Web Letters: Noted.

Electronic voting is distrusted for many reasons, but the principal problem is that the counting process is not transparent. Nobody can prove whether it is working correctly or not. There is no evidence base offered to the public.

New legislation is being proposed that still does not address this problem. It avoids it by proposing paper-based systems that can be used in the last resort in cases where the margin of victory is small.

All computer systems use counting systems that are not in themselves transparent. To overcome this, these systems provide information to demonstrate that they are working correctly by showing that there is an evident relationship between the input and the output.

Electronic voting does not provide this now or in the proposed new legislation. There is no reason why it should not. The remedy lies in additional software. Information technology is highly flexible and can be made to do almost anything. It strength is that it certainly can provide good information about what and how it derives its results.

It certainly could provide a feedback system to demonstrate to the voter that his vote was registered and counted correctly. This is the principal advantage of using electronic systems, since they have the potential to do this.

Because the secrecy of the vote must be preserved, there has to be a way in which the voter can inspect his vote as tallied to support the result without revealing his vote to others. This can be accomplished simply by giving voters secret PINs or confirmation numbers before or when they vote. These secret numbers can then be used to inspect the public record showing all the votes counted to substantiate the result.

Clearly there also must be a public record to inspect, and to show in detail how the result was derived. Further, it should be available electronically or in the press and in each district.

This public accountability also achieves the objective of verifying to the voter that his vote was registered and counted. This degree of transparency achieves another objective, that of verification to a degree unmatched in any other current election method or proposal.

This level of accountability and transparency discourages fraud because any falsification of the record is self-evident. The first and most important principal of security is achieved through the availability to universal inspection and random scrutiny of the complete and transparent record.

Providing the evidence that any result is based on the input is the first principal and common feature of every successful computer application. Why it is not a feature of the current electronic voting system nor a requirement in any new proposal is beyond belief.

Even more difficult to understand is that it is not even part of the discussion.

Many pieces of legislation come to Congress as compromises. These compromises often compromise the effectiveness of the legislation. Even more so in questions of system design, compromises undermine the outcome. Failure to include a feedback system as suggested above will invalidate the effectiveness of Congressman Holt’s new bill and will provide the loophole of continued secret vote counting. When challenged on this point, his office said the legislation was going forward as is, and that they knew that further legislation would be required. System design, on the other hand, is a discipline where any staged implementation still requires the system to be coherent as a whole and does not permit a piecemeal approach from a compromised beginning.

Unfortunately, this legislation provides enough encouragement for the hand-counted paper-ballot enthusiasts because it promotes the idea of a physical document that can be turned to in the last resort if the margin of victory is small. It also pleases the scan-data lobby because they can see business for them. It has successfully divided the election-integrity community, especially as this plays to those who distrust electronic systems. Ironically, this system still relies on electronic systems and secret vote-counting conducted by a handful of private sector companies. No requirement is placed on them to substantiate or demonstrate conclusively the derivation of the result.

The new proposals will cost $1 billion without alleviating this problem. Worse still, another four years will pass before it becomes apparent that the reform failed to achieve very much.

The feedback system proposed here is based on sound system-design principles and requires some software additions, which can be developed at far less expense and without the cost of specific hardware. Further, it is strategically sound to steer any new investment in electronic voting to general-purpose software technology that can be used on equipment of all sorts--the PC, Blackberry and iPhone--which will give greater access to more people wherever they are, and relieve pressure at the polling stations.

It was a mistake of the previous administration to steer this application into the technological cul-de-sac of special-purpose equipment and a handful of crony supplier. This made it a security exposure in itself. Also, special-purpose equipment is always more expensive and always less secure, especially when the frequency of use is at best once a year. Using technology that is hardened by frequent general-purpose use and supported by mainstream research dollars thus ensures greater overall security and design options.

Congressman Holt’s Bill repeats this mistake with a new set of suppliers, which is why it has this $1 billion price tag.

For this and any other proposal, there is a need for federal standards and a federal authority that can set requirements for systems that are to be used in the next presidential election. No adequate federal authority exists, and the execution of what little authority there is is dissipated to the states and even the counties.

Rather than throwing more dollars and another ill-conceived idea at this problem, Congress would be better advised to begin by setting up a small competent professional computer systems group to manage this challenge and give them the authority to develop a sound solution.

peter oliver

Neptune, NJ