I want to go on record as being in favor of electronic voting

[Stinky The Clown]

Let's face it. Computers are probably the singularly fastest and most dead-on accurate number crunching devices out there. And lets also face the fact that they're here to stay. There is no good reason not to use them to help us cast, then to count, our precious votes. I know hand marked and hand counted paper ballots hold much favor (they surely do with me right now), but they really are, when considered dispassionately, outdated.

The problem is not the machines, per se. The problem is their secret, proprietary programming and the cobbled together system by which the votes are collected and tabulated.

The system I favor would have open source, federally certified, publicly owned code. The machines themselves can be proprietary. They're just dumb boxes until the code gets into them. They would, of course, also need some form of certification to be sure they can perform accurate basic number crunching, but that's about it, really.

These machines would also be capable of producing a paper receipt in duplicate. The voter would observe the receipt and sign a (separate) document stating that s/he a) voted and b) voted as shown on the receipt. The original of this receipt would go into a box and the voter would keep the copy. The receipt would have a way to be optically scanned in the event of some need for a recount. If we can check the validity of dollar bills in vending machines, we can do this. But the receipts could also be, literally, hand counted by actual humans with pencil and paper.

At each precinct, the machines could be networked so as to allow them to be polled locally from a central machine, but they would have no modems and no connection to anything outside the room in which they are physically located. Precinct totals would be certified in the presence of the official party-designated poll watchers. These totals would then be called in to the central tabulating office by human voice. No modem or networked transfer of this data would be allowed. The vote counts would be entered in a handwritten log and then entered by hand into the central tabulator. All such activities would be done only in the presence of official party-designated poll watchers.

There would also be a requirement for some random optical scanning and physical hand counts of complete precincts, at random, to back check the machine vote count's accuracy.

Think of the advantages of a reliable and trustworthy voting system. Rapid results. Fully auditable. The ability to have ballots in any language at the touch of a button. Pictures on ballots, if that seems the right thing to do. No over- or under-votes.

I'm sure I've missed some details, and I don't intend my amateurish attempt at describing a reliable system to be taken as a firm proposal - it is just a suggestion of what might be possible. My main point is, that even in the face of the overwhelming evidence that the proprietary machines have been far from reliable or accurate - and most importantly have shown themselves vulnerable to hacking and cheating - the basic concept is simply not wrong. We just need to have qualified people who are concerned with only the public good to design them and design in all the safeguards and capabilities we need to ensure that we all have confidence in the system.

Vote cheating has always been with us, and probably always will be. But it is my view that a computer based system can be made into the best and most bulletproof system thus far .... if only it is done right.

[Rabrrrrrr]

1. People counted, paper ballots only

That's my opinion.

Doesn't require electricity, nothing is hidden, and easier to recount. Longer, sure - but easier.

[ClassWarrior]

2. Paper and pen, hand counted. Maybe not fast, but verifiable by eye.

NGU.

[Stinky The Clown]

4. The system I described has a way to be hand counted and verifiable by eye

Why would that not be an improvement over pencil and paper?

[ClassWarrior]

15. It also has more insidious opportunities to be cheated.

Nope. Pen and paper.

NGU.

[ABaker]

34. Some disabled people

I know say pen and paper only is a violation of the ADA.

I don't know if that's true, but that's what they say.

If they're right, I think some sort of middle ground might have to be found.

[Rabrrrrrr]

38. So if you can hand count it and verify it by eye,

then why go through all the bullshit of hiring some jackass criminal software company to overcharge for software that doesn't work and requires equipment that the volunteer poll workers don't how to troubleshoot or work properly or that could go down in a power outage?

Jesus.

[The Wielding Truth]

39. YES

[sinkingfeeling]

3. I'd rather have the technology of adding machines, which cannot be

'programmed'. Paper ballots, hand counted at the precinct by a panel made up of a rep. from each party in the election and observed by public. Each ballot could be read aloud by each member, so everyone agrees with that vote..then it is added to the candidate's total. Precincts take their certified count to county. County panel that tallies up all precincts. County takes certified totals to the state. State panel totals all county votes.

[Stinky The Clown]

5. In the event of any hint of chicanry, the system you described .....

could be done with the system I described. We lose nothing by doing it the way I described. Or have I missed something?

[msongs]

6. will your machines run without electricity? nt

[Stinky The Clown]

8. Can you fill out a ballot in the dark?

That's just silly.

[ClassWarrior]

17. Yes. Candles.

Occam's Razor.

NGU.

[Stinky The Clown]

23. Okay, this subthread is just getting silly.

[ClassWarrior]

25. Occam's Razor is anything but silly.

NGU.

[JAYJDF]

7. What's up with all the registering to vote problems? I like the idea

that everyone can vote. Come in and vote, dip your thumb/finger into that purple ink that will last for weeks. Probably get more voters. Sure maybe, just maybe one/or a few might be able to figure out how to vote twice, big deal. Out of millions, no problem.

[Donald Ian Rankin]

46. Because some people don't have the vote.

It's generally agreed that the very young, foreigners, and I think certain categories of the insane and criminals although I may be wrong about those, should not be allowed to vote.

That approach is also unworkable for any ballot organised on less than a national level, because it lets people choose where they vote, which can make a big difference - a vote cast in a "safe seat" is worth less than one in a marginal.

[eppur_se_muova]

9. Why do you think speed is so important? It never was before.

Originally, it was only the broadcast networks who were responsible for the pressure to call the elections quickly. The Republicanic vote-riggers used this as a plausible excuse to "overhaul" the mechanics of voting, and snuck their fraudulent software in thanks to lax (nonexistent) oversight. No one else's interests are served by increasing speed at the risk of fidelity.

I will have to disagree with you, firmly. Fidelity should come first and foremost. Speed is a luxury, nice when we can get it, but we can easily live without it (and have, for over two centuries).

[Stinky The Clown]

13. Please make the case that an open source ...

.... publically owned system is less reliable?

Look, I'm on your side. I really am. But I can't help but see technological advance - with a complete ability to sit there and hand count by eye and voice each and every vote - as an improvement.

What are your *specific* objections to the system I described? I don't see where it allows for *any* less fidelity than a pure hand system.

[endarkenment]

10. I want to go on record as being in favor of hand counted paper ballots.

What exactly is the compelling reason for any other system? Are they more secure? Less expensive? What? Oh they are faster? So what?

[eppur_se_muova]

11. Seconded, and call for votes. nt

[Stinky The Clown]

14. What do you mean by 'call for votes'?

I understand that you're endorsing the idea of the poster to whom you replied, but I don't understand what you mean by 'call for votes'.

[eppur_se_muova]

16. Post below if you agree.

[Stinky The Clown]

19. I still have no idea what you're talking about

Not being snarky, but your terseness is hard to understand.

[unpossibles]

12. well, it seems an unpopular choice

but I agree with you as long as it can be kept clean.

[Stinky The Clown]

18. Unpopular doesn't mean wrong .......

.... sometimes it means it simply isn't conventional wisdom.

[SaveElmer]

20. The paper receipt is the key...

I think if that was required the opposition to electronic voting would largely disappear!

Here in Fairfax COunty, VA Republicans and Democrats are both pushing for this (surprisingly)

[Stinky The Clown]

24. That's the crux of it, isn't it?

I agree with your view. If we have paper print-outs that are legally the 'ballots of record' and if these are checked and attested to by the voter, how are they any less reliable or true than a paper ballot marked by hand? They're not.

These paper print-outs that form the 'ballots of record' are treated much like we now treat ballots, can be easily counted by some electronic device OR by eye/hand. As I see it, we give up nothing and get a system that's easier and faster.

[derby378]

21. "Publicly-owned code?" Ain't no such thing

The corporations won't let it happen. It's paper or fight for me.

[Stinky The Clown]

26. Of course there is such a thing

The government pays to have the coding done and then owns the code.

Who says the corporations won't allow it? They may not want it, but they really don't have the control. I know .... things today are pretty bad, but if one is intellectually honest, one has to agree they do not have the power to prevent this. Now ..... our elected reps could be another story ......

[derby378]

29. Under a Democratic administation, you may be right...

...but I guarantee you that any attempt to create a publicly-owned code under these jackals will be privatized and then hidden from public view faster than you can say "Halliburton."

[Stinky The Clown]

31. Under an *honest* administration I'd be right ......

..... and we'll either get one of those or we'll no longer need to even worry about the 'quaint' act of voting.

[Rabrrrrrr]

37. Yeah, right. The government pays some asshole software company

and, like the University of Wisconsin - Madison that bought a $2 million software system for $36 million after the software company finally decided to stop fucking things up - the government pays $2 billion for software that's buggy, unreliable, and has to be scrapped so that at the last minute the paper and pen ballots that are hand-counted are brought out.

Why do you have to make it so goddamned complex? Occam's Razor IS the proper system here - go the easiest, most reliable route.

Why do you insist on inventing a 900 HP Jaguar with a stereo system that cuold be used in a concert hall with tires that will handle sharp cornering at 190 mph, has radar-proof paint on it, is a convertible, and can accelerate fvrom 0 to 100 mph in 0.07 seconds when all you need is a pair of shoes because all you're doing is crossing a goddamn country road to turn in your coke bottle for the 5 cent deposit?

Fuck.

[Stinky The Clown]

40. Good rant

Crappy, meaningless debate. But whatever .....

[meldroc]

22. I can think of better ways.

For one, I don't trust electronic voting machines. Being a software engineer, and having followed the Diebold scandal, I know it is far too easy for a technically knowledgeable person to compromise the machines and steal an election.

Computerized voting machines such as touch-screen machines and other machines do have the advantages of speed (as mentioned) and of being accessible to the disabled.

I would be in favor of setting up a system as follows:

Have a touch-screen/accessible electronic machine with a shiny, friendly GUI that is completely non-networked, and comes with a printer. All this machine does is accept the voter's input and print out a paper ballot. The paper ballot would be similar to the optical scan ballots used today in many places (aka a fill-in-the-bubbles scantron ballot, with the bubbles automatically filled in for you.) The voter would then read the ballot, verify that the ballot has the correct votes on it, then put it in the ballot box.

Once the polls close, the ballots are counted. First round of counting is by optical scan machine, to count the ballots quickly and accurately. If an election is close (less than 2 percentage points, or statistically insignificant), or there are any sort of challenges to the count, including a public petition with enough signatures, then the ballots are recounted, by hand (no you can't just run them through the machine again), in a place where the public can watch. Observers from all political parties will be present, and the final count will not be submitted until all observers agree the count is accurate.

Basically hand-counting, but with automation where it makes sense.

[Stinky The Clown]

27. I don't see where your system and mine are very much different .....

We both have essentially the same system .... initial count is electronic and all recounts are by hand. Voter verifies the 'offical ballot' before it gets counted.

On edit .... all the big problems seem to have been in the central tabulators, not the machine that we see/use when we vote. Both your system and mine break the electronic tie between the voting machines and the central tabulator. Mine makes the central tabulator data entry a public process with a hand written record.

[meldroc]

35. A couple of things.

One, your system simply provides receipts, but records & counts the votes electronically. At least there's a good paper trail, but I prefer to go one step further and make the receipt the actual ballot, and remove a few ways for the touch-screen voting system to cheat.

Going open source is definitely a must - no more proprietary Diebold cheatware. All machines must be installed from a certified binary image compiled from the open source code. Not only that, but the compiler, build tools and all software installed on the reader or used to create the software installed on the reader (gcc and other existing GNU tools will do, provided they pass a security audit & survive the scrutiny of security experts) must also be certified. The code must be audited by the state, third parties, auditors representing the political parties, and anyone who wants to download the code and look at it. The public and representatives (preferably software engineers with skills in security) should be watching every step of the build process to ensure that the code isn't substituted with corrupt code, or the compiler and build tools aren't replaced with ones designed to generate binaries with security holes or backdoors. Once the image is certified & burned to CDs, only that image is allowed to be installed on voting machines. Cryptographic hashing must be used to ensure the binaries match those on the certified image, and auditors should be able to check them at any time.

Even with these precautions, it is possible that unscrupulous individuals may introduce backdoors into the voting systems, or an unintentionally introduced security hole may be exploited. Switching to hand voting and hand counting must be an option at any step of the process. Ballots & paper trails, as well as data files used by the systems must all be human readable

I also don't like the idea of giving the voter a receipt with his votes recorded on it to take home - that violates the prinicple of the secret ballot, and opens the door for all sorts of corruption. You'll have thugs working for various unsavory organizations checking the receipts of exiting voters and breaking the kneecaps of those who didn't vote for their candidates.

I do like your tabulating processes. Use the tabulating machines, but human-verify the intermediate counting steps, and be prepared for hand counting.

[Stinky The Clown]

41. Than you for adding some considered thought to this thread

As I said, my attempt was amateurish at best.

I wonder if, instead of loading the software from CD, the machines could only use hard coded firmware ..... chips with no way to change them ... like pre-flash bios chips (whatever these gizoms are called). The only thing the machines could store in a volitile memory would be the office name, the candidates' names, and the votes.

[Awsi Dooger]

32. Extra machinery and extra steps, so it won't happen but I really like it

I never thought of that possibility before. The conventional wisdom post-2000 was electronic machines would help us, eliminating human error via filling out punch cards or optical scan ballots. Your solution essentially allows the machine to fill out the ballot, then sidesteps the potentially fraudulent aspect of the electronic machine providing the tally. Intriguing.

meldroc, you should copy that suggestion and start a new thread on the Election Reform forum. I'm sure it would be well received. Or if you don't want to do that, there is a recent thread titled something like, "Possible compromise, hand counts and optical scans." Your post would fit in well in that thread also.

[eridani]

44. You've just described a very expensive disenfranchisment machine

You've conflated the act of voting which can be very time consuming, with the much faster process of tabulation, slowing the whole process down to the speed of the slow step. One lunch-hour voter reading and re-reading all his campaign literature and checking everything 20 times could disenfranchise dozens of people who have to get back to work.

[AtomicKitten]

28. Until trust is restored,

paper ballots and pencil, with precise protocol for counting is mandatory.

And since trust will never be restored - and for good reason - the above works for me.

[Stinky The Clown]

30. Either trust or its common substitute ........

..... apathy .... will, indeed be restored.

[Zorra]

33. "An open and honest examination of the flaws in electronic voting

will lead us to only one possible conclusion: electronic voting machines are dangerous to democracy because there is no way of ensuring their accuracy" -- Dennis J. Kucinich

I have to agree with Rep. Kucinich - IMO, the possibility of instantaneous vote switching on a mass scale is just one drawback that makes computerized voting unacceptable in a democracy.

[politicat]

36. I'm with you on this one.

I've worked the polls. Our ballot is a double sheet of 11x17 paper, densely printed on both sides, that we have to mark with pens. Bubbling the thing in takes 25 minutes, reading and bubbling it can take up to an hour. That's insane, and it makes a mockery of secret balloting for the blind (who have to have someone else fill out their ballot for them, and trust that the ballot is filled out as requested) those who cannot stand for an hour (not to mention the waiting time), and other people with disabilities. Heck, my only problem is a couple missing ligaments in my knees, and I felt tired and in pain after voting last time! (And not just because the candidates were losers...)

As long as we continue to expand the ballots, we're going to have to improve the way we count them. We had over a hundred issues in just my county in 2004. That's a massive amount of time to count. We have optical scan, voter verifiable, hand counted ballots and they did us no good - we still had problems with how ballots were counted because there were so very many of them. The 2004 election cost the county several million dollars when all was said and done, money we could have used elsewhere, and for that matter, to buy a bunch of computers, put a linux based OS and open source vote tabulation system on it, add printers, and train a couple hundred poll workers on how to keep the machines running... and still have had money left over for parks, open space and poverty abatement!!

Pen and paper, hand counted ballots are not the answer to election fraud. Weeding out those who would like to "fix" elections and those who are incompetent to be election workers is the answer to election fraud.

I don't mind if it takes a couple weeks to get election results, but in my county, it took several months to certify the election. That's too long and causes far too many problems because government has to continue to run, and if the mayor, school board and Clerk's office are all shut down because the election results can't be verified, then everyone's hurt.

[jayctravis]

42. There is no reason computers can't be used to speed up the

voting process. The problem is how do you know what goes in comes out, and how do you know that they aren't using simple human nature against us? Calibrating the screen so that one candidates "touch area" is bigger than another so it's easier to vote for them, or programming it so touching one candidate "accidentally" reads as the other one. If the person thinks to double check it will let you vote for them, but 90 percent of people won't.

I came up with a voting method that I think would be the best. Use electronic voting machines that show you a serial number at the end of your voting. This number can be written down, or the machine could issue a printed stub with the code.

After the votes are tallied, a spreadsheet of every single vote is published for public viewing on the internet on a .gov site and mirrored at several other places. This spreadsheet would also be published in an almanac and stored on microfilm in libraries. Anyone who feels like it can search for their serial number and verify how their votes were counted. The votes would be public, but it would be very difficult to identify how a specific person voted unless they gave you the serial number of their ballot. Essentially *anyone* would be able to recount the ballots on their own if they felt the need to, and this would be the official record of the results.

Simple and easy, the voting machines just essentially are a database, and the results are made public immediately after all the polls are closed. If a voter sees a mistake on their ballot they call an 800 number, put in the serial code and their voter registration card number and officially dispute their ballot...which basically just marks it red on the website. If the the number of disputed votes grows to equal the same margin a candidate won by before a certain period of time, it triggers an automatic revote.

[endarkenment]

45. Why is speed important?

It seems to me that verified accuracy is the most important quality of a voting system. Electronic system make verification difficult and expensive. Speed is important to the media, it sure is less important to me than knowing that my ballot, as I recorded it, has been accurately counted.

Why should my town buy an expensive unverifiable systems when we have a cheap verifiable paper based system right now?

[eridani]

43. It's flat out not possible for touchscreens to EVER be reliable--

--simply because they aren't used often enough. No complex machinery that is stored rather than used for most of its service life can be expected to function properly, ever. It can't happen, period. Think of how badly cars would suck if people only drove them for a couple of hours twice a year.

No complex piece of software (open source or not) can ever be free of bugs either. With word processing programs, spreadsheets, ATMs and the like, the user base is huge and the programs are in constant use. Therefore, any bugs that compromise function seriously pop up quickly and can be patched due to lots of real-time user feedback. The programs are never bug-free, but they eventually become robust and stable. This process just plain can't happen with infrequent elections and no real-time user feedback, and it is flat out UNACCEPTABLE for our elections to be beta tests, period! No compromise on this, ever!