By Rob Kall 

originally published in Bucks County Courier Times Daily Newspaper  and OpEdNews.com, moveon, the peoplesvoice.org.

The use of computerized voting software has emerged so quickly. With some states and the federal government banning punch card ballots, the explosion of computerized voting will continue at a greater pace than ever.

As an experienced software developer, it amazes me that there do not appear to be any laws that consider the ways voting software can be inappropriately manipulated. It is incredibly easy to cheat the system, for one bad apple or rogue programmer to literally steal an election, or a lot of elections.

Private ownership raises more questions about the safety of the integrity of the ballot. But even if companies were publicly held or government owned the risks of corruption are too great.

It’s easy to manipulate the underlying algorithms that do the counting and analysis within the software. It’s easy to add switches which can be remotely turned on or off, converting a program from normal functioning to cheating-- to distorting the data. It only takes one person who knows what to do. And the nefarious switches can even be set to disappear after they’ve been used.

If a voting software company did allow corruption of the vote counting process, what could we do? How would we even know, since these companies insist that we trust them and then refuse to allow full inspection and checking of all revisions and software "fixes."

If, by some miracle, illegal vote count modifications were detected, what recourse would we have? Fine the company? Pursue civil litigation for damages due to software malfunction? Most likely, a hacker would be blamed. Just think about how many hundreds of viruses are created by people who are never caught. The likelihood of identifying the actual person who did the deed would be almost impossible, and catching anyone higher up in the company who enabled the vote count corruption would be even more difficult. How can we possibly entrust our election process to such risk, to such naive trust in a system so vulnerable to criminal hijacking?

But there’s a way to keep voting software honest.

Have each person register his vote selections on the computer. Then have the computer print the selections out so they are easily readable.

The voter, after checking that the ballot matches his intended vote, personally inserts his one ballot into a reader, like the airlines now use when people board a plane. Election officials count the votes from the printed ballot. Then, they check to make sure that it matches what the computer reports.

Only then is the computer allowed to transmit its results to a central computer. Afterwards, the state or county or city prints out final counts as reported by all the computers. Local voting locations will then officially confirm that the final count matches their hand count. If there are any discrepancies, they will be publicly reported. This approach would provide the precision of computer power and the trustworthiness of joint counting by poll watchers from both parties. It would neutralize any possibility of cheating, and, at the same time, insure that people actually get their votes counted as they intend-- no more dangling chads or confused votes for the wrong candidate because of poor ballot design. The voter will actually be able to read his or her final selection before submitting the actual ballot.

Bev Harris operator of http://www.blackboxvoting.com/ and author of the book, Black Box Voting; Ballot Tampering in the 21st Century, to be released in May, offers these suggestions

"1) Retrofit current machines and require new ones to print a voter-verified paper receipt, which goes directly into a ballot box with the same protections we've always had on the ballot box.

2) Because many states have (inexplicably) passed laws that prohibit election officials from ever looking at the paper ballot -- in states like Nebraska, EVEN IN THE CASE OF A RECOUNT-- we need to revise the laws, getting rid of restrictions preventing humans from doing ballot counting. Allow ANY precinct to volunteer to do a hand-count and compare it with machine totals. Also, allow ANY citizen to pay for a hand-count (make it reasonably priced and have universities provide the service as part of a political science course). If the citizen's hand audit reveals a significant error, he does not have to pay for it, since he's done a public service.

3) Require companies to disclose errors in sales presentations and to the media.

4) Require companies to disclose ownership and key people

5) If the machines show a large error, it needs to trigger a wider audit, and if the machines elect the wrong candidate, the right candidate needs to be installed in office when it is discovered. (Sounds

obvious, but right now these steps aren't usually taken.)"

There are many ways where it is handy to replace human work with computers. But for some sacrosanct activities, like the election process, computers, working alone, are a bad idea. We can increase the accuracy and reliability of voting by using computers, but it should always be done with inclusion of real humans doing the double checking and counting. The vote is too precious to trust to computer error or the risk of intentional or accidental software corruption. Congress should pass a law banning any use of computers in voting that does not include double checking by real humans. Anything less endangers the integrity of the voting process.