Indian authorities have arrested a computer scientist for refusing to divulge the source of an electronic voting machine that he and a team of researchers used to expose holes in the country's election system. The Hyderabad home of Hari Prasad, managing director of Netindia LTD, was raided on Saturday morning at 5:30 by …
...from my experience with Indian customs bureaucracy and customs procedures, it's remarkable that the pressure from higher up had time to reach the police before the voting machines were replaced by newer models and the researchers retired.
Surely the simplest hardware attack is a large muscly geezer with a stick making sure you press the button he likes?
I've been to Injia.
Voting machines are always going to be risky
Please someone shoot me down if this is a ridiculous suggestion, but wouldn't it be a cleaner and less corruptible approach to use machine readable cards/paper in the same way that a money teller or a multiple choice exam machine works?
If the machines fail, then so what? It can be manually counted and verified. It is impossible to verify something that doesn't physically exist on a semi-anonymous system.
Can you prove that the system hasn't voted for the people that didn't bother to turn up?
You still have the same problem with if the software can be broken in the machine reading the cards it can just ignore the results and make its own votes.
It would only take a single recount to expose this though.
But with a paper trail
in this case the printed chits, you can always do a manual recount and compare its result with what the counting machine is programmed to spit out.
Whether or not a manual count can be seen as sufficiently precise and fraud-free, at least it's a process that people know and have kindof trusted.
Follow the money...
...from "the top" back to the vendors, of course.
What a suprise!!!
Stealing the vote
No surprise there. Machines that will cheat you out of a candy bar or soft drink are certainly not to be trusted with an election. These corrupt mechanisms even used to sell cigarettes to minors before an enraged organic population tore them into scrap.
Digital Graphite Technology
The only reliable, trustworthy mechanism involves pencil, paper and scrutineers.
Perfection in an imperfect world...
Unfortunately, pencils can stray to the wrong marks, voters can have brain farts or dyslexic moments and fill in the wrong box, and worst of all, who's making sure the counters are honest (and better yet, who's making sure the ones "counting the counters" are honest, too, ad nauseum).
"pencils can stray to the wrong marks" - so can a mouse pointer or finger
"voters can have brain farts or dyslexic moments and fill in the wrong box" - I fail to see electronic voting makes a difference here
"and worst of all, who's making sure the counters are honest" - the counters themselves, who are taken in from all the parties represented in the election. They'll have quite an interest in watching one another. And if anyone is in doubt, you can always have a new team come in and tally the vote once again. In eVoting, on the other hand, those very people are prevented from ensuring that things are what they seem to be.
Care can be taken
to minimise the risk of dishonest counters. Having a lot of them, randomly selected, and having several unconnected people count the same votes for example.
I'd go with a pen rather than a pencil though :-)
....provided there are more than two staff and the stations don't close before hundreds have had chance to vote. But that would never happen in Blighty or the US would it?
Don't quite know how you got a downvote for completely justifiable cynicism.. can anyone name an election without controversy surrounding counting mechanisms? and if you cite UK elections, we frequently have problems even though we stick to the paper and pencil routine.
Every single solution has the same problem: restrictions and mechanisms only control honest voters.
Re: Perfection in an imperfect world...
But you *can* go watch the votes being counted, you know - in fact, each party has people there overseeing the counts.
Once you've put that piece of paper in the box you can watch the box sit there, you can see it being taken to the count, and you can see the votes coming out of it. How can you do this electronically?
As for those who will say "what about the long queues in May," electronic voting is NOT the alternative to this.
Actually I think I can.
I have never heard of an election issue in Canada. They use paper ballots, they are counted quickly (results are usually in within 4 or 5 hours of the poles closing), and recounts are automatically done for close results. Sure it takes a lot of people to pull off, but it works. Voters can register on the spot (unlike say, the USA), so persons who choose to go vote, can do so.
There simply is no reason to use electronic voting machines like these, other than to attempt to save money or to in fact try to commit fraud.
The Fakir 2000?
The machines are flawed, posibly by design,
in every country where they have been implemented.
Why do you think they are designed behind closed doors, using closed source hardware and software? It is inexcusable for a voting machine to be closed source. With a paper ballot we know how they are counted and we should damn well know how the electronic ones work too. But we don't. We don't and it just so happens that every machine tested can be rigged half a dozen ways.
WHAT a surprise.
I have seen a few headlines on the net about this and finally got round to reading the story. So why was I surprised that he was arrested for having a stolen voting machine in his possession (Stolen becuase they are Govt property and they diddnt saction the release, or have very strong reason to suspect thats the case)?
Exposing vulnarabilities on websites is one thing, knicking kit to expose vulnerabillities is aosmething else (dont be surprised if you get nicked!)
"finally got round to reading the story."
Really? so why did you ignore it all and make up such a silly opinion?
Pencil and paper works just fine......
"who's making sure the counters are honest"
Having attended the last election count (i was a candidate) I can tell you that the candidates, and there agents, are watching the count and making sure that it is honest. Any vote can be disputed by the candidates or their agents, these are put into a pile to be adjudicated on later.
I doubt that any of us who actually work in IT would be prepared to entrust our votes to a machine.
Are there any 'religious focused' countries that aren't corrupt?
Sorry - Are there any countries that aren't corrupt? - there, fixed if for you
Our head of state is also the head of the established church. They say prayers in parliament and take oaths on bibles.
So, tell me, is there corruption in the UK?
Yes there is definitely corruption in the UK. My girlfriend's former lover was a barrister and she witnessed him bribing the judge in a criminal court with a brown envelope and her partner saying to her "watch this.."
I won't tell you the religion of the barrister, because the moderators of this forum will reject the post claiming it's racist. But the mods have a habit of rejecting posts of any mention of that religion.
Re: @Francis Fish
Oh please. We can all make the distinction between race and religion here, although you seem to have a little more trouble.
just because it's not your religion...
don't forget that the UK, CommonWealth and US are also "religion based" (or is that "biased")
when you start allowing your imaginary friends to set the rules there is going to be a problem
there's a lot to be said for a Libertarian Government funded by a Fair Tax ... but there's a whole lot of vested interests in stopping that ever happening so it'll just continue to spiral down
@Rob Farnell ... Hanging chad? Dimpled chad? ... punch-card systems are flawed too!
FFS: Just stick an X in the box and count it already, it's worked for years ... if it aint broke ;)
Besides if all the election results came in instantly it would ruin election night & all those natty graphics :D
I do worry for India though, they are trying hard to be viewed as a first-world country, but clapping people in irons for standing up for democracy is going to ruin decades of hard work.
Cue Streisand effect
Arresting a man who refuses to say how he managed to get a hold of a voting machine to test it, after the election authorities refused to provide one will almost certainly give rise to the suspicion that the election authorities and politicians are up to no good.
Paris Hilton, as she would have more of a clue than those trying to cover up the controversy.
Post your own message
kudos to these chaps and may the indian govt forgive them their " enthusiasm " !
when we booted round the footy oval in the 60's conceptualizing " keyboards at the end of domestic phone lines ", instant referrenda, electronic voting etc were on our minds. every development since convinces me that political power offers too many temptations to too many interested parties to allow that power to be assigned by any means not as close-to-foolproof as possible. i.e. pen/cil, paper, privacy, transparent scrutiny of results etc.
remember dubya in '04 !
big brother - because if we ever vote electronically, we will have to log in......
Stating the obvious, because it's needed.
Voting is supposed to be an open, transparent, and anonymous process, for obvious reasons. So, any voting commission that accepts voting machine manufacturer refusals to be open and transparent when asked is not doing their job. Further, any voting commission that refuses to be open and transparent or worse, as here, actively suppresses attempts at open voting process review, is entirely incompetent and needs to be gotten rid of and blackballed from government wholesale.
So, how open are our voting commissions really? The one in India is clearly unfit for service, for one.
Paper doesn't help
Paper trails don't really help.
If the machine prints out a receipt saying that you voted for Candidate A (which is in itself a bad thing; imagine this on a company notice board: "All employees who wish to take time out of work to go to vote must show a valid receipt showing that they for the factory owner's brother-in-law, under penalty of dismissal"), then nothing is stopping it from recording a vote for Candidate B.
In fact, it doesn't even matter what the machines record. A really corrupt election administration could just make up any numbers they wanted, and you'd have no way to verify it.
Suppose Candidate A receives 500 votes, B receives 390 and C receives 110. These are the *actual* votes, remember. The *announced* result, however, is: A 380, B 500, C 120. (Note that those figures are not so far out as to be utterly implausible. If they wanted to get a candidate elected in the face of very strong opposition, they might have to field a few extra candidates of their own just in order to split the vote.) You voted for A. You go with your receipt to the Town Hall to check how your vote was recorded, and are correctly told you voted for A. And that’s as far as you can take the matter.
Even if all 499 of the other people who voted for A go and check, they’ll be told — rightly — that their vote was for A. And because (1) they all go in one at a time to check their vote, and (2) there are also many B- and C-voters in there, not one single one of the A-voters will be the slightest bit the wiser that there are really 500 of them, as opposed to the 380 that was announced!
You could only ever determine that something was amiss if all those A-voters produced their receipts for Candidate A *at the same time*. And a lot of those receipts will have been lost or destroyed.
At least if the counting is done manually by the actual candidates themselves and their representatives, we can be sure that it's correct; because as long as none of them trust any of the others, the only result they can ever agree on is the truth.
@A J Stiles
Yes, you're right, just giving voters a receipt for them to keep it is completely useless (besides the fact that it would nullify the secret ballot). What would be useful is if all the voters get a receipt to confirm their vote, and then everyone drops their receipt into a normal sealed ballot box.
Normally the result would be correctly predicted to within a few %age points from polling. In the example you give there is a 23 point swing against A (from +11 to -12), this result would be clearly anomalous compared to the polling, and candidate A could request a recount based on the paper receipts in the ballot box. There could be a rule that obliges a manual recount if the gap between candidates is less than X% or the result not within X% of the polling ( margins of error could vary, of course). This would require a manual recount of only a small percentage of votes unless there has been widespread fraud.
Manual recounts should be public (news crews allowed to film etc) and overseen by scrutineers of the interested parties with a judicial body adjudging disputed votes depending on rules that are well-defined and clearly published BEFORE the election.
The flaw with all of this is the usual one where elections are involved, ie if the electoral commission is actually the one perpetrating or facilitating the fraud (see Iran last year). But in this case, the electoral commission would anyway not permit the introduction of a system that they cannot easily manipulate.
I may have missed the original point about a paper trail but...
..I don't think I did. I read the suggestion that cards be used and kept (still anonymous) and used in cases of dispute or for batch sampling (e.g. I dispute the result - well lets count the cards then). The only problem here is that everybpdy would end up disputing the result and we are back to manual counting of bits of paper/card.
Anyway, you very precise explanation fails by assming that the card is kept by the voter as a receipt - which I don't think was the suggestion (or maybe it was...dunno now).
In summary I agree that manual counting is still the best we have that satifies trustworthiness, openness and reliability.
I don't think you understand how the paper trail works. Well, the way it's actually worked with voting machines in the US is a junky printer prints on a paper strip inside the machine, and then it jams up so there's not actually a paper trail. Using receipts like you suggest, voters wouldn't take the receipts with them -- they would drop them into a ballot box. Then, when voters start saying they think the voting results are bent, and demand a recount, people recount using the printed receipts. At this point, I do wonder "why even bother with the voting machine?" But *shrug* people like shiny machines.
There's still nothing in principle to stop the machine recording a vote for one candidate in the "main" memory, and a vote for a different candidate on the internal paper roll. (Unless, perhaps, the paper is printed visibly, then fed on before the next voter arrives. But you, the voter, still don't know that there isn't *another*, hidden roll being printed differently.)
Anyway, it still doesn't solve the problem. If anything, it makes it worse, because you now have more than one set of data to count: the "main" record and the paper record(s). And you don't know which is right.
If we look at the UK system, every voter is given a unique token (poll card) which is unalterably marked and exchanged for one of a many identical tokens (ballot paper) under the scrutiny of the presiding officer. The non-unique token is then unalterably marked, in one of a finite number of ways, by the voter; so as to make it one of a finite set of sets of identical tokens. The important thing being, it is the *actual* tokens issued to voters which get counted; *not* some copy -- because the fundamental issue is, a copy can never be guaranteed to be exact.
ok, challenge time
For future Posterity, (in 50 years time someone will look at these archives and have a good laugh.)
How would you design an open-source, untamerable e-voteing system?
It wouldn't be universally comprehensible, it wouldn't be verifiable in practice (how do you know the machine is running code compiled from the published Source Code?) and it would still suffer from all the other problems associated with e-voting systems -- which stem ultimately from counting a *copy* of the vote, rather than the *actual* vote.
Serge Humpich ...
... will not be forgotten.
- Bugger the jetpack, where's my 21st-century Psion?
- Something for the Weekend, Sir? Why can’t I walk past Maplin without buying stuff I don’t need?
- Review 'Mommy got me an UltraVibe Pleasure 2000 for Xmas!' South Park: Stick of Truth
- The land of Milk and Sammy: Free music app touted by Samsung
- Privacy warriors lob sueball at Facebook buyout of WhatsApp