidate for township supervisor.

ALAN S. BRAU

 

http://www.mcall.com/news/opinion/anotherview/all-right_col-a.5849915may16,0,2719339.story?coll=all-newsopinionanotherview-hed