Home ShopTalk Can VoteHere Save American Elections – and the Free World?

Can VoteHere Save American Elections – and the Free World?

This November, for the first time in American history, around 50 million Americans used electronic voting machines similar to ATMs. What separated these automated transactions from the ones we bank on everyday is paper thin, but it’s as important as our US Constitution. When you vote electronically, there is no receipt. Without that accountability, our venerated democratic institutions are undermined and our faith in government drastically diminished. The questions are these: How did it come to this? Is there a better way? And what if you could verify your vote was accurately recorded by going online or picking up the telephone?

Technology from a Puget Sound company known as VoteHere has a powerful role to play in reshaping the new era of electronic voting. VoteHere’s “Technology inside,” (VHTi for short) is ready to go to work in voting machines from Diebold, AVS and Sequoia, among others, to guarantee the sanctity of voting accuracy with total transparency and accountability. What’s holding things up? A slow moving bureaucracy for one. Stop-gap thinking for another.

Electronic voting is still in its infancy, and not without growing pains. In some states, legislation has been far overreaching and has mandated that paper ballots [not just receipts] be part of any electronic voting system. “It’s as if the politicians saw the cockroaches in the cupboard, and instead of calling the exterminator, they decided to burn down the house,” says Jim Adler, VoteHere founder and CEO.  “It’s like the 1920s when they required a horse escort for every automobile. If you go back to paper ballots, you effectively invite the kind of ballot fraud and error we’ve had in this country for 150 years. You can’t improve elections by going back to a flawed system.”

Adler found his way into the voting booth from the technology side, moving from signal processing in the aerospace industry to “a more terrestrial mission-critical application — elections.”  Since then his developmental efforts have been among the most transparent of any technologist. In September 2003, VoteHere went so far as to publish its cryptographic technology online, inviting scholars and hackers alike to poke holes. An even greater revelation followed when VoteHere posted its source code online in April 2004.

We talked with this cyber-age patriot about working at the crossroads of democracy and technology.

Seattle24x7: What is the difference between digital signature technology that authenticates a signer and the technology that validates a ballot cast by a voter?
Adler: We use digital signature technology within our approach of securing elections. But we have an added requirement, which makes it much more difficult, which is that we also have to maintain the secrecy of the voter’s ballot. That requirement, in and of itself, makes voting extremely difficult which is why somebody like our chief scientist, Andy Neff, is a Ph.D. in the field of theoretical math. We can’t use the generic cryptography tools that are on the market.

Seattle24x7: You made the leap from space-age technology to information age election IT. How did that happen?
Adler: I came out of school playing with 300 foot booster rockets and doing advanced avionics on commercial launch vehicles for what is now Lockheed Martin. A lot of it was signal processing. When I left to go to graduate school, I moved into data security, which is mathematical kin to signal processing. From that, I left the aerospace industry, and came to a more terrestrial mission-critical application — elections. I found that the science of elections had not advanced at the same rate as other information technologies, and started to apply the core competence of our team to voting after founding this company in 1996.

Seattle24x7: Your VHTi technology stands for what exactly?
Adler: It stands for “VoteHere Technology inside” which is technology that our engineering team has developed that goes inside any electronic voting system, or any voting system period, to secure that election. It provides verification properties for the election and auditability properties so that as a voter, you can verify that your vote was counted properly. Anybody can audit the results. As a software technology company, we license that technology to electronic voting machine manufacturers, system integrators, other security companies, or even software companies, anyone who is involved in the electronic voting marketplace in the US and around the world.

Seattle24x7: The other acronym we should know is “DRE.”
Adler: DRE stands for Direct Recording Electronic voting machine. The common term is really the touchscreen voting machine. Other terms are EVM or electronic voting machine, but in the US they go by the term DRE. Up in Snohomish County, they have these touchscreen voting machines.

Seattle24x7: How did touchscreen voting machines become implemented without the ability to create a paper trail?
Adler: A long, long time ago, in 1999, (which seems like forever from now), it was believed that these electronic voting machines were fine. They are not hooked up to the Internet. They are application-specific. There was really no reason to have any concern. That was the position of even the current electronic voting critics back in 99.

I was one of the ones who said, “Wait a minute, that’s not true. These systems are just as susceptible to any sort of manipulation by insiders that could install malicious software. They could also make accidental mistakes. At the time, we had developed this technology and the industry really wasn’t ready for it. Many of the vendors, the electronic voting machine manufacturers, told us, you guys have great technology, but the market really isn’t asking for it.

Fast forward about 2 1/2 years, and this starts becoming a huge issue. There were certain things that I think conspired to make it an issue. The simple analogy is to ATM machines. It would be as if they didn’t give you a receipt. Would you use an ATM if you didn’t get a receipt? I don’t need to get a receipt at the ATM, but my wife, who does the family books, won’t allow me to take out cash and not get a receipt. I mean, a receipt is just standard procedure for anyone using an electronic machine.

Seattle24x7: Your other product, RemoteVote, is a total E-voting platform that will work on any digital device via telephone wireless phone, kiosk and more?
Adler: That’s right, and if you look overseas, especially in the UK, but mostly in Europe, they are doing quite a bit of remote voting right now. We won a three-year contract in 2003 to provide remote E-voting services to the UK government where they want to move completely to electronic voting by 2008. RemoteVote is a complete application election management system that is installed in a data center, and can take votes through any of those channels.

Seattle24x7: How is RemoteVote able to authenticate the one person-one vote requirement?
Adler: The authentication is similar to an absentee ballot vote. In the UK, they have a relatively easy situation. They do a mailing to every voter during every election cycle. Through that mailing, they mail out PINs. In the UK, there is a PIN and a birthday authentication mechanism. Having said that, RemoteVote doesn’t impose any authentication on the jurisdiction. They can use whatever they want. If they want to fill out smartcards, we can use that. If they want to use PIN and passwords, we can use that. If they want to use biometrics, we can incorporate that. We plug-in to whatever they want to use.

Seattle24x7: In the role of founder and CEO of VoteHere, you are an evangelist for democracy and accountability in the election process. Did you ever think that technology would prompt you to play this type of role in government ?
Adler: That’s the most rewarding part of what were doing. We are at the threshold of bringing transparency and auditability back into elections. Back when we had elections 150 or 200 years ago, society wasn’t that complex, the scale was small, you went into town to vote, it was one ballot box, and it was pretty simple. What’s really exciting is that you take these constitutional values of transparency and accountability, and you bring them to the information technology world. The constitutional values should be independent of technology. The fact that we get to say that these values are so wide in their construction, that we can apply them to our more complicated technology world, that is our role.

Seattle24x7: You must be as disenfranchised with the lack of progress as the American electorate. The United States currently ranks 139th out of 163 democracies in the rate of voter participation. How do you explain this failure to reform the election process?
Adler: In many ways, we’ve tended to follow our nose as far as voting is concerned. Instead of doing what we ought to from a democratic value perspective, we do what’s technically possible. Punchcards are a typical example. We’ve used a system that didn’t capture the votes correctly. It didn’t count the votes very well. And yet, we used it for 40 to 50 years. No one stopped to think, wait a minute, do we have enough transparency? Do we have enough audit here? What is the vote capture error rate? Those questions are now being asked vehemently. It’s great to be part of that discussion.

Seattle24x7: You’ve also been critical of election audits?
Adler: We did an analysis of the Venezuelan audit. One line of analysis says that the Venezuelan recounts or the Venezuelan audit was inadequate to the point of allowing an 11% swing in the election. In other words, the audit would not have caught a fraud that would have allowed an 11% swing in the election. What I find fascinating is that no one in the world right now is looking at the efficacy of these audits to discover fraud. In the case of Venezuela, they audited a half of a percent of the machines. I didn’t see any analysis that said, what was the efficacy of that audit?

I presented a paper at NIST, the National Institute of Standards and Technology, that goes through that argument. It points out that with these DRE’s, there’s just no way to know the “fraud rate” which is the allowable fraud rate that could go undetected by an audit. Unless you have some way, some receipt mechanism, to check, you will continue to have an unknown fraud rate. It turns out that California, even with their punchcards, have been doing hand recounts over the last 40 years as an audit, so at least you can calculate what the allowable fraud rate is. It happens to be exorbitant, but at least you can calculate it.

Seattle24x7: The solution is clearly a marriage of electronic access and instant verification with paper authenticity, the election equivalent of click and mortar. The apparent problem is that government sees it as an either or proposition?
Adler: You can look at electronic voting and say oh my God, we need some audits. But if you just go back to paper ballots, in effect, you will invite the kind of ballot fraud and errors we’ve had over the last 150 years in this country. If you go back and look at, say, the New York Times, and the cases of ballot fraud they’ve documented over the last 150 years, it turns out to be around 800 cases of paper ballot fraud, about one every 70 days. Paper ballots are obviously broken. It’s how we got in this mess in 2000 to begin with. But to say that electronic voting is so bad that we have to go back to this other flawed system, is very shortsighted.

Seattle24x7: You’ve also issued a challenge to see if anyone could hack into your system. Can you explain that?
Adler: What we did in September 2003, was that we disclosed all the technical documentation of our technology. In the spirit of transparency, we said here is all of the mathematics. We are protected by patents so we felt very comfortable doing that. We let every academic in the world try to poke holes in it. They could either implement it or they could just say theoretically, there’s a hole in it. We didn’t  receive any criticisms or any attacks that our technology would not have detected.

Then in April of this year, we actually released an implementation of the source code over the technology. So first we gave out all the documentation and then we disclosed all the source code. That got us a lot of kudos for being open and transparent, but, most of all, no one has come forward and said your system has a hole in it. It’s been over a year now that we’ve had that challenge out there. It’s as if we gave them the architecture to the safe and said ‘Propose a way to break the safe.’ It’s one thing to put a safe in the middle of a city and say break into it, it’s another to give somebody the drawings of the safe.

What’s interesting is that if any of those DRE’s (Diebold, Sequoia or the others) had our software providing a way for voters to track and prove that their vote was or was not counted properly, they wouldn’t have to disclose their source code, it wouldn’t matter. You can go on the Internet or use a telephone and verify that your vote was counted properly in the final result.

Seattle24x7: Last question. What will voting technology look like 50 years from today, how will it have changed?
Adler: In many respects, if we take these democratic values, and project them onto today’s technology, it will serve us for centuries. I think you’ll see incremental improvements going forward. It will get better, faster, and authentication and biometrics will get stronger. Most of us out West vote remotely. So we want to be able to cast a ballot, but not be confined to a polling place. That, again, is a democratic value from the 1800s, the Australian ballot. We take these democratic values that date back to the Greeks in many respects — one person, one vote, transparency, audits, secret ballot, and we look for technologies that allow us to implement those democratic values. The good news is that the technology is available now, today. Now we just have to do the business of getting them into voter’s hands. [24×7]

Larry Sivitz is the Managing Editor of Seattle24x7.

Previous articleMercent: Minds Over Markets
Next articleNlighten Me, Please