A Note from the Editor:

I include some ten introductory paragraphs to give the Voting Train Wreck editorial context.

Trust in Computer Networks

I have come to the conclusion that it is necessary to end this Executive Summary with a discussion of Trust and trust issues in electronic or Internet based voting. When I began this exploration, early December, I did not plan to include any discussion with Ed Gerck on trust. He was in my private mail list only because for the first time I offered access to the list to all our subscribers as a benefit of being a subscriber. About one third of our subscribers took me up on my offer.

A long discussion on trust ensued. I have published the results on pages 92 through 103 above. It ended with the following commentary on use of a paper ballot in voting machines. Although it goes a bit outside the scope of this issue on the real time corporation supply chains on RF-ID, I include it here and have gone so far as to editorialize on it because it is quite possible that the failure in public policy that we are experiencing will push us into future electoral disasters and goes to the heart of the value or lack of value of computer technology in other fundamental activities in our society. Read on.

Trust Requires Corroboration by Independent Channels

Gerck: To solve this question in electronic voting, some advocate printing a paper copy of the ballot, which the voter can see and verify that it is identical to the ballot she intended to cast, and then sending the paper copy to ballot box A while an electronic copy of that same ballot is sent to ballot box B. The idea is that ballot box B could be tallied quickly while ballot box A would be used as a physical proof for a manual recount. Such a suggestion is oftentimes advanced as the sine qua non solution to voting reliability in electronic voting.

But what makes the introduction of a paper ballot special is not the fact that it is paper instead of bits. It is the fact that the voter is actually casting his vote twice. We now have two independent channels of information for the ballot, one from the terminal as source B, the other one from the printer as source A. So we have N = 2.

In other words, this design provides for two outputs: ballot A and ballot B. However, in the event of a discrepancy between the two, no resolution is possible from within the system. Technology provides no answer in this model. Thus, between two conflicting outputs, each of which is the result of a trusted system, the decision of which “is correct” must be made independently of the system, by policy.

This means that such a paper/electronic system does not work when it is needed most ˆ i.e., when the system reveals that it is not a “perfect clerk.”

The situation can be corrected with a better model, as I have discussed elsewhere. The solution considers not only machine-machine communication channels but also human-machine communication channels because the voter can act as a source and as verifier in more than one part of the system. Further, human-human communication channels must also be considered because we do not want machines to have the potential to run amok, unchecked.

Information Theory can be used to describe such communication channels and, as previously noted, the concept of qualified reliance on information can be introduced as a formal definition of trust in order to rate such channels in terms of providing effective proofs when all operational factors (collusion, hackers, buffer overflow, bugs, etc.) are taken into account.

As a result, the only provable solution to increase reliability in communications (e.g., the communication between the voter as a sender and the ballot box as a receiver) turns out to be to increase the number/capacity of independent channels until the probability of error is as close to zero as desired (direct application of Shannon’s Tenth Theorem in Information Theory). To be complete, the solution considers not only machine-machine communication channels but also human-machine and human-human. Thus, if an electronic system is able to provide N proofs (human and machine based), these N proofs for some value of N larger than two will become more reliable than one so-called “physical proof” even if this one proof is engraved in gold or printed on paper.

I note that the same applies to other communication protocols, not just voting. Public voting is, however, one of the hardest -- because it needs to be anonymous, secure and independently verifiable.” Editor: Gerck’s quote above comes largely from Ed Gerck, C. Andrew Neff, Ronald L. Rivest, Aviel D. Rubin, Moti Yung: The Business of Electronic Voting. Financial Cryptography 2001: 243-268. Springer Verlag. Note that this is a compilation of five separately authored short papers presented as part of a panel discussion.

Part Two Voting Train Wreck: an Editorial

This is my description of a handful of well-meaning people, people who nonetheless are actors in what is shaping up as the greatest technology policy failure of my lifetime. The story is also that there is another side to the story, a side that is being systematically underreported by the technology press. The problem: expensive and faulty voting systems, with no solution in sight. The solution: according to Dr. Peter Neumann, Dr. Rebecca Mercuri and others, either vote using paper ballots or do not vote at all. I have a problem with this solution.

It goes to the heart of the value, or lack of value, of computer technology in fundamental activities in our society. Even though I see no basic reason why computers cannot be used to tally votes, my problem with Neumann, Mercuri and others is not one of technology or mathematics or algorithms. They are the computer scientists after all. My problem is one of method and procedure and the reasoned applications of scholarly research when presented with new ideas. Their minds are not open to a solution that would not use paper in computerized voting (aka DRE) or, heaven forbid, use the Internet to vote. When someone like Dr. Ed Gerck challenges their world view, with qualified arguments summarized in this Executive Summary, we hear not much more than “it’s impossible on its face.”

I decided to enter into discussion with most of the principal actors in order to lay out their basic positions by showing what the beginning of a dialogue might look like. I have no intention of shifting the focus of the COOK Report to electronic or Internet voting. However in view of my own personal knowledge of what was happening, and the compelling human story that is developing, I have found it impossible not to take a public stand on the basic issues of scholarly methodology. I will watch and report on the extent to which they and others as well will make the effort to look more closely at these complex issues.

This voting train wreck comes in five parts:

Domination of the Electronic Voting Machine Debate By Dogmatism

A Brief Flashback – The Expertise of Eva Waskell

A Proposal to Maintain the Responsibility of the Scholar and Scientist

Ed Gerck’s Responses to his Critics, and

Ted Selker: An Academic Who Has Done his Job

Given both my observation of the train wreck that we have had with elections during the past five years and my knowledge of Ed Gerck, Safevote and Eva Waskell’s four years of work for Safevote, I have a problem with the way in which Peter Neumann, Lauren Weinstein, Rebecca Mercuri and Avi Rubin have managed to use their positions as academics and respected scholars to dominate the public discussion. They are, of course, entitled to use their scholarship to advocate any positions they choose and can defend with due intellectual rigor. But they also have responsibilities. They are generally expected to read a wide range of primary source material and to discover new sources if possible. They are expected to study, to debate, and to think. They are expected to probe and keep an open mind until independently verifiable facts move in to drive their analysis to a defensible conclusion.

Domination of the Electronic Voting Machine Debate By Dogmatism

These four individuals have done us all a great service in bringing their intellect to bear on exposing the risks of the pell-mell rush to DRE voting machines. They have correctly attacked the problems behind Sequoia and Diebold and ES&S. But they have become dogmatic in their approaches to the issue and have taken public positions that maintain it is impossible to prove, for anything but the most trivial computational module, for all voters that the code that is supposed to be running in the voting machine IS actually running there, and that nothing else is. Further, they maintain that anyone who claims to have solved the Internet Voting Integrity problem is seriously misguided, in the face of untrustworthiness everywhere throughout the Internet.

Thus, what happens, if there is a person who, through means that have not occurred to these four, steps forward and says: “I have a solution to these problems,” they will not hear what is suggested. They have said as much repeatedly. They are so publicly committed to their positions, that they will consider no other views or solutions. They maintain that they are computer scientists but if someone floats a paradigm shifting idea, they are not willing to hear or analyze it. And they will use the prestige of their positions and credentials to cast aspersions on ideas that they find to be uncomfortable.

My problem with Neumann, Mercuri, Weinstein and Rubin is not one of technology or mathematics or algorithms. They are the computer scientists after all. I am not. My problem is one of method and procedure and the reasoned applications of scholarly research when presented with new ideas. Their minds are not open to someone like an Ed Gerck who challenges their world view. If they can dominate public discussion, his ideas will never be heard above the cacophony. For the past five years I have watched this occur with great dismay.

The first three drove their stakes in the ground early. (Avi Rubin became prominent on the voting issue only in 2001.) In December 2000 with Risks 21;14
http://catless.ncl.ac.uk/Risks/21.14.html they concluded

"In the wake of the recent Presidential election problems, the knee-jerk reaction of "gee, can't we modernize and solve all this with electronic and/or Internet voting?" is predictable, but still wrongheaded. The shining lure of these "hype-tech" voting schemes is only a technological fool's gold that will create new problems far more intractable than those they claim to solve."
"Peter Neumann, Rebecca Mercuri, and Lauren Weinstein" (Risks 21.14)
See also
http://www.mail-archive.com/risks@csl.sri.com/msg00049.html

And by January 19, 2001 Safevote is safely tarred on Privacy Watch as just another “election company.” http://www.pcworld.com/resource/article/0,aid,38262,pg,3,00.asp
The battle over Internet voting is far from being won by the election companies: "There will be people who will plow forward in defiance of common sense and be the poster children for what doesn't work," said Lauren Weinstein, a co-chair of People for Internet Responsibility who has coauthored anti-Web-voting articles with Neumann.

Safevote's Gerck says that he believes key requirements, including voter privacy and election integrity, "can be guaranteed in a system using a protocol that requires both technology and human intervention," but Weinstein vehemently disagrees. "Internet voting, it's not even a close call: It's a garbage concept," Weinstein says. "From a technical standpoint it's easy to say it's impossible to fulfill the requirements you need for an election (using the Internet)."

Editor: The position these highly respected people took was that Gerck was just another DRE vendor pushing a flawed technology. Now my doctoral training is Russian History. Not mathematics, nor physics, nor computer science. I will not be so foolish as to debate these people on technical grounds. I will debate them on the grounds of scholarly process.

Furthermore although I have known Dr. Ed Gerck personally for five years, I would not make a public protest based just on his word versus theirs. Ed and Eva Waskell tell me that these folk have never seriously examined Ed’s technology design – dismissing it out of hand in the press and in face-to-face forums at conferences and hearings. Eva, who has given the last 20 years of her life to these issues, after working for Ed for six months told me: ‘what he is doing is real. It works. It is fundamentally different than any other approach. I have read all Ed’s work and the documents that support it and I fully understand at a technical level what is going on.’

These words from a person so dedicated were difficult for me to ignore. Approximately six months before the Florida recount debacle, I contacted Peter Neumann via email for I knew that he had known Eva for many years and believed that, if this person whom he knew well and presumably trusted, was convinced that this was new and different solution, Peter would undertake a serious study of Ed’s position.

A Brief Flashback

On December 1, 1999 I made an introduction of Eva Waskell to Ed Gerck. Up to that point both were in a private email discussion with Larry Lessig about his newly published Code and Other laws of Cyber Space. Before Eva responded to a comment about voting, neither person knew that the other was deeply interested in the application of computers to voting. Ed hired Eva almost immediately.

But why you may wonder do I care about Eva? I care because we have here a question of broad expertise lacked by the academics who are monopolizing the discussion. I first learned about her interest in the risks of the applications of proprietary black boxes to count votes in elections in 1994. There is a serious problem. In the 20 years since Eva became involved we have so many layers of new technology being pushed onto the public scene that it becomes very difficult to get some sense of the interlinking and therefore the reliability of the whole. Never mind issue of electronic voting at a walk in voting polling place versus casting one’s vote remotely using the Internet. Or using the distributed nature of the Internet to run interlinked processes serving as checks and balances to the elections process. Issues that, as readers will see, are not clearly demarcated in the comments of Neumann and Mercuri.

Policy experts must ask themselves who is an expert on the way elections have been held in the past? What about the entire system of the election process as it has been done in this country for the past fifty years? Who is saying that elections in the past have been perfect? They have not of course. Eva knows that because she has spent over 20 years in the trenches observing the totality of all procedures at the local state and national level. She has experience as a poll worker and through research, writing, investigative reporting, public speaking, grassroots organizing and election consulting. She is one of a relatively small handful of people who have credentials in the totality of the electoral administrative process – a process that Avi Rubin glimpsed for the first time in his service as an elections judge in Baltimore on March 2, 2004. It is not clear that the opponents have this experience. The breadth of her expertise should not be ignored because it gives her unique qualifications to make judgments about what Ed Gerck is doing. Computers? She wrote code in assembler in the 1980s.

A quick review of some of that experience is in order. In May 1985, while working as a stringer for The Economist, she accidentally stumbled onto a lead regarding a 1982 court case in Elkhart, Indiana in which the plaintiff alleged that the counting and the certification of the votes were "false and fraudulent.". The case caught her interest because the court ordered the vendor to turn over the vote counting software to the plaintiff for inspection.

Her research on the lawsuits filed against the vendor Computer Election Systems (one of the "ancestors" of ES&S) served as the basis for an article by David Burnham that appeared in the New York Times on July 29, 1985. She was quoted in that article: "Even when local officials learned of the problems [with computerized voting systems], little apparent effort was made to correct them." Over the next eight years, she read thousands of pages of transcripts from this lawsuit and others, and talked at length with the attorneys and computer experts involved. This education, along with hundreds of hours of interviews with election officials, formed the core of her knowledge base.

During the remainder of the 1980s, she went on to field study as a consultant to the Texas State Attorney General's office in their investigation of a Dallas City election. This Dallas City election was audited by Terry Elkins, the first person to ever perform a citizen's audit of election results. See David Burnham's September 23, 1986 article on page A26 of the New York Times, "Texas Looks Into Reports of Vote Fraud".) Eva worked closely with Elkins for many years.
In 1993 she helped compile, edit and distribute a Source Book on Computerized Voting, a joint project of Computer Professionals for Social Responsibility and ELECTION WATCH, an election watchdog group based in southern California and founded by Mae Churchill.
She helped organize the first conference on computer security in elections which was held at Boston University in 1987. As a result of this involvement she met Peter Neumann who published some of her work in the ACM Software Engineering Notes. From 1989 to 1999 she served as the volunteer director of the Elections Project for the Washington, DC chapter of CPSR. Mae Churchill and Marc Rotenburg led Rebecca Mercuri to Eva in the 1989-90 time frame. Rebecca and Eva kept in touch and shared information with each other for a decade until some months after Eva went to work for Ed Gerck in December 1999.

By 1994 Eva had also gained much experience in local Florida elections. In May of 1994 Eva left four years of full time employment at TeleStrategies and from that point until going to work for Ed in December 1999 she devoted herself full time to work with a nationwide network of election activists.

During this time, I was a friend who kept in close touch and was continually apprised of her efforts. I think it is fair to say that she was well known to the vendor and generally despised by them. At both the state and Federal level she knew very well the entire national election infrastructure including the widely respected Roy Saltman at NIST. I have never known an individual who dedicated herself with such total devotion and single-minded integrity to the public good – in this case to preserving the integrity of elections.

There is no way that Eva would ever have sold her soul to just “another DRE vendor.” After all at this point she had dedicated 16 years of her life to warning of the risks of black boxes being allowed to count votes using proprietary software. She would never have gone to work for a computer vendor whose approach was one and the same as the companies she fought for so many years. She interviewed Gerck as much if not more than he interviewed her. They found each other acceptable.

At about this time I contacted Rebecca Mercuri by email and told her that in my opinion Eva would not have gone to work for Ed if he were just “another DRE vendor.” I pleaded with Rebecca to study Gerck’s ideas with care. Then in the late spring of 2000 I emailed Peter Neumann and asked him out of his long knowledge of Eva and her dedication to the risks of DREs to sit down with both people and try to grasp how Gerck differed from the other DRE vendors. He politely declined. In December 2001 I met Avi Rubin in person at the UseNix LISA Conference in San Diego. I told him what I have outlined here. He declined as well.

My issue is one of what, if any, effort these people have made to understand the approach of Gerck and Safevote that they were lumping together as “just another DRE vendor.” In email correspondence they have told me that they believe they did understand what Gerck was saying and that they believed his work was simply not credible on its face.

I have a problem, not with the technical mathematical substance of the issues that I do not claim to be qualified to decide, but with the methodology being used by the four. Lump Gerck together with the other DRE vendors and dismiss him. In particular I have a problem with Peter Neumann and Rebecca Mercuri who, have known Eva for many years. If they had kept the open and cautious minds of true scientists, they would have stopped and asked themselves what it meant when a person who was one of the very first explorers of these issues with a 16 year record of single minded dedication to the dangers involved, appeared to endorse just “another” DRE System. Any scientist with an open mind would have said whoa! How can this be? They hopefully would have called Eva and asked her what happened?

There appear to be two possibilities here. Either that Eva had sold out or she had not. If she had sold out, then what on earth would that mean about the dedication of her preceding 16 years? A scholar trained to weigh evidence carefully and make a decision on the basis of information that he or she had personally verified, should have confronted Eva and interrogated her until it became clear whether or not Eva had ‘gone over to the enemy.’

Of course there are two problems here. The easy outcome would have been verification that Eva had betrayed what she had dedicated herself so selflessly to for so long. Neumann’s and Mercuri’s suspicions would have been verified with a triumphant case closed.

The other possibility was far more troubling. What if Eva had not sold out? What if Gerck’s technology were real and it worked? That would mean that a lot of lines that Mercuri and Neumann were drawing in the public sand would have to be redrawn. That public assertions made as to the impossibility of doing what Gerck claimed would have to be retracted. A very unpleasant prospect. Furthermore the other problem was that confronting Eva and Ed in extended discussion and debate would be unpleasant and fraught with unforeseeable outcomes.

I have not looked forward to the prospect of having to throw down the gauntlet in front of four very prominent and highly respected people. But I have done that now. I have corresponded with all four. It has not been pleasant nor did I expect it to be. But in doing so, I have learned a great deal. I suggest that it is time for Peter and Rebecca to do the same with Ed and Eva.

A Proposal to Maintain the Responsibility of the Scholar and Scientist

I began by sending Peter a draft of my proposed editorial on Wednesday night and on Thursday morning I sent him a second message explaining that my grounds for challenging him were based on the methodology of his scholarship.

I said: “I well understand your reluctance to trust computers in voting of any kind and absolutely agree that considering where we are now, [with both sides sharply polarized and with insecure DREs scheduled to record the majority of the votes cast in November] we'd be better off with the old analog paper based machines. Therefore I am pleased that you and Lauren and Rebecca and Avi are making life miserable for the Diebolds of the world. You should.”

“But I have a problem nevertheless with the objectivity of your scholarship. In Risks you publish one of the most famous mail lists on the globe. And where do most of the risks come from? From the computer doing its own thing under conditions that were not anticipated and without a human in the loop to intervene when trouble occurs. Computer scientists have not designed their systems to take human behavior into account. They are maintaining, in effect, that they cannot be so designed.”

“And what does Gerck say on my list? "The solution considers not only machine-machine communication channels but also human-machine communication channels because the voter can act as a source and as verifier in more than one part of the system. Further, human-human communication channels must also be considered because we do not want machines to have the potential to run amok, unchecked." “

Gerck is not a dumb man, nor is he a charlatan, Peter. Have you ever spent more than a few minutes in serious conversation with him? I have. . . . Almost four years ago I reminded you that Eva was working for Ed and pleaded with you out of respect for your many years of knowing her, to take a closer look. You responded with a politely worded refusal.

I am reminding you again today one of the earliest and most dedicated activists in this field still says that there is a computer/human-based answer and that the technology that you have never seriously examined works. I believe that Eva would never go over to the “enemy,” Peter. Therefore, until you can say that you have sat down with both Eva and Ed for as long as it takes to get to the bottom of this; conclude either that it does work or that it fails; and do so in a documented scholarly essay, I am prepared to say in public that I believe you have failed in your most fundamental responsibilities of scholarship.

To his credit Peter responded on March 4: “Come on. I have read Ed's stuff. It proves nothing. I have spent various times at various meetings with Ed. We seem to agree that we disagree on some things and agree on others. . . . . I will be very happy to take another look at what Ed and Eva are doing.”

I find Peter’s response commendable, but, in view of assertions, he has made in his follow up conversations with me, views which I present below with Ed Gerck’s comments, in order to have any reasonable chance of an outcome that makes any progress toward a resolution of the issues, there needs to be not just another look but a real, in person, dialogue.

As a suggestion if Ed and Peter agree to meet for a face discussion I would gladly publish a recording of the discussion. They could sit down in private and engage in a dialogue that would give Peter a chance to probe and test Ed’s ideas. Each individual could make his or her recording of the dialogue. Each could write up their own summary of what they felt they learned with the understanding that: I would publish the two resulting summaries and that, should they be unable to do this because they remained in serious disagreement, I could publish from the tape a transcript of what was said. If Peter wished to ask Rebecca Mercuri to participate, he would be welcome to do so. I am sure that Peter and Ed would agree that the future of free and honest elections in the United States deserves no less than an extensive effort by all parties to reach agreement.

Ed Gerck's Response to His Critics

I asked Ed to respond to comments sent to me by Rebecca and Peter on March 5,
Rebecca Mercuri: I have read Ed Gerck's work and have iterated with him over email on numerous occasions. His theories are flawed, on various counts. Peter has attempted to explain to you the problem with Gerck's work, but I will reiterate that Gerck has never supplied a strong mathematical proof (this is quite different from theoretical musings, it is a FORMAL proof) for his crypto-based voting system.

Gerck: My papers describe an information-theory based voting system, both for electronic voting as well as for Internet voting. In both cases, formal proof was already provided by Shannon 50 years ago. The fact that Mercuri mentions a "crypto-based" system" just underlies a basic difficulty she has in reading the material.

Mercuri: As well, for a voting system to be acceptable for use in a democratic governmental election, it must be transparent and hence completely understanable by all citizens (let's assume a U.S. High School education).

Gerck: When systems are implemented using what I propose, anyone can verify that the system is working as specified with a confidence level (statistically defined) as close to 100% as desired. If the system is not working as specified for its voter-verified records, its own checks and balances shall indicate a malfunction. This is much better than what current paper-based voting systems provide, where no one can have any confidence level that the system is working as desired.
That said, the principles behind what I propose are very intuitive and the Moguls in India already used them successfully 500 years ago. They used at least three parallel reporting channels to survey their provinces with some degree of reliability. Their additional efforts were paid for by the improved results.

Mercuri: So this mathematical proof must not only be correct and complete, but there must be an independent way for all voters to assure that the code that is supposed to be running in the voting machine IS actually running there, and that nothing else is. Since this is theoretically impossible for anything but the most trivial computational module.

Gerck: "All voters" can be assured that this is true in the systems that I propose, to the confidence level required, by principles already well-known to the election community. Roy Saltman published the mathematical analysis while he was at NIST, in 1975. The mathematical basis was supplied by Shannon 50 years ago.

Mercuri: It will probably take Ed Gerck a LONG time to come up with this proof, but Peter and I (and others) have encouraged him to do so, and we will look forward to seeing his proof of correctness and transparency when he is able to accomplish this task.

Gerck: The proof was given by Shannon 50 years ago and exemplified in a series of papers by Roy Saltman almost 30 years ago. My work was, actually, much simplified and made shorter in time by following what these pioneers have done.

Neumann: Anyone who claims to have solved the Internet Voting Integrity problem is seriously misguided, in the face of untrustworthiness everywhere throughout the Internet.

Editor: Peter is missing the issue here that there are two levels - electronic voting by walk in local polling places and the use of the internet in polling and administrative collection of data. He has just dismissed it with the assertion that on its face it cannot be done. He needs to give his own proofs rather than flat assertions. The quality of these exchanges goes to the heart of my concern about the methodology in play.

Gerck: Neumann, Mercuri, Rubin, Weinstein and others say that one needs to assure reliability for all parts of a system, including hardware, OS, and voting software in all hosts, for all possible failure modes, in order to assure reliability for the system. Which, of course, is impossible. Hence, a solution is impossible, no matter how it is proposed.

This is the blindfold that prevents everyone who puts it on from seeing a solution. Reliable systems can be made of unreliable parts. This principle was known to the Moguls in India 500 years ago and was mathematically modeled by Claude Shannon some 50 years ago. And this is crucially important because there is no other way to build a reliable system.

Anyone with the required technical background, who would seriously consider reading what I have ALREADY published (there is no need for new material) should be able to understand that there are indeed: (1) paperless solutions for electronic voting; and (2) solutions for Internet voting. A search on Google should also provide replies that I have ALREADY given in public lists to ALL the questions posed so far by Neumann, Mercuri, Rubin, Weinstein and others.

For those interested, here are the main references:

1. DRE - ELECTRONIC VOTING: On August 2-30, 2001, I presented an invited paper at the WOTE'01 conference in Tomales Bay, California. The conference was about trustworthiness in voting systems. My paper was on the Witness Voting System, a provable, reliable solution for voter-verified electronic voting (DRE), providing integrity and anonymity proofs, that does not use paper. Peter Neumann was present and had no questions. The conference proceedings is available electronically at http://www.vote.caltech.edu/wote01/pdfs/ , look for gerck-witness.pdf
2. INTERNET VOTING: Published in 2002. Please see Ed Gerck: Private, Secure and Auditable Internet Voting, chapter in "Secure Electronic Voting: Trends and Perspectives, Capabilities and Limitations", Edited by Prof. Dr. Dimitrios Gritzalis, Kluwer Academic Publishers, 2002, ISBN 1-4020-7301-1. See also http://www.wkap.nl/prod/b/1-4020-7301-1?a=1

Now, critics have suggested that all aspects of such systems be understandable by voters with a high-school education, so that voters could know why they are relying on such systems. The answer has already been given above. When such systems are implemented ANYONE can verify that the system is working as specified with a confidence level (statistically defined) as close to 100% as desired. This is much better than what current paper-based voting systems provide, where NO ONE can have any confidence level that the system is working as desired. Election systems have been the subject of fraud in this country for 200 years.

When I asked Ed to summarize, his view of the current state of the dispute, he wrote:
"The whole issue of electronic-voting and Internet -voting has been addressed by Peter Neumann, Rebecca Mercuri, Avi Rubin and other critics with a complete determination to arrive at a negative answer. That any attempt undertaken will fail, since by their own definition all needed infrastructure is insecure. And that anyone who claims to the contrary is incorrect by principle. Consequently it should be no surprise, that my technology fails to meet their requirements, if by their definition, it is impossible to do what I am in fact doing.

I welcome, however, their expose' of faults in current electronic voting systems and their stabs at Internet-voting. Trying to cure all the faults (which would take them a LONG time to find) will nevertheless, fail to lead to a secure electronic voting or internet voting solution.

The answer is not to demand perfection for all parts. Reliable systems can be made of unreliable parts. And this is crucially important here because there is NO OTHER WAY to build a reliable system, as Peter, Rebecca, Avi and other critics will eventually demonstrate by the exhaustion of all other possibilities.”

Editor: Having to use expensive and faulty voting systems without any solution in sight, and hearing these arguments against Ed, I would demand a much more rigorous explanation than just “it’s impossible on its face.”

Ted Selker: An Academic Who Has Done his Job

I asked Ed Gerck for an example of someone in the academic community, who he felt had done his homework on the questions of election technology and who understood where he, Ed, was coming from. He named Ted Selker of the Media Lab at MIT who has researched both electronic voting (DRE) as well as Internet voting.

The best documented evidence available of Selker’s ideas that I have found comes from a September 8, 2003, “interesting-people” message from Stephen D. Poe, CEO of Nautilus Solutions to Dave Farber. Poe attached an interview that Simson Garfinkel had done with Ted Selker. Poe also wrote: “I spend too much time in my consulting practice trying to convince clients to pay less attention to the hype about all the issues that could arise in the electronic world and pay more attention to the simple questions of "Is this proposed method more or less secure than what you have now? Does this proposed method provide more or less benefit for your buck?"
Editor: Poes’ statement above deserves serious thought. Given the disaster of the last election, with emotions understandably running high and with enough found out about the flawed Diebold machines to indicate that there is strong evidence of a serious threat in the application of the technology, one still has to wonder how careful the efforts of these leaders of the anti DRE movement have been? But given the heat of the battle, and the press’ inability to run with more than simplistic sound bytes, I can understand why, with the best of intentions undoubtedly, they have been using baseball bats in a slugfest.

My intention in this editorial is to appeal to a stronger approach that will better serve the public interest of maintaining the integrity of our electoral process.

Poe to Dave Farber: The following article starts to address those questions. Editor: In his article “Campaigning for Computerized Voting,” September 3 2003 Simson Garfinkel decided to go beyond the usual critics of electronic voting. In his own words Simson met one “who isn’t opposed to DREs. In fact, he’s positively enthusiastic about them. And this man isn’t just anybody; he’s Ted Selker, an award-winning inventor with many patents, formerly with IBM Research, currently a professor at the MIT Media Lab, and member of several panels and commissions that looked at the issue of voting following the debacle of the 2000 presidential election.”

Here is what Simson Garfinkel had to say: "I met Selker a few days after he had attended a meeting of computer scientists and election officials in Colorado. He was livid. He had just spent two days listening to the experts of the field talk about all of the failings with DREs and how these systems could be used to steal an election.

“’What these people don’t realize,’ he told me, ‘is that automated tabulating machines were invented for a reason that is, because paper is a fundamentally bad way of making and keeping accurate records. Paper is bulky and heavy. It can be hard to read something recorded on paper, no matter whether the marks were made by hand with pen-and-ink or by a computerized printer. Paper rips and gets jammed in machines. Paper dust gets everywhere.’ Eliminating paper, Selker explained to me, has the potential for dramatically improving elections.”

“But what about all of the ways that you can hack the voting machines?” I asked him. Selker laughed. Politicians, he told me, have been hacking elections in America for more than 200 years. The geeks are focusing on the abilities of hackers to steal elections by reprogramming DREs because electronic attacks are what these folks understand. But if your goal is truly better elections, he says, the DREs can do more good than harm.

Editor: Selker is suggesting that by careful design the equipment can become more useable and offer an increased possibility of reducing fraud.

Garfinkel: Before talking with Selker, I was squarely in the anti-DRE camp. After listening to him, I realize that there is another side to the story that is being systematically underreported by the technology press. Did he convince me? Well, let’s say that I’m no longer convinced of the inherent correctness of the anti-DRE position.[The entire interview is found at: http://www.interesting-people.org/archives/interesting-people/200309/msg00054.html ]

Editor: As with most technology and policy issues this one is horribly complicated. I want to close here with a reminder that my complaints are based not on the technology substance but rather on the methodology employed. I have indicated what my problems with the methodology are. There is always time to temper what is shaping up as the greatest technology policy failure of my lifetime
I have ventured into this minefield because, in the heat of ongoing battle, I believe that some academics who have positions of public trust, have, nevertheless, suffered an inadvertent loss of perspective and are not keeping their minds open to the ideas of another man with equally outstanding academic credentials: Dr. Ed Gerck. Yes, Ed Gerck represents with Safevote a commercial business, a vendor. However, I disagree with the tacit premise that individuals who sell products are basically dishonest about their product's limitations. I think that being accurate is good business. These days those who stretch the truth often get caught and lose their reputations and business. That seems to be potential problem, now, for Gerck's critics who have been saying that it is impossible to do what Gerck claims to be doing. They have dismissed his work out of hand. I have talked both to them and to Gerck and have found that it is clear that what review of Gerck's work they have done has been marked by the unceasing assertion that what he was proposing was not possible. However, many experts, including Waskell, agree with Gerk. Also more than 300,000 voters in 35 elections, reportedly, have used Gerck's system successfully. Arguments that say, it is impossible to do what Gerck claims to be doing, remind me that everyone knew, 101 years ago, machine powered conveyances cannot possibly enable us to fly.

I have no intention of shifting the focus of the COOK Report to electronic or Internet voting. However in view of my own personal knowledge of what was happening I have found it impossible not to take a public stand on these issues of methodology. I want to close with another thank you to everyone I dialoged with on this issue. I will also watch and report on the extent to which they and others as well will make the effort to look more closely at these complex issues.
Disclaimer: Ed Gerck has been a $300 a year subscriber to this newsletter for four years. I have never received any income from any other financial transaction or compensation for anything from Eva Waskell or Ed Gerck. Nor have I any ownership stake in any enterprise owned by these two people.