Slashdot Mirror
E-voting Trials and Tribulations
Alex Susor writes "This article is about the new digital touch screen voting system in Georgia, the first state in the nation to adopt this method of voting statewide. Demonstration machines were set up at the recent primaries to teach voters about the new system (to be in place for the November general election) and had some big problems." Compare and contrast to systems in Florida and Germany.
See the smudges from other people's fingerprints?
Depends... When's the next election?
you should really first develop a paper system that nobody challanges to not work properly, it's not that hard. power shortage and s*** is bound to happen somewhere even with ups. + the (illusion) of real privacy goes straight out of the window.
world was created 5 seconds before this post as it is.
Comment removed based on user account deletion
As long as there is a talking paperclip at the bottom of the screen to help people out, I think everything will go smoothly.
this was an interesting quote: In Fulton, poll workers also reported the machines mysteriously switching from demonstration mode to election mode, Champion said. But state election officials and the company that makes the machines, Diebold Election Systems of Ohio, said that's virtually impossible and instead suggest untrained workers were to blame.
Exactly. It goes without saying that remote voting of any kind will NEVER exist, and even this system is way too easy a target for cries of foul from the losing parties. Anything from sabatouge, hacking, or even something as benign as a short-circuit of the system could happen, and the losing candidate will bring up all 3 if given half a chance, rest assured.
It just leaves the door wide open on challenging a vote, whether you believe someone lost fairly or unfairly due to the machine. I consider myself a reasonable person, and I would challenge it if I lost in a second.
Finally, math books without any of that base 6 crap in them.
Until advertising is sold on the kiosks..
until pop under ads for the X10 camera appear
never ending pornsite loops to entertain grandma (since young adults don't vote.. I know.. I waited in line to vote last november, and was saddened by the turnout.. I was the only one under 40 it seemed)
I am the lord of the pun. Dance Knave!
Stupid sexy Flanders.
I don't want to leave my house. Why can't I vote over the internet?
/. population agree's with me. Are you listening politicians?
Just mail me my username/password, i'll go to whatever website you want me to go to and vote. I'm sure 1/2 the
-T
The nice thing about digital voting is that you know that there's a problem with your vote (a frozen computer screen, etc.) before you walk away from the booth. With the current system, how are you know if your chad is punched all the way? ;) Coding errors can be debugged. It's great to be able to _see_ the problem.
has power."
- Joeseph Stalin
With a computer voting system, there profile of risk for election fraud changes so radically that the folks used to policing these systems will never know what hit them.
We've already had one US election stolen by outright electoral fraud (I'll let y'all verify that Gore won from your own preferred, trustworthy news source).
This just opens up the door for more trouble ahead.
Hexayurt - open source refugee shelter,
I don't mean to sound like a Luddite, but I'm not sure technology is the best solution in a situation like this. Technology is great for many uses, but for a task as simple as voting, it is much easier and more practical to simply use existing methods which have been proven by their use in the past hundreds if not thousands of years. Voters who are not computer savvy will likely become confused by the unnecessary complication of the new voting machines and many are likely to cast their ballots in error, possibly voting for a candidate they had no intention of supporting. Clearly, in a situation such as this, current paper voting mechanisims are much more accurate and reliable. Furthermore, if voting is to be computerized, we're leaving ourselves vulnerable to all sorts of hacking and digital manipulation of the ballots which otherwise would not exist. It's been said many times here before that no computer system is 100% secure, and I, for one, do not want to trust my country's elections to the likes of Microsoft of Red Hat. Paper elections are much harder, if not impossible, to tamper with.
-atrowe: Card-carrying Mensa member. I have no toleranse for stupidity.
I wonder if it's really legal to have votes counted by a machine that has secret software inside that voters are not allowed to examine?
Shouldn't voters in Georgia be able to file an FOI request to find out what's happening to their votes?
I usually wouldn't bash windows but this is not the right solution for this. Why would you rely on such complexity for a system that is supposed to be simple and easy to use? Just imagine how easy it would be to break into this system and change it.
A better solution would be to use an embedded microcontroller or other simple hardware device for each voting station and then connect that to a central database server running a much more secure operating system. I think that voting and it's integrity deserve as much mission critical attention as safety systems in an automobile. There simply shouldn't be any failure here. Relying on an OS with several millions of lines of code just to input a few votes just doesn't make any sense.
What really matters is that they use Instant Runoff Voting; please see:
The Center for Voting and Democracy
the Instantrunoff mailing list
and the California Instant Runoff Voting Coalition for an example of a good local activism site.
P.S. You can create your own web-based IRV web surveys with DemoChoice.org (also includes free downloadable php scripts for your own site.)
so now our election system is run by microsoft, yay! now because it's closed source, couldn't microsoft run a service in the backround that changes the vote tallies? or even some of the techs working on it. techs need access to the basic parts of the system, and im sure one could change the number of votes, it has to be stored somewhere outside of the ram. if its stored in ram and the power goes out, the election's screwed. there are so many places where this can go wrong it's sad.
yeah, and a tetris for those who haven't decided yet and still need some time.. or quake! There are some possibilities.. with skins from the politicians.. "Humpz.. that bush guy fights like a wimpy girl.. He's not gona get my vote!" And of course a random-choise button for those who realy don't know or care.. .if you're the thirt in a row to cast the same vote you win! The possibilities of getting people actualy to vote!
Hmm.. Maybe a link with a casino.
Nobody expects the spanish inquisition!
This is good stuff.
Hexayurt - open source refugee shelter,
But what happens when all that oil builds up and blocks my keypress???
Ok, if it works like this, can't you do this remotely as well, it should be rather easy make the conduct just before the real vote is given. Then just, voilâ and thank you!
Knowing the average voter ;))) no-one will notice :)
It would be nice to see someone with more knowledge write about his :)
Though they blame the mistake on the Windows OS for crashing, it's stupid to believe them. Basically whomever decided to push this out there didn't test their product enough. Everyone who has used computers for any length of time realize that a closed system like this should have zero problems if properly tested no matter what the underlying software is, be it windows, linux, HP-UX, or mac.
What does this really mean? That the voting system should go back for yet more testing. QAing software is probably the most boring part of the job, but it's also the more important. If we are to even pretend that we live in a fair society then any voting system should work and work fairly. Be it paper or computer based.
Is America ready for a computer based system? I think a computer based system should be able to replace a paper based system. I think that possiably we should also use paper in addition to the computer system, meaning that they should actually print a reciept of your vote so that in the case of a recount, they have physical proof that you voted for (Gore and not Bush?) the person you said that you did.
This is really frightening. The poll workers couldn't attach a power cord to the PC? That is a really basic interface, and we then trus them to operate the mechanical systems that drive the legacy election process?
I suppose that one possible issue is that there might not have been outlets near the voting booths...
But Herr Heisenberg, how does the electron know when I'm looking?
The next president of the United States: {FATAL EXCEPTION IN 0x0E4F}
Instant run-off voting is a step in the right direction, but it too is still leagues away from being able accurately representing the will of the voting populace. What about Approval voting? It is just one of many options out there.
Remember "Bring 'em on"? *sigh
The Vivendi hijacking seems like what this system is waiting for. http://www.businessweek.com/technology/content/may 2002/tc20020521_3291.htm
Bill, can you factor this prime number for me?
Compare and contrast to systems in Florida and Germany
Germany = state of the art open source based system
Florida = unauditable mystery box system
No surprises here, I would expect such systems in America's 'joke' puppet government owned by corporations verse Europe's 'real' and refined governments.
GoatPigSheep, the 3 most important food groups
What a sad commentary, that home computers should be the obvious metaphor for an unreliable piece of junk.
Sheesh, evil *and* a jerk. -- Jade
It had a touch-screen with a display that was probably about 18" high and 9" wide. There was a card reader to the right where you inserted your voting card. I'm not sure how the distribution of the cards will work. I don't know if they will issue every voter a card or if you get the card when you go to vote. It looked like it had a smart chip on it instead of a magstripe.
The user interface was pretty easy. It would present one or more categories and all the candidates for each category. You just touched the one you wanted. Once you selected a candidate, it greyed out the others. It took me a few seconds to figure out that if I changed my mind, I had to touch previous selection to undo it. There were "Next" and "Previous" buttons to navigate through the various pages.
At the end, it showed a summary of my votes so I could give a final yes/no to my choices. It printed out some kind of receipt, I think, but I didn't really look at it.
If I had to guess on the platform, I have to say that I did see an hourglass icon that looked just like the one in Windows. Maybe they're running WinCE or something. It looked a lot like one of those "pen computing" devices that never really went anywhere.
I would probably feel a little more secure about the system if it printed out a ballot that I then had to put in a ballot box, so it wouldn't be any worse than what we have now (from a fraud standpoint). It is certainly easier to use than the punch ballots we have now.
You won't have to imagine it in MS-Election 2004!
If you read the article, you will see quite clearly that Windows is not the only thing blamed for the errors. First and foremost were the problems by the people running the polls. Apparently the ones in the county that had the most problems (only one county, not all of them) hadn't had all the training necessary to operate these machines.
Also note that the machines that had the problems had not received the most recent updates from the vendor, whereas the machines that worked well did have the most recent updates. So the fact is, the company must be doing some good QA work to get the upgrades ready in time, but upgrading machines across a whole state, and training workers across a whole state, takes time.
And lastly, the person who blames the problems on Windows was not a spokesperson for the company; he was a spokesperson for the secretary of state's office. I highly doubt he's qualified to make any sort of pronouncement as to the technical cause of these problems. Hell, he may have just been told that there were problems with the unpatched versions of the software running the machines, and assumed that the software running the machines was Windows. Nowhere do I see an official for the company that makes these machines blaming Windows, although I agree with a previous poster in that Windows is probably overkill for a situation like this.
Also, I think a receipt would be a good idea, with both the voter and the polls office keeping a copy so a manual count can be performed if necessary. That would make a good intermediate step before going to a totally paperless voting system.
Yeah, this was well thought out. Remind me what happens when the power goes out?
I voted (my first time as I am 18 yay!) in the primary and got to see one of these thing first hand. Very nifty. The interface is VERY clean, and it takes no time at all. The old method however was to basically fill out a form similar to a scantron sheet (darken the bubble...) where you constantly wondered if it was dark enough. I'm all for faster, more accurate voting counts too. Talk about instant results! Anyhow, again, I liked, everyone is entitled to their own opinion.
Derek Greene
> Assuming this new system is completely secure, there will need to be explicit instructions and examples to ensure even the most brain-dead person can cast their vote.
And auditable enough to ensured that dead-dead people aren't casting their vote.
> With some luck we won't see a repeat of the election insanity we saw in Florida and elsewhere.
Unfortunately, that election insanity has merely brought out the snake oil salesmen.
The general rule of thumb for understanding the USA is that whenever something goes wrong it is eagerly embraced as an excuse to do lots of other things wrong. Particularly if it can be used as an excuse to feather a businessman's bed. (The reader shouldn't have any trouble thinking up lots of recent examples.)
Sheesh, evil *and* a jerk. -- Jade
The nice thing about the above system is that, except for the printing, it could be prototyped by anyone with undergraduate EE skills. The circuits are dead simple, it adds speed to the counting process, it allows for double checking of results, and it doesn't require the voter to learn new skills.
With all those things going for it, there must be something wrong with the idea.
=Brian
There is nothing so good that someone, somewhere, will not hate it.
First and foremost, we have to remember that this is a government venture. What that means is that you have to lower your expections by about 60 IQ points. I theorize that it might go something like this:
Step 1: Acquire, pay for, and install thousands of new, electronic voting machines. Ignore the obvious, like the inability to audit the manner in which votes are tallied and reported by the software.
Step 2: Experience initial problems during a "demo day" held at some point before the election. Disregard the notion that this may very well be the beginning of a very bumby road.
Step 3: Use the newly-acquired machines during the next election, experience more problems, and be sued by a public interest group questioning the results, and demanding a detailed audit.
Step 4: Be dissed by the company that manufactured the machines, who claims that disclosing the process by which votes are tallied and reported would result in disclosure of proprietary trade secrets.
Step 5: Be backed into a corner. Wonder why no one took this issue seriously during the initial planning.
Step 6: Scrap all 19,000 voting machines, kissing the $millions they cost, goodbye. Replace them with machines from a company with a more open disclosure policy.
Step 7: Lather, Rinse....but hopefully, avoid repeating the same sordid tale over again.
This was settled 2 years ago. BUSH WON! Get over it.
Fine. If the Bushies and their ilk will stop blaming everything on Clinton!
Fascism starts when the efficiency of the government becomes more important than the rights of the people.
Ireland has a reasonably complex voting system. Each voter has a single vote, but can vote for several candidates in their order of preference. (Each constituency has between three and five seats). Even given that complexity, the system seemed to work well. There was about the expected turnout in each count center, and there were few concerns expressed about the usability of the system.
Some info is at http://www.environ.ie/electronicvote.html.
I've got flamebait 2, and I'm trying to raise a serious point here: electoral fraud is a real concern in US elections, and computerizing the process is not going to make it more transparent.
Moderation is supposed to be for post quality, not for political or other content.
Hexayurt - open source refugee shelter,
I voted early in Colorado Springs in 2000 at one of the local malls
Californina had some counties that used them as well. IIRC some states on the east coast had experiments as well.
You are right, this is old news.
Pen and Paper. Clearly marked.
Easy to use. Easy to count. Easy to Spot Errors. Easy to Secure.
Dosnt even need elctricity.
Why isnt it used?
I would love to see this implemented nation wide. I don't think it will happen. Too many people in power have too much to lose from an IRV system.
Finding God in a Dog
1. We have a large network of vendors for the National Lottery... there was talk a while back about using this system for e-voting. It's secure, handles large numbers of transactions, uses proven technology, and each machine is capable of scanning hundreds of lottery tickets per hour. Most people in the UK knows how to fill in a lottery ticket...
2. It's extremely easy to get multiple votes in the UK. My girlfriend received two voting cards for the 1995 General election, and could easily do so again... so any electronic version surely must be better than the current mess.
The more advanced the technology, the more open it is to primitive attack
I had been advocating IVR without knowing what it was called. I found a very explanatory Flash demo of how it works at http://www.chrisgates.net/irv/
Kind thoughts do not change the world
1) Be wealthy
This is enough in the US. Step 2) and 3) or not required.
What I like about this idea is that it could finally allow the US to move in the direction of being a democracy instead of a republic. On the other hand, people are dumb in large groups and I know everyone would vote for a "gimmee" without considering the consequences (sure, let's ALL get $5000 tax rebates this year)...
I do find it odd that they are using a Windows OS though. Maybe they wanted an excuse for a recount? [/sarcasm]
Honestly, I don't see an electronic format as being a bad thing. In the future, we might actually get to vote on more issues and take more direct control of our government: wouldn't that be a plus?
Murphy was an optimist.
Great, a proprietary system recording my vote. I have no way to audit it for correctness. Even if it was open source, if a problem is found, there is no way to recount.
Any voting system needs to be auditable and recountable. My local county (Dane County in Wisconsin) has a great system. The ballot is a big piece of paper with a broken arrow next to each candidate. Something like this: President
Albert Gore (Democrat) <-- ---
George Bush (Republican) <-- ---
Ralph Nader (Green) <-- ---
You use a provided pen to complete the line pointing to the candidate you want. You then take your sheet and feed it into the locked tabulating machine. The machine refuses your ballot if there are obvious errors and you're sent back to try again with a new ballot.
The result: The interface is easy for anyone to understand. The tabulating machines make it possible to quickly generate tallies. The system is auditable since you can randomly hand count the ballots in a particular machine to verify the totals. In the event of problems, you can simply hand count the easy to read ballots (unlike trying to read holes in a punchcard).
Unfortunately shiny computer screens are easier to sell that boring grey boxes and paper ballots.
Chris Riggall, a spokesman for the secretary of state's office, attributed the problems to errors by poll workers, a glitch in the Windows operating system that runs the machines and problems with electronic cards that replace paper ballots and ballot boxes.
I wonder if they got some BSODs on those Windows boxes. I bet they are just locked up e-Machines running IE connected to IIS/ASP in the back room. Gimme a break. Maybe we need to have a vote on how to vote.
What a bunch of idiots, IMHO. Can't they figure out a way to use a system similar to what people have used for YEARS and just remove the issue of hanging chads, etc?
Now they want Granny and the other old farts to deal with touch screens and the likes. What happens when they touch two places at the same time or leave a hand on one part of the screen?
One step at a time folks. At least lets change this with the older generations in mind. Aren't they the ones who started this when a modified ballot style was used?
I mean really. Insert a digitized pad with LED's or something. If you push a button the LED will show you that you pushed it correctly and track your vote. When you're done, there's one last selection( done / not done ) and your personalized iButton won't eject til you select DONE. You won't be able to exit the polling place til you hand over the iButton.
Now who is behind all this new fangled voting system anyway? Some business is surely pushing it....
LoB
"Anyone who stands out in the middle of a road looks like roadkill to me." --Linus
Maybe not an obnoxious Clippy offering to "help" someone vote for President but just information buttons.
Not sure about an intiative? Click here to the entire text of it along with submitted Pro/Con statements.
Who are these candidates? Click here to view statements from their campaigns.
I don't see any reasons why the electronic voting machines shouldn't be able to include *already existing* information from the voter's guide.
Well from what it seems to me (though it could be happening the article did not mention) is that this whole thing is missing something.
This something is called Validation and Qualification. These 2 processes are used in many types of industries (espicially pharmX). Basically the concept is to evaluate and log everything ever done to every system.
There are checklists that should be followed to ensure each system has an identical setup (these checklists are so detailed they tell you where to click).
The setups are then tested and re-tested for consistency and accuracy. Once a system has been qualified and validated, any change no matter how small to the configuration must be logged and the machine must undergo another validation process.
Through all this an accurate paper trail is kept with each change, update or install.
This way if anything is ever challenged there can be records showing what goes on with the machines and thus either disproving or proving malfunction. This would allow us to avoid the whole problem with debating what could hav happend and track down the real issue if there is one.
This way a canidate cannot "wait out the clock" so to speak.
-I am too lazy to spell check today
"Don't mess with him, he taunts the happy fun ball."
Actually, I've seen a BSOD on an ATM... It's been a while, but I think it was at a Wachovia bank. I don't know how common it is, but some ATMs apparantly use (or used) Windows.
Bushies and their ilk will stop blaming everything on Clinton!
He's blaming forest fires on Clinton? I thought Dubya was blaming trees that he wants to cut down.
--
"Outlook not so good." That magic 8-ball knows everything! I'll ask about Exchange Server next.
It would if the database(s) is(are) irreparably corrupted or otherwise compromoised. Then there is nothing left to count at all. Far far worse than what happened in florida.
Contrary to popular belief, coding is not all free blow-jobs and beer. Those things cost MONEY!
You can find some information on better solutions (particularly approval voting and Condorcet voting) at Electionmethods.org, including and explanation of why Instant Runoff isn't a much improved voting system.
Hahahah...they act astounded that the computer crashed 11% of the time. DUH! They're running MS Windows.
If we're going to have computerized voting, it should be done using a Linux or *BSD OS. This way, you won't get crashes. Also, any software which the government uses or is used by citizens interacting with the government should be Open-Sourced.
social sciences can never use experience to verify their statemen
If only I had mod-points. Guys look into this. It's so important that people understand the power of ranked voting. IRV is flawed, but it's a start.
Take a look at Condorcet's Method for information about an even cooler (although more complicated) voting system.
A speech...
The more things change, the more they will stay the same.
1) Microsoft will integrate the voting softare into the OS.
2) In Chicago, computerized voting booths will be set up in cemeteries due to high voter registration in those areas.
3) Florida election are a mess due to old people forgetting to hit the submit button. Younger people that immediately follow a senior citizen notice that the touchscreen have either the Dem (left) or Rep (right) icons continuously flashing.
4) Losers of elections will demand a manual recount of the digital votes. State officials eventually declare the vote to be 0xdead to 0xbeef.
5) RIAA and MPAA will attempt to stop the digital transfer of votes for candidates who are former musicians or actors.
Oh yes, the "good old days" of paper ballots. When seals on ballot boxes would be "accidentally" broken, ballot boxes would "accidentally" fall off the back of a truck into a swamp, and the county courthouse would "accidentally" burn down after the ballots had been counted and reported.
Mea navis aericumbens anguillis abundat
Geez, someone moderated me as a troll. What ever happened to a sense of humor on slashdot? I started regularly reading slashdot for the satirical posts.
Slashdotters out there....lighten up a little. You got into technology for fun, not to fix all the world's problems.
If you want to do that, become a politician.....
****That is also a joke.****
-ted
I've worked the polls once before and i'm just commenting on my previous experience....it was a nightmare.
I'm sure my anecdotal experience isn't a statistically accurate sampling of poll workers, and i'm sure most are competent induhviduals.
-ted
Approval voting has some nice properties, but doesn't take into account something voters can easily express, the ranking of their preferences. Thus approval voting loses very useful information.
The best option for many kinds of elections is Condorcet voting. It's used by the uk.* Usenet hierarchy and, I've been told, by debian, but I've seen no confirmation of that.
I have my own rant about voting systems. (It's a bit rambling, but does anticipate and respond to some objections to Condorcet voting).
Prime numbers are exactly what Alan Greenspan says they are -S. Minsky
We were the first country to have 100% electronic voting (and the 1st to have any kind of it, I guess) . This year's election (federal and state representatives, state governor, senate and president) about 80% of the voting booths will be transmit the data and give the results a few hours later. ;). :)
Yes, we're poor but we know about digital democracy
BTW, we *don't* use M$ OSs on it. It uses VirtuOS sort of multitask DOS. Old but works
Adilson.
Faith can move mountains. I prefer dynamite.
Maybe i'm late in replying to this story, and hence this post won't be modded up.. (and i don't have that +1 bonus yet :-) but i just thought of putting here the fact that India has successfully used evm's in its elections. more details can be found at here And it does not contain Microsoft software :-)
Go read up opn Arrow's Paradox. Method doesn't matter; democracy is impossible. Funny how it works...
People like to pick a first choice as "their" candidate. Approval makes you say yes to some set, and no to the rest. I agree it has some mathematical elegances, but I prefer IRV and so do the majority of reform activists judging by the initiative measures which have actually made it onto the ballots.
Before you guys get too caught up promoting IRV, please read Chap 2 of Political Numeracy by Meichael Meyerson for some stiff discussion of the caveats of all voting "algorithms" including IRV (although he doesn't refer to it by that buzzword).
Sure, from a purely theoretical point of view. No real-world system is going to be perfectly accurate.
However, as a practical matter, the pareto imbalances upon which the proof of Arrow's theorem depends are very rare. When they do occur in practice, it usually means that the situation is similar to the 2000 U.S. presidential race, where the difference in the top candidates' votes is much smaller than the election's margin of error.
The fact is that IRV is the least expensive system that precludes the spoiler effect, and in doing so relieves the all-too-common prevalance of completely inaccurate outcomes, is the big win. I'll take a big jump in accuracy at a minimal cost any day of the week.
Condorcet is absurdly expensive for large elections. Can you imagine how long it would take to do a Condorcet count for the Governor of California, even from punch card ballots? (Hint: all of the ballot preferences have to be colected and centralized before the count can begin.) On the other hand, many districts in Canada do IRV counts of paper ballots, by hand. And they finish it by midnight, in most elections. Try doing that with Condorcet.
Approval is so different from the traditional plurality method that it is unlikely to be adopted for anything bigger than city council elections (for which it is already used in many districts, including my own.) Approval is already widely used for corporate board elections -- not exactly known as bastions of democracy, those.