Bob Graham for President! Urrrr, whatever.

December 11, 2003 11:58 PM

Graham to go out with a bang--will Diebold go out with a wimper?Senator Bob Graham, D-Florida, today introduced the Voter Verification Act, legislation that would require computer voting systems to produce a paper record.
posted by Ignatius J. Reilly (46 comments total)

I prefer open source voting over diebold any day, but on the paper record thing, what's the big safeguard that everyone is battling for? I understand that paper record of your vote is better than nothing, but what good is your paper record if someone can hack the db and move your vote to someone else? Is it that your receipt would come in handy in a recount?

If we need a recount in a county, is everyone expected to save their receipt and bring them back if the city asks? I could see this making recounts a lot worse and elections go badly in the future, since I doubt many people would save them.
posted by mathowie at 12:09 AM on December 12, 2003

good for him. fuck diebold, their secrecy, lack of integrity, and intent to profit from their subversion of the democratic process.
posted by specialk420 at 12:13 AM on December 12, 2003

If we need a recount in a county, is everyone expected to save their receipt and bring them back if the city asks? I could see this making recounts a lot worse and elections go badly in the future, since I doubt many people would save them.

On a micro level, election monitors can have volunteer sample groups save their receipts for later verification. I'd sure as hell save mine. Even in your worst-case scenario, a nationwide recount is unlikely. A 100% paper recount of certain precincts in, say, Florida, wouldn't be that tough if the paper verification part was "opt-in," though I don't know how to deal with things like dead people voting.
posted by Ignatius J. Reilly at 12:15 AM on December 12, 2003

I'm just not sure it's the be-all-end-all solution to Diebold's problems.

I'm not a big open source guy (this site is hosted on windows, and my workstation runs XP), but for something as fundamental as voting in the world's biggest democracy, why shouldn't the software be completely open to everyone so the bugs can be all squashed and everyone can ensure a safe, secure election? When it comes to voting, I have to say open source all the way or no electronic voting at all.
posted by mathowie at 12:19 AM on December 12, 2003

The voter-verified paper trail works much like ordinary paper ballots today. Your vote is recorded electronically, but the machine also prints out a parallel paper ballot that you can look at to make sure your vote is recorded correctly. You turn this ballot in at the polling place, just like you would with an ordinary ballot.

If there's any reason to suspect that the electronic count is faulty, or if the election is close and there needs to be a recount, this is done using the separate but parallel paper ballots, which were verified by the voters as they were cast.

When counting these paper ballots, there would be no ambiguity about voter intent, since they would be machine-printed without any handwritten marks or punches. Potentially, there could also be some sort of bar-code or other security measure printed on the ballots to verify that they were cast during the original election period.

I know this issue has been talked about here before, but VerifiedVoting.org, BlackBoxVoting.com, and BlackBoxVoting.org are great resources to learn more.

The other good thing about Graham's bill (and Rush Holt's similar bill in the House, I believe) is that it sets a deadline of next November's general election for these voter-verified paper ballot machines to be in place.
posted by stopgap at 12:35 AM on December 12, 2003

What stopgap said. This is a bit of good news in a sea of, well, not-so-good news.

And, since I was just listening to it yesterday, here's a link to a This American Life show that talks about this a bit.
posted by stavrosthewonderchicken at 1:23 AM on December 12, 2003

Normally a company with this kind of reputation would be declaring bankrupcy about right now, but its deep connections with the GOP means they'll be around for a while and your state will probably be buying Diebold machines.

Soon we'll be hearing, "No one ever got fired for buying Diebold." Just like the apocryphal "No one ever got fired for buying Microsoft/IBM." When the opposite should be true.

Corruption, cronyism, quid pro quo, etc we've have it all here. The market should be eliminating reckless companies like these, instead we've got the GOP doing their best to keep them afloat. That's wrong, Americans of every political stripe should be demanding their state to divest from Diebold ASAP.

So what ever happened to this whole free market thing and voting with your dollars? Somewhere Adam Smith is weeping.
posted by skallas at 1:26 AM on December 12, 2003

Also, open source voting machines are a hit in Australia. It lends lots of credibility to the OSS argument when its been demonstrated to work.

"We'd been watching what had happened in America (in 2000), and we were wary of using proprietary software that no one was allowed to see," he said. "We were very keen for the whole process to be transparent so that everyone -- particularly the political parties and the candidates, but also the world at large -- could be satisfied that the software was actually doing what it was meant to be doing."

It took another year for changes in Australian law to allow electronic voting to go forward. Then in April 2001, Software Improvements contracted to build the system for the state's October election.

Software Improvement's Matt Quinn, the lead engineer on the product, said the commission called all the shots.

"They, as the customer, dictated requirements including security and functionality, (and they) were involved at every step of the development process, from requirements to testing," Quinn said. "They proofed every document we produced."

The commission posted drafts as well as the finished software code on the Internet for the public to review.

The reaction was very positive.

"The fact that the source code had been published really deflected criticism," Quinn said.

posted by skallas at 1:31 AM on December 12, 2003

"legislation that would require computer voting systems to produce a paper record."

it's hilarious -- first you design and build sophisticated electronic machines. then you figure out the guy who makes these machines may be a little too committed to a certain candidate (and you think hard about a few recent "upset" victories in Diebold-machine districts). so you crap your pants and ask for a law where those machines must actually produce a paper record.

a paper record like, say, a voter simply entering the voting booth and writing an "X" on a paper ballot, like in most countries

cool
posted by matteo at 2:32 AM on December 12, 2003

Matteo, yeah it seems pretty dumb. But given the financial investment in these machines already, they're not likely to just go away, no matter how rushed or ill-conceived they were. Having the machines produce a voter-verified paper trail lets you keep the benefits of electronic machines (fast counting, accessibility for people with disabilities) while having a reliable audit trail that's hard to tamper with.
posted by stopgap at 3:13 AM on December 12, 2003

Voting machines should be subjected to at least as stringent verification measures as slot machines. Rather than just complaining here, write your congresscritter and tell them that's what you want.
posted by Xoc at 3:14 AM on December 12, 2003

What matteo (and others) keep saying. What a huge fucking waste of money. Paper ballots, lots of volunteers. It's worked here in the UK for centuries. The only valid argument for electronic machines would be if they improve accessibility. Do they?
posted by cbrody at 4:57 AM on December 12, 2003

Ken from Diebold speaks!

The e-mail from "Ken," dated Jan. 3, 2003, discusses a (Baltimore) Sun article about a University of Maryland study of the Diebold system:

"There is an important point that seems to be missed by all these articles: they already bought the system. At this point they are just closing the barn door. Let's just hope that as a company we are smart enough to charge out the yin if they try to change the rules now and legislate voter receipts."

"Ken" later clarifies that he meant "out the yin-yang," adding, "any after-sale changes should be prohibitively expensive."

Now that's a company with ethics!

http://www.gazette.net/200350/montgomerycty/state/191617-1.html
posted by skallas at 5:49 AM on December 12, 2003

they really should be barred from supplying machines at this point...how many more things do we have to learn about them? (and how much are they contributing in every state and in congress and to Bush's campaign to make their machines the ones used?)
posted by amberglow at 5:55 AM on December 12, 2003

Amberglow - Sure, and even leaving questions of potential vote tampering aside, Diebold should be permanently barred from ever selling it's voting technology again - purely on the basis of it's recent record of atrocious incompetence.

Canadians I've talked to are proud of their paper ballot voting system - with national votes tallied within 24 hours. I don't know if this is always true or if the Canadian system always works so seamlessly, but the US reliance on computer voting machines seems like a bad cliche from Philip Slater's "Pursuit of Loneliness" critique of American culture : American obsession with technology often overrides all practical considerations.

Open source voting technology in Australia, paper ballots in Canada and the UK.....

Here in the US, you'd think computerized voting machines driven by secret, proprietary software were the sine qua non of democracy, somehow.

Paper vote counts can be tampered with, of course - so a combined open source software driven electronic voting system which produces a parallel ballot record could be the most robust, tamper free voting system of all.

As far as Graham's bill goes, paper ballots would simply restore basic, sound accounting practices (which - in accounting - are designed, through redundancy, to prevent random errors at least, if not intentional manipulation) to voting. Without any vote tampering whatsoever, mistakes always will happen - and without a redundant backup paper record it is impossible to verify voter intent in the event of random software glitches (common in Microsoft software products) or actual computer hardware failure.

And - I'm sure I'm beating a dead horse which now more resembles a flattened squirrel pelt on a highway, but still - If we cannot verify voting results, then the public trust in American Democracy will be deeply and perhaps fatally compromised. ( *slow drum roll* )
posted by troutfishing at 6:12 AM on December 12, 2003

Hey folks, click the link. This is about more than just paper receipts:
Key provisions of the legislation include:
# Requiring that all voting systems produce voter-verified paper records for use in manual audits
# Banning the use of undisclosed software and wireless communication devices in voting systems
# Requiring mandatory surprise recounts in .5 percent of both domestic and overseas jurisdictions.
# All voting systems would have to meet these requirements in time for the November 2004 election.
posted by Outlawyr at 6:12 AM on December 12, 2003

Confirmable audit trail? Yes

But Paper record = Horse and buggy

I think Diebold is wrong, but legislating a specific technology for the auditing process is just as wrong...
posted by geist at 6:57 AM on December 12, 2003

I can't understand why anyone would think this is a bad idea.

Would you deposit money in an ATM that didn't give you a paper receipt?
posted by bas67 at 7:11 AM on December 12, 2003

If we need a recount in a county, is everyone expected to save their receipt and bring them back if the city asks? I could see this making recounts a lot worse and elections go badly in the future, since I doubt many people would save them.

The "receipt" isn't for you to take home. The bill specifies:
"(i) The voting system shall produce a permanent paper record, each individual paper record of which shall be made available for inspection and verification by the voter at the time the vote is cast, and preserved within the polling place in the manner in which all other paper ballots are preserved within the polling place on Election Day for later use in any manual audit.
(ii) The voting system shall provide the voter with an opportunity to correct any error made by the system before the permanent record is preserved for use in any manual audit.
(iii) The voter verified paper record produced under subparagraph (A) and this subparagraph shall be available as an official record and shall be the official record used for any recount conducted with respect to any election in which the system is used."

So it's more of a "vote, double-check, and we'll store it for verification if needed."
posted by nickmark at 7:28 AM on December 12, 2003

"Paper record = Horse and buggy"? - yeah, that's the sentiment that Diebold is banking on.
posted by troutfishing at 7:33 AM on December 12, 2003

Hm. The linked statement on Graham's website indicates that it will go to the Governmental Affairs committee, but the Senate Record says it went to Rules and Administration (which would be more appropriate, since they deal with federal elections). That's a little disappointing to me personally -- since both of my senators are on Governmental Affairs, I was all set to do a ton of lobbying. But one of them is on Rules as well, so that's good.
posted by nickmark at 7:49 AM on December 12, 2003

geist, can you think of a secure, voter verifiable verifiable audit trail other than paper?
posted by zsazsa at 7:55 AM on December 12, 2003

A paper trail probably isn't the answer. This seems like a case where strong mathematics are needed, i.e., cryptographers. Each ballot really only represents a few bits of data. You vote for persons X, Y and Z, against everybody else.

The problem with a paper ballot is that there are two scenarios where falsification can take place. First, the tallied ballots can be tampered with might be caught by a random sampling. This is only true if the paper records aren't tampered with either though. Remember that either side can have an agenda and either side can be crooked.

So the data needs to be recorded such that any changes to the record in either event invalidates the ballot. This is actually possible at any given point in time even if the mathematics are entirely known. There still needs to be care that the protocols that implement the algorithm are trustworthy (that's the lots of eyes looking things over - the typical Open Source mantra) and that the algorithm is trustworthy (that's the cryptographers job through peer review - any secret algorithm should be treated as snake oil).

I'd throw a mathematician in the group to determine what the sample size would need to be to detect fraud as well. Maybe 0.5% is more than enough, maybe it's small enough a sample that wide spread corruption could go undetected. I don't know, but a statistician would.
posted by substrate at 7:55 AM on December 12, 2003

zsazsa today no, but 5 years from now maybe. Why should we be legally tied to a specific technology and have to go through the Congressional process all over again if technology produces an acceptable technical solution in the near future? If you create an opportunity the market will likely find a way to make it reality. You don't need to specify how. Just that there is a verifiable, auditable, and reliable system smart businesses will tell you how.
posted by geist at 8:05 AM on December 12, 2003

Diebold is only fighting this because they know Bush is unelectable and therefore needs to cheat.
posted by nofundy at 8:07 AM on December 12, 2003

About friggin' time. Why is this such an issue - a receipt should've been a standard part of the design of any such machine.
posted by FormlessOne at 8:15 AM on December 12, 2003

a paper record like, say, a voter simply entering the voting booth and writing an "X" on a paper ballot, like in most countries

Plenty of places in the US vote with paper ballots. Usually they're machine-readable, though. The ones I'm most familiar with have the various candidates' names (or the responses to ballot propositions) in a column with incomplete arrows (<--- ---) next to each one, and you complete the arrow you want.

Machine-counting gets used here because of the number of elections going on at one time. It's easy enough to hand-count in a precinct, under the watchful eye of maybe 5 candidates or their representatives, if all you're electing is your MP and maybe a local council.

In the US, it wouldn't be uncommon for a single election to involve the President, US House, US Senate, Governor, Lt. Governor, state Secretary of State, state Attorney General, state comptroller or treasurer, state Agriculture secretary or commissioner, state insurance commissioner, a state judge, another state judge, another state judge, another state judge, another state judge, another state judge, another state judge, another state judge, a state senator, a state representative, maybe another state representative, a state board of education member, a local judge, another local judge, another local judge, another local judge, another local judge, another local judge, another local judge, another local judge, another local judge, another local judge, another local judge, another local judge, a county commissioner (or maybe several), a county prosecutor, a mayor, a city councilor (or several), a local school board member (or several), and then a ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, another ballot proposition, and maybe another 10--30 ballot propositions.

Which leaves out other possibilities for sherriffs, tax assessors, idiosyncratic state offices (TX railroad commission, which regulates oil and gas extraction), and so on.

A fair number of states run state elections in the off-years (ie, 1998, 2002, 2006), but that just saves one office. And Texas seems to do most of its ballot propositions in special elections.

The point being that elections in the US are, in that sense, fundamentally dissimilar from elections in Europe and British settler colonies.
posted by ROU_Xenophobe at 8:27 AM on December 12, 2003

Excellent point, ROU_Xenophobe -- and in addition to all that hassle, there's no option in American elections to simply "vote the party" the way one can in Australian (and I imagine, other countries') elections.
posted by nickmark at 8:46 AM on December 12, 2003

What's up with the meme where people say Bush in unelectable and then link to his bio? I guess I'm missing the gag.
posted by McBain at 9:12 AM on December 12, 2003

nickmark: there's no option in American elections to simply "vote the party" the way one can in Australian (and I imagine, other countries') elections.

Actually, in Texas you can vote the party. I was quite surprised when I moved here and was presented that option.

And McBain, the Bush->unelectable is part of a googlebomb attempt similar to the 'miserable failure' one [type miserable failure into google and click the 'i feel lucky' button].
posted by birdherder at 9:22 AM on December 12, 2003

Ah... I see. In that case, Bush is unelectable.
posted by McBain at 9:27 AM on December 12, 2003

Well, even though Bush is a miserable failure, I don't know if I'd say Bush is unelectable....there are plenty of people in this country who are uneducated and ignorant, and he's the right candidate for them.

Also, in New Jersey, you can vote the party, also. I think that many states that have voting machines have a lever at the top to do it.
posted by taumeson at 9:51 AM on December 12, 2003

This is similar to the "paperless office" myth. Everyone thought we'd be fine with just computers -- yay! no more paper! -- but it quickly became very clear that we needed some sort of back-up in case the computer has a problem. Hence paper remains.

I'm surprised at the number of people here saying, essentially, we don't need a back-up. The revelations regarding Diebold make it abundantly clear that we do -- and paper is the most reliable means, unless you'd like to engrave ballots in stone.
posted by me3dia at 10:00 AM on December 12, 2003

Does this still work if I say that Donal Duck is unelectable?

In Illinois the old machines allowed straight ticket voting (please, no jokes about the Daley machine), but I can't remember if our hanging chad style ballots have that option.
posted by Outlawyr at 10:02 AM on December 12, 2003

Hm, interesting -- I stand corrected. At least in Minnesota, and Oregon when I lived there, it's not an option.
posted by nickmark at 10:20 AM on December 12, 2003

What matteo (and others) keep saying. What a huge fucking waste of money. Paper ballots, lots of volunteers.

The trouble with this approach in the U.S. is that here, we insist on holding our elections during the work week. The pool of potential volunteers is therefore limited to retirees and the unemployed.

Of all the things that piss me off about American democracy, this has to be #1. Why the hell can't we hold elections on a Saturday?
posted by mr_roboto at 10:29 AM on December 12, 2003

Xoc writes: Voting machines should be subjected to at least as stringent verification measures as slot machines.

You're referring to this story, I take it? This is a thoroughly praiseworthy decision by Nevada, and a richly-deserved black eye for Diebold. The latter can disparage the analysis of Johns Hopkins' computer science department if they like, but when in addition the Nevada Gaming Commission's slot machine analysts say that a machine is susceptible to corruption, there's not a lot of weasel room left.
posted by George_Spiggott at 10:36 AM on December 12, 2003

Why the hell can't we hold elections on a Saturday?

Working people might vote?
posted by Ignatius J. Reilly at 10:36 AM on December 12, 2003

in addition to all that hassle, there's no option in American elections to simply "vote the party"

You can, but only for elections that are partisan. In the nonpartisan parts of normal elections (usually some judges, some local offices, and ballot propositions) you can only vote the party ticket on the partisan part of the ballot, and in elections that are entirely nonpartisan you can't do it at all.
posted by ROU_Xenophobe at 10:43 AM on December 12, 2003

in addition to all that hassle, there's no option in American elections to simply "vote the party"

Thank you for pointing out a good thing about American democracy. What on earth is the value in a democracy of a vote by someone who doesn't give a damn? If you don't care to do your homework, why vote at all? We shouldn't be making it easier to vote, we should be making it -- in some respects -- harder. I'm opposed to discrimination on the basis of physical ability, language skills or what hours you have to work, but for me the ideal system would vigorously discriminate against wilful ignorance and apathy.
posted by George_Spiggott at 11:00 AM on December 12, 2003

This bill is pretty important, especially since electronic voting machines are mandated by federal law.

Ergo, without this bill, everyone must have electronic voting machines installed by (I think) the 2004 elections...and there will be no audit trail for the presidential election.
posted by dejah420 at 11:00 AM on December 12, 2003

Of all the things that piss me off about American democracy, this has to be #1. Why the hell can't we hold elections on a Saturday?

I second that emotion!
posted by nofundy at 12:17 PM on December 12, 2003

Why the hell can't we hold elections on a Saturday?
We should hold them on Saturdays, and malls and shopping centers should all be polling places. (maybe an extra 10% off at stores if you show your voter reg. card that day?) : &gt;
posted by amberglow at 12:50 PM on December 12, 2003

i always figured Saturday and Sunday were out of play b/c of Jews and Christians, respectively.

i *really* like this bill, btw. i'm a writin' my senators and congresswoman today!
posted by mrgrimm at 1:05 PM on December 12, 2003

instead of weekend voting, i always liked the idea of a national holiday.
posted by mrgrimm at 1:13 PM on December 12, 2003

Actually, around here the Mall is one of the early voting stations, open for a period of about 2 weeks before the Big Day.

I'm sorry if anyone disagrees, but this proposed legislation is absolutely necessary. Paper receipts verify the voters' intent and make an independent audit possible. This circumvents the arguments against electronic voting, to wit: how do I know the machine got my vote right, how do the election officials know the machine doesn't make up its own numbers, how do any of us know it hasn't been hacked by the Democrats/Republicans/Anarchists/Terrorists/Cylons.

The fact that this bill would appear to provide the ability of voters to correct a vote, and the provision of mandatory random audits of sample vote batches is strictly yummy gravy.
posted by ilsa at 2:27 PM on December 12, 2003