Verify the Vote: Tell Congress to Fight for Secure Elections! [EFF

Electronic Freedom Foundation ^ | November 10, 2003 | EFF

[risk]

TAKE ACTION! SEND A MESSAGE

Verify the Vote: Tell Congress to Fight for Secure Elections! The 2004 presidential election might not be flawed like the last one; it might be worse. Communities across America are purchasing electronic voting (e-voting) machines, but the technology has serious security problems that need to be addressed. Most of the machines use "black box" software that hasn't been publicly reviewed for security. Almost none of the machines provide voter-verifiable paper ballots to detect fraud. And despite the efforts of one voting technology company to stifle its critics, the public has become increasingly aware of the problems in e-voting. Send your representative a letter supporting the Voter Confidence and Increased Accessibility Act of 2003 (HR 2239) to require openly reviewed software and voter-verifiable paper audit trails for all new e-voting machines. Note: If your representative has already pledged his/her support for this bill, you will not be able to take action. Please click on the "Learn More About This Issue" button to see if your rep has already done the right thing.

November 10, 2003

Your U.S. representative

Dear Representative,

I am a constituent who cares deeply about security and trust in elections, and I urge you to cosponsor the Voter Confidence and Increased Accessibility Act of 2003 (VCIAA, HR 2239). As touchscreen voting (e-voting) technology is adopted across America, it's absolutely vital that these new systems meet basic standards of accountability and openness. I am particularly concerned that many systems do not use openly reviewed software and cannot provide a voter verifiable paper audit trail. Unless a paper record is generated by publicly reviewed software, verified by the voter and retained for potential recounts, I believe that this technology is unacceptable for use in our elections.

The public should be allowed to review the software that runs these machines in order to confirm that they act in the way that the manufacturer claims. Right now, however, the leading technologies are not only proprietary, they are covered by trade secret claims. This kind of closed source, or "black box," software lacks sufficient quality assurance. In fact, two recent analyses of one company's software revealed gaping security holes that could be exploited by anyone from a technically proficient insider to an average voter with a few dollars worth of technology in his or her pocket. This underscores how important open source software is for our elections. HR 2239 would require voting machines to use publicly reviewed software.

HR 2239 would also mandate voter verifiable paper audit trails for all new e-voting machines, a prerequisite for accountability and accuracy. The 2000 presidential election was a painful lesson in the failings of current voting technology, but at least there was a back-up system that allowed a manual recount when evidence emerged that the regular voting process was flawed. Without a paper audit trail, a compromised e-voting system could not provide even the cold comfort of Florida's manual recount. Luckily, adding this protection to the machines is not hard and does not have to add a significant amount to the cost of each unit.

I strongly urge you to cosponsor HR 2239 to ensure that all new e-voting machine purchases provide a voter-verifiable paper audit trail and use publicly reviewed software. Thank you for your time.

Sincerely,

<click here to visit the eff page and send the letter>

[risk]

I don't know if HR22390 goes far enough. But it's worth considering, and worth alerting your congress critter if you agree it's a good idea. I'm interested in people's perspective. At this point, I support it. Is there a better choice?

Some useful HR22390 Links

What	Where
Introduction to HR22390	house.gov (Rep. David Price)
Election Reform	house.gov (Rep. Robert Wexler)
HR 22390 propsal	humanfactors.com
HR 22390 propsal (PDF)	acm.org
EFF on Diebold	freerepublic.com (from EFF site)

[risk]

HR 2239 ping.

[zeugma]

I haven't had an opportunity to read it yet, but I've looked fairly closely at this whole idea of electronic voting. The average person out there has a piss-poor idea of even the issues surrounding voting itself, and will sign up for just about any stupid scheme that is foisted upon them by their political masters.

I can't even imagine how anyone can support a voting system that doesn't include a verifiable audit trail that is, at the same time allows for truely secret balloting. Any system that compromises either of those principles is a non-starter. For this reason, I seriously doubt I'll ever see Internet voting implemented in a satisfactory way. There are well developed and thought out protocols for doing secure, anonymous, verifiable, and auditable electronic voiting, but they are very difficult to implement in any but the most ideal conditions.

Hell, there are protocols for playing poker on the net through email that can be proven secure and honest. That doesn't mean anyone wants to waste the computational power it requires.

Another issue with electronic voting machines is that there is a large vested interest in fraud. Noone will admit to it, but that doesn't mean it is not true. All in all, I don't have a lot of short-term optimism in these systems, because I just don't see the proper mind-set being brought to bear on the issue.

[TheOtherOne]

BTTT

[HangThemHigh]

I haven't taken time to study this in great detail, but it requires an auditable (paper) record of the vote which is to be verified by the voter. I think this is essential to any voting system.

I don't see that it contains any safeguards against stuffing the ballot box, so risk may be right in wondering if it goes far enough.

Intersetingly, HR 22390 requires disclosure of the soruce code to any ctizen demanding to see it. I wonder if any software supplier will go along with this?

[PetroniDE]

I don't know how many times I have discussed this, but here it goes.

Electronic voting (e-slate in Texas) may not be perfect, but systems like the Florida votomatics are FAR WORSE. I have experience as a precinct judge with both systems.

With non-electronic systems, you don't even have to be at a voting booth to cast ballots. Just take some ballots to an unmonitored location, poke the right hole(s) with a thumbtack or message board pin, and PRESTO (More Votes for the Dem Candidate).

[antiRepublicrat]

Electronic voting (e-slate in Texas) may not be perfect, but systems like the Florida votomatics are FAR WORSE.

With current electronic voting, you are asked to simply trust a large corporation, not with just anything like your money or such, but with your vote. Diebold (the company quoted as using lawsuits to silence its critics) is already suspected, with lots of obvious evidence, of bypassing the standard government software auditing procedures.

Do you trust a company with your vote? Do you trust some programmer in that company with your vote? Just think of one software tweak resulting in millions of purposely changed votes throughout the country, and no paper audit trail to confirm.

I say this not as an anti-technology luddite, but as a programmer who loves tech stuff and sees the great possibilities inherent in electronic voting, if done right.

[arthurus]

The electronic voting machines should be much easier to rig because one person can control many machines in many districts.It centralizes fraud and makes it much more efficient costing many jobs in the votrigging industry. People no longer have to directly manipulate paper and individual machines. It can all be done by pushing "F3".No one gets caught with vote counters in the trunk of his car. No one has to coerce aliens and the comatose to vote.

[PetroniDE]

Do you trust a company with your vote? Do you trust some programmer in that company with your vote? Just think of one software tweak resulting in millions of purposely changed votes throughout the country, and no paper audit trail to confirm.

I say this not as an anti-technology luddite, but as a programmer who loves tech stuff and sees the great possibilities inherent in electronic voting, if done right.

It sounds like the problem is a state specific issue. I have no idea who makes the e-slates used in Texas (but feel certain it is NOT Diebold).

I shall use the Texas e-slate as an example. The vote totals are stored in THREE different places (the main JBC, required end-of-day printouts, and each e-slate machine). These systems were beta-tested and used in previous elections with no problems, except for incompetent and un-qualified/trained democrats precinct judges.

It is not the company that should be in question; it is the people in-charge that is the problem.

IF YOU HAVE CORRUPT PEOPLE RUNNING THE SYSTEM, YOU WILL HAVE CORRUPT RESULTS, REGARDLESS OF THE SYSTEM.

The reaction of "no paper trail" is an exaggeration. At least with this system, one must cast a vote at a booth (unlike the votomatics).

[PetroniDE]

I agree that Internet voting is NOT the way to go. I have NO problem with electronic voting otherwise.

If voting procedures are properly followed, there are multiple counts of any vote totals. See post #10 for an example regarding Texas e-slate. It is much more difficult to commit vote fraud electronic than by a simple F3 command.

With traditional paper, you don't even need the stinkin' machine. Give a democrat a stack of ballots, message pin, and the correct punch number, and ANY democrat can steal ANY election. No voting booth even necessary.

[zeugma]

I have no particular problem with using electronic devices to record the vote if they are properly set up. I've actually given this a bit of thought. The problem comes in when there is no paper trail to facilitate recounts and audits. There is no reason at all that a computerized voting booth couldn't be set up in such a way that the vote, once made, is printed out on a piece of paper that is both machine and human readable. This paper woud be considered to be the actual ballot which would be fed into reader that would actually tally the vote. The voting computer could also tabulate votes as they are made, and should be useful as a cross-check against what is tallied by the reader. Any descrepancy would be flagged for later retabulation.

Both the voting machine any the tallying machine should be completely open source to satisfy voter questions about the security and accuracy of the vote. There not be some kind of 'black box' that we are just supposed to trust because fedgov tells us to. There are many other safeguards that can be implemented easily to lessen the possibility of fraud. THere is no way to eliminate all fraud. After all, the dead still vote in many places throughout the country. I'd make sure that voter identity was more closely verified than it is today, and would probably require voters to re-register at least every 4 years, with complete purges of vorter rolls along the same cycle.

[zeugma]

The reaction of "no paper trail" is an exaggeration. At least with this system, one must cast a vote at a booth (unlike the votomatics).

Not trusting balloting systems is not an over-reaction. I have worked with computers for many years now, and have managed systems from mainframes, minis, to workstations. I simply know that any computer system can be gamed by a determined adversary. Granted, such gaming would be harder with an electronic system than simply taking a pin to stacks of ballots to manufacture votes. However, it would also be almost impossible to prove absent a non-electronic audit trail. Having both electronic and physical tallys would make fraud at least more difficult than it is today.

[PetroniDE]

Voting roll cleanup is good, but outside the scope of the electronic vs. non-electronic vote machine discussion.

In Texas, we have a paper trail (the open-poll and close-poll printout reports, see post #10). However, I must dissent regarding "reading of paper". There are many more way to rig paper controlled counts compared to computer counts. Computers are not biased; people are. I have demonstrated in an earlier post how to rig a punch-card ballot without even going to the voting booth.

No disagreement regarding open source info regarding computerized code. Let the company patent it to protect their copyright.

Vote fraud is the people running the system and not the system itself. Leave a precinct in complete control of corrupt democrats with no GOP "eyes and ears", and ANY vote totals are possible.