The electoral reform Green Paper covers a huge array of territory, from the enfranchisement of non-citizens to the placement of posters outside polling places on election day. It canvasses many important issues that relate to the shape our electoral system, from compulsory voting to enfranchisement of prisoners and more besides.
Even by the standards of green papers, this one seems to be very much a discussion starter rather than a strong indicator of what reforms the government is likely to implement; substantial reform to the architecture of the House of Representatives or Senate voting systems – for instance, the introduction of multi-member House electorates or single-member Senate electorates – seem highly unlikely.
One of the areas canvassed in the report is electronic and internet voting. Electronic voting is being used for an increasing proportion of votes in Australian Capital Territory elections, and has been tried for some disabled electors in Victoria. Internet voting has not been used in Australia, though there is some interest in it, particularly with the rise in pre-poll and internet voting. For what it’s worth, however, I strongly oppose the mainstream introduction of electronic voting in Australia, and even more strongly oppose internet-based voting.
Put simply, the trouble with pure electronic voting is that it is far too easy for a few technical experts – or even a single individual – to alter vote counts in an election-deciding but undetectable manner. Worse, proving beyond a reasonable doubt that such tampering did not occur is very difficult. Worse, proving beyond a reasonable doubt that the election results were accurately recorded to an audience who aren’t experts in software engineering and security – in other words, 99.9% of the electorate, and 99.9% of those active in electoral politics – is all but impossible.
Internet voting simply increases group of people able to do the tampering from people with physical access to anything that goes near the machines, to anybody with an internet connection. While internet voting is viewed as a substitute for postal votes, a process with its own integrity issues, the potential for undetected, systematic vote tampering, or even just unintentional, but systematic erroneous recording of votes, is such that in my view it should not be implemented now or at any time in the future.
Our current, paper-based voting systems have proven themselves highly reliable. The desire to fiddle with them is hard to understand, with the exception of better accommodating the special needs of the physically disabled. For the disabled, the best solution is simply to create technological aids that assist them in the filling of a regular ballot, which can then be deposited and counted with all the others, and hang the extra cost incurred.
The US experience 2 elections ago [and maybe others] with electronic voting should be a salutary warning against even considering such here.
Our system is pretty durn good already, “If it ain’t broken don’t mend it”.
Fine tune perhaps but that’s all.
http://en.wikipedia.org/wiki/Premier_Election_Solutions
Some details of the problems potential and real [?] involved in electronic voting.
I disagree that electronic voting would necessarily be easy to tamper with. We run ATMs without serious concern about the integrity of the system. Doubtless you’d want a system that allowed for the anonymity associated with voting, but I can easily imagine a system that could be
a) anonymous after the fact of voting
b) capable of independent audit and completely redundant verification
c) immune to effective tampering
d) user friendly to those intimidated by e-tech
Personally, as I’ve said elsewhere, were it up to me, I’d have a fundamentally different system for ensuring inclusive governance which would not only be robust against tampering and easy to understand but supportive of political pluralism.
Of course, it doesn’t occur to me that these objectives would be of interest to official reformers.
Fran, electronic voting is very, very easy to tamper with. ATMs are a completely different issue. I can use an ATM ten times in a day and no-one cares. I authenticate to my ATM with two factor security (something I have, and something I know). And it dispenses cash; if I don’t get my money, I complain, and a record that Alister accessed an ATM (complete with photo) can be found. Reconciling my vote not getting recorded correctly with its anonymity is problematic.
But the real reason why it’s not worth bothering with is that it would be a very, very expensive way of implementing a less secure, less reliable system. A voting machine would work if it produced a valid paper ballot, minimising informal votes. But the informal vote count is low anyway, and introducing optional preferential voting would probably go part of the way towards reducing the informal vote count.
Fran, the only way to achieve those things is to have the system produce a voter-verified audit trail – that is, a paper ballot that serves as the definitive count.
The elecronic voting device becomes, in essence, a fancy ballot printer.
I don’t know how much background you have in IT, but you may be interested in the writings of Rebecca Mercuri on the matter.
As the classic paper on how hard it is to verify that a system has not been compromised, you might have a look at this – again, it’s rather technical.
Fran, there’s an enormous literature about tampering with voting machines, which I’m sure Robert is aware of. I’ve dipped my toes in it over the years as well. Bruce Schneier’s work is a good place to start. (puts on best IT Professional voice) Trust me, voting machines are way too easy to manipulate, and the Internet is even worse.
I’d actually be prepared to bet that ATMs are frequently tampered with as well, it’s just that it is not in the banks’ interest to advertise this widely. Also, they probably spend more money and effort on security because money is involved (which, as we all know, is far more important than luxuries like electoral integrity and democracy).
You say that Alister, but of course you’re only thinking of existing systems in use. You can separate the input from the voter from the paper record easily enough and use that paper record as a copy of the recorded digital vote. At each election you could take a sample — say 10% randomly and non-random audit if results at odds with exit if other polls showed up. In either event, if you get an anomaly, you recount all the paper ballots. Auditing is independent.
As to the machines themselves, the image for the drive is kept in a secure location, encrypted and the drive at the polling booth is imaged on the spot on the day of the vote when the private key supplied to the electoral officer in secure form is entered. I see TABs and banks as good places for these.
Voter ID could be by resort to the public/private key system harnessed to a sealed letter which could be read by a a wand or MICR or entered manually.
Simple.
What is the compelling reason for changing to e-voting? I can’t think of any, and I haven’t heard of any good reasons either.
One point to note is that any software for e-voting must be published so we can see that it doesn’t have ‘back doors’: but this allows the ‘black hats’ the opportunity to find bugs that they can exploit.
Just because the yanks made a meal of it doesn’t mean the idea of electronic voting is a bad one. Remember they’ve a long history of making a meal out of manual voting too.
As for the “if it aint broke don’t fix it” line – well it is broken. It’s dead easy to vote multiple times, for example – something that a bit of electronics (not even full electronic voting) would make impossible. And with preferential voting its only a matter of time before we end up with a constitutional crisis when a knife-edge election is not resolved for many weeks – plenty of time for people to seek court injunctions etc to play games with the counting a la the US in 2000.
A bullet proof audit trail is certainly a sine qua non though – and that includes a paper printout for the individual voter. I’d never privatise the sytem as the septics have – keep it under the control of the Electoral Commission. And it’s gotta be open source – what you lose from hackers seeing the code you gain many times over from increased public confidence that there’s no behind-the-scenes dirty work (plus, of course, the usual open source advantage of any bugs or security holes being spotted quickly because many different eyes look at the code).
Internet voting I’m less sure about – I think it would disadvantage the old and other groups who are afraid of or don’t have easy acccess to computers. Plus the identity verification issues are a bit tougher. But maybe one day it might work.
It’s interesting that the IT professionals agree that e-voting is a shitty idea, while people who don’t have a background in IT are enthusiastic about it. (This is a possibly naive extrapolation from a known sample of two – me and Robert – but I’d bet it holds.)
Still, think about that, and about how we say “listen to the experts on climate change.”
It’s not just the IT professionals who think electronic balloting is insanely dangerous, DI(nr). People with electoral experience do too. Paper is simply the easiest and most unbreakable secret ballot there is.
Here’s a little Did You Know: in Upper House elections, while all balloting is paper, a great deal of the below-the-line vote-counting is done electronically already.
If you’re in one of the major Parties, and you don’t have a Monday-to-Friday job at the moment, and someone really hates you—and I mean, really hates you—they’ll give you a call just as you’re crawling out of bed during the week after polling day, while you’re dosing your hangover with black coffee and paracetamol. They’ll tell you to get yourself out to a warehouse somewhere probably in Western Sydney (or its equivalent in the less important States) where boxes of ballot papers are being shifted around in very carefully-marked piles, guarded by security guards and folks from the State or Australian Electoral Commission. Above the line votes, that way. Below the line, the other way. That’s where, oh lucky hate-recipient, you come in. You’re a scrutineer.
In the next warehouse, there are a hundred or more backpackers and labour-hire workers sitting at rows and rows of desks with computers, ready to methodically key in every voter’s poorly-written ballot. You didn’t really think that Senate and Legislative Council quota was worked out on a blackboard did you? Well you get to look over the shoulder of backpackers all day, forbidden to touch any ballot paper, able to question and challenge votes if you think they’re keying them in wrong. In effect, you’re making sure while preserving the secret ballot that the electronic vote is done cleanly and effectively.
It’s ugly and it’s bizarre, but it works, and it maximises scrutiny. And if it works out to be close within a certain margin… you recount.
Oh, baby, do you recount.
I’m half way between an IT professional and someone without a background in IT. I teach IT at high school level and am involved in concert with IT professionals in managing the school network.
Joe @ 8 – one advantage is it allows blind or near blind people to vote confidentiality. Another is that it would be possible to design a system such that it gives voters a token which they can use to verify their vote has actually been counted. You can even do it without them being able to prove to someone else which way they voted (which I think is important). Try doing that with a paper system.
If you have complex voting system like in the ACT and Tasmania then electronic counting means you get the result much much faster , especially when someone retires mid-term (ACT doesn’t have by-elections, but a recount instead which takes seconds rather than days or weeks). For small booths you can remove the confidentiality problems in manually counting (eg everyone votes for the same party so you know how everyone voted)
I definitely agree with the other comments about code for electronic voting systems needing to be published and preferably open source.
David Irving @ 10 – Well I’m an IT professional (well just call me a programmer) and one who has looked into the design of electronic voting systems a few years ago and I think it should be considered. Internet voting is a different matter with the main problem with internet voting not security is the loss of confidentiality in voting – makes it much easier to buy and coerce votes.
I think dd @ 9 makes a good point about considering the security of the system around voting/counting – we don’t even ask for ID for people voting so its a bit pointless getting overly paranoid about just one aspect of the system.
For those interested in the topic I think its worth looking at the ACT system of electronic voting and not just failures overseas.
Liam – in the ACT they do electronic counting (although people can still submit paper votes). For paper votes two people separately enter the details for each paper and the results compared before so that makes it pretty hard for one person to just cheat.
It’s interesting that the IT professionals agree that e-voting is a shitty idea, while people who don’t have a background in IT are enthusiastic about it. (This is a possibly naive extrapolation from a known sample of two – me and Robert – but I’d bet it holds.)
As someone who’s done both IT and elections (but not at the same time), I should say I’m not favorable to electronic ballots. But if one is going to do it, I opt for the approach the ACT used: make the code open source – up to and including the OS it runs on (Linux or BeOS). In theory, you could run voting machines on Windows, but the cost of licensing would be prohibitively expensive – especially for machines that you only use once every three years.
Most commercial e-voting companies, such as Diebold, produce closed source solutions. This is done to prevent theft and embarrassment. Theft, because their rivals could steal their code base, and embarrassment, because any detailed inspection would show too many security holes for safety.
As for the company known formerly as Diebold – their rep became so bad that they ended up selling their e-voting division for a measly 5 million to ES&S, another e-voting company.
Fran, you have a touching faith that science and technology aren’t tainted with the subjectivity of humans in their creation.
The American experience comprehensively proved that.
Note that using open source code (while a good start) is no guarantee of safety from malevolence. As the paper Robert referenced mentioned, having certainty that the known source code is actually running on a given machine turns out to be hugely difficult.
Another -1 here to electronic voting (from a computer science graduate).
I also wanted to add that from a sociology perspective I would think that elections, like censuses, are excellent examples of long-cycle social enterprises. They give a lot of people the chance at reasonably paid short-term work, and it gets them something respectable on their CV to help towards landing a better option.
plus it works.
Why not have a Peruvian knot system for ID!That way one could use one’s hair,entangled for a year before election!?Get it on video,the back of the head view,thus ,having the original always at home in a safe place,and the government puts the knots on its computers,graphically,pictorially and as pixelated image.Rather hair challenged persons could use a wig,have it stamped from inside by a Government Electoral officer!? Seeing I am a baby-boomer that has nostril hairs regularly,I might be inclined to use a shot of them as ID!? Without knots,of course!? Sorry!I cannot take the Electoral Commission seriously!
I tend to agree with DD – I’m not predisposed against further pilots of electronic voting with a view to making it a mainstream medium for voting at some point in the murky future. The current system, whereby polling booths seem to often simply have a big long printout of the electoral roll and prospective voters are checked off the list by a some kindly old dear with a lead pencil, is charming but outdated.
If in this modern age we can manage our banking online, shop online, do our taxes online, and send personal information through a completely unsecure channel like email online, there is certainly no reason why we couldn’t vote online.
I’m an IT professional and I wouldn’t support electronic voting. Unless we went for two factor and used a smart id card along with embedded chips. oh, and maybe tattoed barcodes.
Now there’s an OpEd waiting to be written.
I’m an IT professional and I wouldn’t support electronic voting. Unless we went for two factor and used a smart id card along with embedded chips. oh, and maybe tattoed barcodes.
Now there’s an OpEd waiting to be written.
Oops…forgot to say great post! Looking forward to your next one.
Guy @ 19 – I think with internet voting there is a significant problem of lack of enforced privacy and as a result you’d have a spouse voting the same way as their spouse or a young adult voting the same way as their parent as they don’t want to cause problems. Currently they can just say they’re voting one way and vote another. It also means people can pay others to vote for a candidate and verify that they do so. At the moment a briber can give out the money but there is no way of checking that the person actually does what they are paid to do (which is a good thing).
darin @ 20 – as opposed to the “trust who I say I am” which we use for the current paper based voting?
I’m not opposed to computerized checking of the electoral roll. That’s fine, in principle. However, it means that you need to do a complex IT rollout nationwide once every three years or so. That’s both costly and prone to failure.
I’m both an IT professional and a party hack (for any given definition of ‘hack’). I’m also a graduate student in economics. I’m strongly opposed to electronic voting.
IT – systems fail. They can be compromised. In addition, the combination of full, failsafe auditability and complete confidence in a secret ballot is a tough combination to match. Fran, I think all your proposal would do is give us fractionally faster counts for the HoR, and we get them all pretty much on the night now anyway. The Senate doesn’t need a quick response, as in every case except a DD the new Senate doesn’t sit for months – more than enough time to count the votes properly.
The economics of the matter is simple. If you’re going to spend millions and millions of dollars on a machine that effectively replicates a pencil (which is what an e-voting machine that returns a completed paper ballot would be), you’ve got a curious set of priorities. And you’ve got to maintain your expensive, electronic pencils. You’ve got to upgrade them, patch the inevitable holes, make sure they turn on when they’re supposed to, run test regimes to ensure that they return valid responses, and so on. All this for a system that would have only one practical purpose – to perhaps halve the number of informal votes. Because, while you could make it impossible to cast an informal vote, should you? Do voters have the right to not mark their ballot, to not cast a valid vote?
Me, I’m more tempted to write in and ask for multi-member HoR seats. Seriously, is there nothing else in the paper that we want to discuss other than electronic voting?
As far as the comments about keeping the source code of the software running on election machines public, it’s simply not adequate. As Fmark noted, the paper I linked to shows (one aspect of) why it is almost impossible to guarantee that the public source code is the one actually running on the machines.
As I understand it, the ACT system does not produce a voter-verified paper ballot. As such, it is inadequately secure, and I strongly oppose any attempt to extend its use until this is rectified.
Alister, there’s lots of things I’d like to discuss!
For instance, should the incarcerated be entitled to vote? In my view, yes – and it poses an interesting question – where should they be enrolled to vote?
Some interestingly detailed coverage of the issues and complexity involved in even a simple “choose one option” voting scheme, along with some cryptographic solutions in this youtube video.
darin @ 20 – I like the tattooed bar codes. On the back of the neck. Of course, it’s the end of the secret ballot.
Alister @ 23, you’re absolutely correct. I remember some years ago (around US election time, iirc) on Slashdot there were heated discussions about the merits of various e-voting
schemesscams, and the few Australians (including me) and Canadians participating asked why on earth you’d replace a perfectly good pencil-and-paper system with a hugely expensive, error-prone, high-tech solution to a non-problem.Robert @ 24 – what are you going to do with the voter verified paper ballots – is the intent to use them for random checks of booths or to turn the electronic voting system into a fancy printer so the votes have to be manually counted? I’d point out there could be other ways of checking the system is working such as electoral commission people adding votes on the day (anonymously) and following them through the system to make sure that they end up in the system correctly (and removing them at the end under scrutineer supervision so they don’t affect the result).
I think you also have to clearly distinguish between internet voting, electronic voting and electronic counting. The last of which we already have and somehow manage to tackle the question of whether the system is working and running the software it should be.
btw AFAIK no one of the political parties (major, minor or even independents) have complained about the accuracy or reliability of the ACT electronic voting system.
As for verification you can randomly test voting stations before and after and verify that the various parts (vote collection/accumulation/counting) work as expected using random data which can be verified separately. If you were really paranoid on the day you could feed the vote collection data into multiple separately developed and maintained backends as well.
How robust is our paper based system at the moment?
Robert @ 25, yes. Without question, incarcerated people should be able to vote. I might accept that people in for life shouldn’t, and can understand barring people who are in for 5-6 years (ie, the rest of one cycle and all of the following), but don’t we want people who have committed a crime to feel like they have a stake in society? Won’t this help them stay out once they get out?
As to where they should vote, my suggestion would be the address they were in immediately before being jailed/held in remand.
Alister @ 29 – for voting purposes you could treat prisoners like ex-pats who have similar “which electorate” issues.
Oh and a couple of other advantages of electronic voting:
- can eliminate the donkey vote problem which does skew results in close elections by displaying candidates in a random order for each person
- totally removes the chance of unintentional informal votes and optionally could remove informal votes completely.
- may get more below the line voting especially in situations where the number of candidates is very high (though personally I’d prefer they just add above the line preferential voting)
Electoral roll reform is sound, not much else is to be honest (apart from minor things).
In the 40 years I have been paying attention to the mechanics of voting, I have noticed that every noble, logically justified proposed electoral reform just happens to tilt the scales in favour of the party making the proposal.
Voting fraud, false counts and other chicanery occurs in paper-based ballots in Australia but at least an ordinary citizen has some chance, if they could be bothered, to detect it. A computer based vote will be easier to rort and detection of the rort will require specialised technical expertise.
I think that the voting status of non citizen permanent residents should be discussed. In NZ any permanent resident can vote. Not so in Australia. I think that there is a good case for allowing anyone who has been approved for permanent resident to be allowed to express their opinions through the ballot box.
It is great to see here that most posters here, IT experts and others, are totally against electronic voting. Australia used to lead the world in electoral reform, before Howard enacted his partisan “reforms”, such as removing the franchise from prisoners(overturned by the High Court), hiding the size and source of political donations, and spreading fear of virtually non-existent voter fraud to limit the franchise of the poor and homeless, and it is to be hoped that good sense will prevail again now that he has gone.
Paper ballots are the safest, cheapest, and most reliable means of voting, full stop. Look at what happened over the last decade in the USA if there is any lingering doubt about this. Australia, with a small population dispersed over a vast continent, has always done paper ballots very well, and is used as a model for the rest of the world amongst those electoral experts who are not impressed by yankee ingenuity.
On another matter raised in the green paper, that of extending the franchise to younger people, my reaction has changed over the years, and I am now in favour of it. These days, young people are in reality no more or less “educated” about parliamentary democracy that many adult voters, including those who once voted for Pauline Hanson…and the future belongs to the young, more vitally so now than ever before. What harm can it do?
I am not impressed by Fran Kelly this morning bringing Malcolm MacKerras onto the ABC to press his old man views on the subject. Malcolm is a psephologist of dubious repute, and is no more qualified to express a view on the youth franchise, than the next person. Especially since he is a monarchist to boot!
On one other issue raised in the green paper, I am strongly opposed to extending the franchise to non-Australians resident in this country, as I have always been opposed to the Howard-era introduction of dual citizenship (at the time, it seemed, only to enable Rupert Murdoch to retain his Australian citizenship when he beacme an American).
Chris, electronic vote counting systems aren’t a big deal, because if it comes down to it, the vote count can be repeated with entirely new software, or even manually. If the issue is with whether the vote machine recorded the preferences of voters correctly, no such backup exists.
As for your proposed auditing method, if the tracking is sufficiently precise to be useful, privacy is lost. Completely aside from the question as to whether it provides sufficient confidence that votes have been recorded accurately.
I stand by my view on the ACT’s system.
Grace,
I’m curious (not critical) to hear your reasons for the permanent resident franchise issue objection.
I am puzzled as to why this post and discussion is concentrated on electronic versus paper-based voting. As one or two have pointed out, there is much more in the Green Paper than that. Well, actually I suspect it is because of the left’s continuing obsession with the myth of the stolen Presidential election (it produced George Bush so it must have been rigged, no-one we know voted for him) but whatever the reason it is a second order issue. A how issue, the method of counting, and what and why issues are almost always more important.
The what and why issue that always baffles me about the Australian electoral system is preferential voting. I have only skimmed the Green Paper, but it appears to canvass only playing round the edges of the current basic system. Why? To my knowledge, except for watered down versions in Kiribati and Nauru, preferential voting of the Australian sort is used nowhere in the world (run off elections are not the same). Is everyone else wrong about the fairest way to run elections and only Australia right?
OK, I confess that my knowledge of electoral systems round the world is limited. No doubt someone will put me right about the extent of use of preferential voting elsewhere. But I would still hope for more questioning of it in the Green Paper process. Anything that produces Senator Fielding must have something wrong with it, and, less facetiously, party horse-trading over directing preferences comes perilously close to corruption.
Compulsory voting is also a nonsense in my view, but I have gone on enough already for one comment.
Wozza, the reason I picked out electronic voting is it’s one of these very bad ideas that, zombie-like, refuses to lay down and die. The fact that it is discussed in the white paper is an indication that it still has fans.
As somebody with actual professional expertise in the area, it seems to be a periodic necessity to regurgitate the arguments as to why it’s still a very bad idea so that it doesn’t get adopted.
As for the 2000 election, the disputed ballots were recorded on a badly-designed paper-based system. Paperless electronic voting was a knee-jerk reaction to the Florida problems.
While there are other important issues, I would think the principle that votes be recorded and counted accurately, and be seen to be recorded and counted accurately, is the very basis of a fair election system, no matter where on the party political spectrum you sit.
It is possible to both preserve privacy *and* make it possible for a voter to verify that their vote has been counted after an election. The biggest problem would be being able to convince people that when the system says they’re vote has been included that it really is as from a laypersons point of view the maths is not simple. And we’ve all seen what happens when the fundamentals of even something really important like global warming are not easily understood or have to so simplified for the general public to understand that its easy for fear mongers to cause a lot of problems.
And yes you can have dual systems for the voting station as well with the user input going to two independent systems at the same time if you really want to.
grace @ 35 – I think extending the franchise would be ok as long as it is also made compulsory – compulsory voting for everyone or no one otherwise you risk skewing the results. Like older people, if a 16 or 17 year old really didn’t feel like they could make an informed vote they could just vote informal.
You’ve hit the nail on the head, Chris.
Fancy verification schemes using all manner of complex mathematics may work in principle, but if only a tiny fraction of the population understands them they won’t have confidence in them if a dispute gets heated.
I should say, RM@41, that I’m not particularly in love with e-voting per se. I favour inclusive governance, which IMO would entail a system which facilitates timely feedback when policies are being devised and rolled out and a strong correspondence between the social composition of the governed and those governing.
If someone can show me a way of achieving these things without substantial recourse to formalised electronic input, then I’m for that, but it seems to me that it would be kind of loopy not to use such a powerful communication medium as is afforded by electrionic networks to shape governance.
I shopuld also add that I don’t set quite the same store by the concept of the privacy of the ballot box as some here apparently do. While I certainly think personal and identifying information should be kept at arms length from the bureaucracy and others — and difficult to get at in anything like real time and then only with a compelling reason, I don’t fear the government at a personal level. The real problem with contemporary government is not that they care what you or I think, but almost the opposite — they aren’t interested, not even a little bit and in fact, they would probably prefer that nobody else did either.
People worry about the potential for authoritarian rule if the privacy of the ballot box is infringed, but the reality is that the kinds of people whom putative authoritarians go after have views that are very public, or who belong to whole population groups.
Privacy is a valuable thing, and it should be given robust protection in areas of our lives that really are nobody’s business, but one shouldn’t get obsessive about it, IMO.
What might be of value is voting site “voting aid”. This would be a computer which offered verification of the consequences of ones choice, for those who are confused by the complexity of the voting form. Such a device could help in a number of ways. It could offer the option of starting from primary preference (green red blue black) then flavour that primary choice with special interests (pale green grey youth gun etc) to offer a voting form card to take to the booth. It could offer advice on the consequences of a particular combination of specific candidate choices.
Of course such software could be available on the internet ahead of an election for use at home (and may very well already be).
One thing about a physical voting booth is that it becomes the point of final decision.
“How robust is our paper based system at the moment?”
Exceedingly robust. Look at the any of the Courts of Disputed Returns decisions over close results and you’ll see the amazing level of detail that the Electoral Commissions can produce, all supported by auditable paper trails.
“As for verification you can randomly test voting stations before and after and verify that the various parts (vote collection/accumulation/counting) work as expected using random data which can be verified separately.”
More important, in my view, is that a successful attack on an electronic voting machine will compromise *all* the votes cast on that machine and possibly many more. At the moment, a successful attack on the voting system means you get to cast one extra vote. The odds of a single human security error resulting in an divisional election (or God forbid, a Senate election, being declared a failure goes through the roof.
d
No-one has been able to explain what is wrong with our current system of paper ballots and how long it takes to count ballots. There are lots of ways in which technology can be used to improve the governance process, including through the digital use of the electoral roll (such as marking people off on a digital device rather than a paper roll), but there is a massive danger and little benefit to using electronic ballots.
A lot of talk about electronic systems by little about postal systems, so here’s my penny’s worth.
The present system of postal voting is broken because it relies on it being a privilege and not a right. Anyone, and almost everyone, has, usually, to get the local MP to get the electoral commission to grant postal votes to a person. Just applying, like the law says, is not enough. This statement is being made from personal experience.
Then, when the paper arrives it is often provided at the last minute. If you’re disabled enough to be needing a postal vote you are usually not able enough to shoot out to the post-box, fast disappearing from the landscape, to post the papers in time.
That’s not counting the processing of the voting paper before you get to stuff the envelope.
We could take a leaf from the New Zealand electoral system where postal votes are common and have been for many years. If you want to force everyone to vote why not make it easier? Postal voting is a way of doing that and it is as secure as turning up at the polling booth to vote.
An effective postal voting system can obviate the need for electronic voting.
Why are people allowed to tout for votes right outside the polling booths? Does the need for that show us that the voting system is broken and that people have to carry printed cards into the booth to be able to make a “proper” decision? Should voters have to run the gauntlet of being pestered to take a card to get into the polling place? I don’t think so!
All touting for votes should have stopped by the end of the day before the voting commences – that includes all advertising everywhere.
Why does Australia continue with a voting system that ensures that, largely, one of only two groups get sufficient votes to govern? Why are Liberal and National parties afraid to separate themselves from the coalition? Is it because they all know that the voting system is broken and that both of them will suffer if they break their coalition? And what of each party’s principles?
And that is just one issue about the voting system. Just think it through for yourselves. Test why you think it is reasonable to have one MP holding the majority party to ransom?
Are any of you, commenting here, going to make any representations to the Green Paper’s authors? Or is this “consultation” process so broken and worthless, in most people’s estimation, that it is not worth the bother to do anything in the way of making a submission?
If people do make submissions about the paper, what chance do you think there is that your comment and examples will be read? Analysed? Taken into account? Made available to the law drafters?
Is there a feedback process by which we can tell that our feed-forward is going to make it into any sort of data-processing system that allows it to be considered in a valuable way? I don’t think so!
As a means of gauging public ideas and thoughts the whole “consultation” system in Australian politics is very broken. Many votes don’t count because the voting system is broken and the consultation system doesn’t work because there is no real system to consultation.
Lin, postal votes are not nearly as good as in-person voting.
The trouble is that with a postal vote it is possible to prove to another person how you vote. This is not possible with in-person voting.
Why is this a bad thing?
Say I’m a greedy solar panel manufacturer trying to increase my profits and thus offer people $1000 each to vote for the Greens. Lots of people take my money. But how can I possibly know that they have actually voted as they said they would?
If, by contrast, they agree to submit postal votes, I can look at their ballot papers and watch them seal the letter and deposit it in the post box. Therefore, I can give them the $1000 with high confidence it’s actually helping to rig the election.
Fran@42, that’s a much broader discussion, and there’s been a number of posts related to “Government 2.0″ on LP by Mark recently. But on your point about privacy, I’m going to have to strongly disagree. While governments may be mostly benign, technology has created all manner of tools to make abuses much easier and more efficient if the desire is ever there.
Here are BrassCheck TV videos on how the 2004 US elections were rigged by electronic voting: How we got George Bush, Your vote counts … but it may not be counted, Computerized election fraud, Dying for free elections, Destroying free and fair elections – state by state.
Any suggestion of four year terms, whether fixed or variable should be rejected out of hand.
If we can’t even stop the Queensland Government from carrying out plans to flog off public assets when that is opposed by at least 84% of the Queensland public and when they did not even inform the public of their intentions at the last poll, then how can our politicians be trusted for 3 years let alone for 4 years?
What is needed is greater accountability and tighter controls and we urgently need Citzens’ Initiated Referenda.
Robert @ 41 – yes, I think the problems are more social than technical.
Darryl @ 44 – I wonder what they’d do if someone set fire to a senate ballot box at a busy booth?
Being able to detect if something goes wrong is I think the primary goal. Australia’s paper based system (or an electronic based one) wouldn’t work if there was a concerted effort to disrupt it. Or do you think that if Afghanistan had used our paper based voting system they would have had free and fair elections? Perhaps the existing system has appeared to work so well because there have been no serious attacks on it.
I have not seen NZ collapse because they enfranchise legal residents, so how would that hurt Australia? I see it as a benefit, not only are we recognising productive members of our society we might even kill off a few racist bigots from apoplexy.
The voting setup here is insane compared to somewhere like NZ or Germany. It’s complex and widely misunderstood and partisans have strong incentives to spread the misunderstanding. A bit like the tax system. Major parties gain hugely from voters not understanding how vote transferring works (both in primary vote funding and suppressing minor party votes) and so have no incentive to fix it.
Simple improvements would be allowing exhaustion in all ballots, an intention of the voter test (so just ticking one box would count as a vote for that party and exhausting afterwards), and ideally more proportional voting. If you want wishful thinking I’d like to see three levels of elected reps not four or five, universal proportional voting (either MMP or pure proportional), no thresholds and a republic.
Fran: The problem with lack of anonymity in an electronic voting system can best be described as “once Google indexes the votes, what happens?”. In Australia it’s less a question of ASIO disappearing you for a few weeks and more one of how much it costs to direct mail only people who voted ALP, or explaining to a potential employer that voting Green doesn’t amount to funding terrorism. The voting records leak only has to happen once, and it’s not a great deal of data – a tiny USB stick or less than an hour online data transfer.
That is a good point, Robert, on the payola angle.
NZ uses postal votes for local elections but not for national elections up till 10 years ago. It may have changed since.
I quite like election day, it gives me the opportunity to see people who, in our busy lives, I rarely get to see.
Yes it has, but in practice, what we must do is demand accountability and opnennes about the ways in which data is used, who gets access and the adequacy of the rationales attached thereto.
There are all manner of putatively pernicious consequences to the availability of all manner of goods and services — and the advent of the web is case in point. One could, no doubt, seriously obstruct the exchange of child pornography by shutting it down, but at an unacceptably high cost.
A similar example might be the treatment of heroin. Right now there is a serious shortage of morphine in the developing world, — one that could be eased if attempts to destroy opium poppies to control the illegal trade in heroin, were abated. Local farming villages might become viable and the Taliban could be somewhat isolated from the trade they have now. There’s be a risk of course that there would be leakage from a legal trade into illegal markets, but should that risk be considered worse than the suffering ofg those who go without analgesics, ostensibly because first world regimes have a moral problem with people self-medicating and so create a context in which heroin can be a vehicle for funding criminals?
My point is that the mere potential for abuse is not a good enough reason to oppose something. You have to show that the inevitable consequences of abuse would outweight the inevitable benefits of adopting something.
People here are looking at theoretic vulnerabilities with electronic voting and dismissing the idea on those grounds, while ignoring the fact that the alternative – a paper-based system – has actually existing holes that are much bigger.
For example, take the claim that “we can’t know what code is actually running on the machine”. With paper, you can’t know that the voting slip is not counterfeit or otherwise compromised – but no-one worries about such theoretic possibilities.
Robert is correct in saying we can’t rigorously check the software after the fact (though there are quite simple heuristic techniques – eg Monte Carlo trials of simulated elections – that would take a very sophisticated Trojan to defeat). But the code is written, verifed and tested (including publication) and then loaded by the Electoral Commission under conditions of tight physical security; you can even get the EC to compile the Linux OS themselves if you’re really paranoid. The machine itself is then physically sealed. The vulnerability then becomes only the honesty of the EC, and if you can’t trust that then a paper system is very vulnerable too.
You’d be introducing a system that is possible but tough to beat to replace a system that is easy to beat. The main reason the current system is not often beaten (though how would we know?) is not any inherent security but the fact that almost all players – even the partisan ones – value democratic legitimacy; they are not thieves. Electronic voting would benefit from the same attitudes.
On the economics, maybe the opponents of electronic voting are on stronger ground – I haven’t done any sums. But the Yanks, the ACT and the AEC itself, who all presumably have done their sums, don’t think so. Don’t forget the AEC does many more elections than the Federal parliamentary ones (even union ones), so those machines would be trotted out a lot more often than 3 years apart.
I’m very disappointed by the lack of emphasis given to alternate voting systems — a single chapter without any CONCLUSIONS, just rehashing dot points that seem to be copied from Wikipedia articles.
The argument that it ‘prevents stable governments’ has always seemed oddly patronising, not to mention anti-democratic. You CAN have a majority government. You just have to convince more than half the population to vote for you — something, incidentally, that might not lead to a majority government under the PRESENT system. (Witness the 1954 federal election.) It’s not our fault if your parties suck. If you can’t convince a majority of the public to vote for your party, then you shouldn’t be able to govern as if you had. That the NSW government governs as if it held 100% of seats in the lower house when it in fact received 39% of the primary vote in 2007 is insulting. (Mind you, it’s only in the Top 20 most insulting things about the NSW government.)
Oh, and, um, the above is referring to proportional representation, just in case the awkward segue between the first and second paragraphs doesn’t give it away.
oops …
Careless ….
You have to show that the inevitable/probable consequences of adoption would outweigh the inevitable/probable benefits of adopting something.
Fran 52,
I have to diasagree. Where there is any risk of corruption, simple and safe is always better.
This guy…
http://news.bbc.co.uk/2/hi/americas/1825076.stm
is the front line of what Robert is referring to. But you don’t have to go to Brazil to find corruption, need I mention Australian Wheat Board. Even in our politics we have the ever present shadow of favoured deals.
As Chris pointed out there are possible attacks on the paper ballot system that would indeed invalidate (or potentially invalidiate) all votes within that ballot box.
But I think the point is that while all systems can and do fail, catastrophic failures of the current system are exceedingly rare and minor failures are either easy to remedy (with a recount) or of little consequence (the small number of extra votes cast.
Indeed the trickiest thing to resolve is often intention as in the infamous McEwen racing-car ballot. (Which incidently I’m still not sure what the ruling on that was, if anyone knows.)
And on that note I can’t believe no-one has mentioned the single biggest argument against electronic voting – no opoortunity for write-ons
Just to canvass anopther matter, and without a great deal of thought about it, I say that my initial impulse is to support optional enrolment for 17 year olds.
I support that Martin B. It’s good to be able to show the youn’uns how to vote while the little buggers are still relying on ya for a feed. Vote 1 Liberal and starve.
Let me given you a concrete for instance. This very day I was trying to purchase some replacement keyboards, but departmental requirements in place to prevent corruption meant I had to go to SmartBuy first, which lists all the approved suppliers. The only supplier with stock of USB keyboards was offering them at $62 each, plus delivery. I knew there was a previous supplier we’d dealt with before who could supply at $4.20 plus delivery. The accountant said I had to fill out a special form and attach the SmartBuy price so she could satisfy the auditor. The possibility of corruption was restrained, but did it achieve anything? Well it delayed getting what we needed and wasted about 30 minutes of my time.
Corruption is a problem but it’s not a catastrophe. As always, one has to be mindful of the scale involved. Remember those Latrobe Council workers of 20-years standing who got sacked for using some about-to-be-dumped bitumen to fill in a local community club carpark pothole in their lunch break and were called corrupt, not for soliciting but for accepting a steak sandwich? Madness.
To return to the topic, let’s have good protocols and openness and firewalls to protect sensitive data but let’s not get too precious.
DD@53: “On the economics, maybe the opponents of electronic voting are on stronger ground – I haven’t done any sums. But the Yanks, the ACT and the AEC itself, who all presumably have done their sums, don’t think so. Don’t forget the AEC does many more elections than the Federal parliamentary ones (even union ones), so those machines would be trotted out a lot more often than 3 years apart.”
As I understand it, the Electoral Act allows the AEC to conduct electronic voting for the bases on Antarctica, and perhaps a few other isolated constituencies like lighthouses, because this is indeed economic, and convenient, and the populations very small, so the likelihood of fraud or system collapse is equally very small.
The legislation also allows the AEC to conduct fee-for-service elections for private organisations, including unions and corporations, and the rules and methods are many and various, and not necessarily recommended by the AEC – they just do the job as asked.
The ACT electoral commission (not part of the AEC) does do electronic voting for assembly elections, but this is mostly in special controlled locations, and provides a real world experiment at a relatively small and safe scale, for observation by other electoral authorities. Nobody is jumping on board yet, and most voters in the ACT, including myself, decline the invitation to participate.
None of this constitutes an argument, economic or otherwise, in support of electronic voting for the entire national population, and to date, I have not seen the AEC publicly subscribe to such a proposition, despite considerable pressure to do so over the years. On the other hand, the AEC routinely lends its considerable competence and expertise, through the United Nations, in assisting emergent democracies develop and deploy paper ballot systems.
As for the yanks, I doubt you will find any electoral authority on the planet that believes that they know what they are doing when it comes to electoral procedures. Privatising the democratic process and bugger the consequences is what it looks like from a distance…
Point taken, Fran, some lessons could be learned there from British Rail. In the Barbalho case where vote buying was riff to preserve power the cost to the Brazillian public purse was 2 billion dollars ($US), and a number of lives. But the worst consequence was that the 2 billion dollars was supposed to go to encourage microbusinesses in the northern Brazil region. This affected hundreds of thousands of lives.
Corrupt people look for situations that offer leverage. A few hundred votes bought might provide the opportunity for one individual to massively milk the public purse from the top down. Some one at opporational level can only milk from the bottom up. Much harder.
Fran, sorry to keep harping on about this, but I don’t think too many non-IT people really understand the impossibility of guaranteeing that e-voting is trustworthy.
Just remember, most software except for trivial programs is buggy, and a lot of it is also poorly designed.
Hark back to Ken Thompson’s paper on trust that Robert linked to, and you’ll see what the problems are.
I agree with what appears to be the magority re electronic voting. Just too difficult to make it secure and appear to be secure. Besides, I see this as only a third order issue given how well the pencil and paper system works.
Looking other issues raised:
1. I have no problem with lowering the age to say 16. They are entitles to do a heap of ther things and are effectively treated as adults already so why not vote?
2. I also agree that any one here for an extended period is also elligible to vote. Period, not sure, probably of the order 1-2 years. The corollary here is that anyone who leaves the country for the same period automatically looses the vote.
3. Prisoners. They still have to live here so why not be aloowed to vote. This is just one more factor to ailienate them further from society. Surely the aim is to reintegrate them into society.
4. As for retiring members before an election. My problem here is they made a commitment, so why not stay. I would distinguish between someone who dies in office or becomes seriously incapacitated. In that case allow the party/grouping to nominate a replacement to serve out the term. Otherwise, go back to the ballot and whoever was second in the count gets the seat. No ifs or buts.
5. I also think there are other electoral issues that need addressing that aren’t in the paper. Things such as:
5.1 fixed terms. Stop the shenanigans of the leader having a free hand on when an election is called.
5.2 Double dissolution. Again stop the pussy footing of will they/won’t they. The moment the senate rejects the bill the second time satisfying the DD conditions, causes both house to be dissolved and an election in 4-6 weeks. The only issue is the government prepared to push, is the opposition prepared to stand ground. This also provides the circuit breaker for a fixed term if the two house disagree.
5.3 Allow citizens to initiate referenda and legislation. Simply require a minimum number of signatures to ensure reasonableness of proposals. The pollies may not like this but it ensures that groups with a wide following can get their issues into law if the people will support it but the pollies just won’t go there. Think abortion reform etc.
Enough for now
DKit
No known attack has been made on the current system because the cost and risk are too high for the payoff.
To falsify an election result in Australia requires corrupting multiple officials, at multiple levels, in multiple locations. It requires a lot of physical resources. It requires accessing and modifying records witnessed by mutually-hostile parties who would have an interest in exposing any such shenannigans.
By contrast, in an internet voting scheme, I need only corrupt the person who knows the root password on the tallying server. Or have access to a 0-day exploit that’s not widely known. Then I can make untraceable changes to the count.
Our democratic system has two requirements which become mutually exclusive in any electronic system: secret ballots and systemic integrity. The only way to have both is to produce physical tokens. At which point, you have obtained no advantage over a paper ballot, because hand-recounts will almost always be demanded by the losing party.
This is, incidentally, why internet commerce and ATMs are broken analogies. In such cases your identity is part of the transaction. In voting it cannot be part of the transaction by requirement.
As an IT professional, I completely reject that electronic or internet voting should be made mainstream.
Jacques @ 65 – I think you’re combining the problems of internet voting, electronic voting and electronic counting all into one. We already use electronic tallying systems – how many people have the root password to that machine?
In marginal electorates you may need very few votes to sway a result, and has anyone else noticed how the vote often changes in a recount implying that paper vote counting is often a bit faulty? I don’t know the details of the security of ballot boxes, but in practice how hard would it really be to reseal ballot boxes (if you had a couple of corrupt insiders to help)?
I’d also point out that electronic voting in the ACT over several elections has not resulted in any of the parties complaining. Its also not an all or nothing choice – there will be people for many years not comfortable with electronic voting so at the very least you have something to compare against.
One interesting tidbit from the AEC website seems to say that they are planning on or have already been trialling electronic voting for sight impaired people in federal elections.
http://www.aec.gov.au/About_AEC/Publications/handbooks/scrutineers/2during_the_polling.htm
“But I think the point is that while all systems can and do fail, catastrophic failures of the current system are exceedingly rare and minor failures are either easy to remedy (with a recount) or of little consequence (the small number of extra votes cast.”
As usual, you’ve found something resembling a point in my ramblings and expressed with more precision and clarity than I could.
d
Chris@66: When did you last look at a ballot box? They are are very big and obvious, and planted in full view of the entire polling booth. They are made out of cardboard, with plastic corner seals that are numbered and registered and cannot be re-sealed once cut. Ballot boxes are never out of sight of voters, polling officials and party scrutineers in the polling booth, from 8 am when they are sealed, until 6 pm when the seals are cut and the boxes opened for counting the ballots. Ballot box tampering just does not happen under these completely transparent circumstances. In fact, the seals are hardly necessary, with everyone watching all day, and are really only there to appease the nervous nellies.
Fran, sorry to keep harping on about this, but I don’t think too many non-IT people really understand the impossibility of guaranteeing that e-voting is trustworthy
I set out a way in which people could be confident of the integrity of the process. I could put in further redundant real-time checking to ensure an even higher standard if necessary.
In the end of course, it doesn’t alter the fact that in a place like Australia, you can have any government you like as long as its conservative. No matter who wins, it is going to be much of a muchness, which is why nobody is going to bear the considerable risks and costs of trying to corrupt the process. All their corrupting is best done by making up scare stories about boat people and terrorists and porkbarrelling marginal seats. That’s all legal and far more effective.
I was sure I’d put that in quotes!
“I don’t know the details of the security of ballot boxes, but in practice how hard would it really be to reseal ballot boxes (if you had a couple of corrupt insiders to help)?”
Really quite extraordinarily hard. The Presiding officer is issued with sets of numbered plastic tags that lock the plastic ties keeping the lids on the boxes so that they can’t be opened without breaking the seal or tearing the box. Before the booth opens, scruitineers inspect the boxes, watch the presiding officer seal each box in four places, and both the presiding officer and a scruitineer sign forms listing the individual numbered seals on each box. At the end of the day, the seals are checked again by the presiding officer and scruitineers, numbers checked against the form, then the boxes are opened for the preliminary count. Finally, the number of ballots at each booth is totalled and compared to the number of ballots issued at each polling place.
but before you’d bother with that, you’d have to get very good forgeries of ballot papers, which is even harder.
d
grace @ 68 – ah, I didn’t know if votes were counted in every booth separately or if the ballot boxes were transported to a central location especially from the smaller booths. There must be some serious privacy issues in the smaller booths as I can imagine in some places nearly everyone voting the same way. And I had thought that if you vote outside of your electorate area they were transported to your electorate rather than counted at that booth. I also thought there was some centralisation of senate vote counts.
The solution seems obvious. To ensure the integrity of IT voting systems, you provide a ballot paper as well which is filled in by hand and tabulated against the electronic results.
True it’s an additional expense but you could always save money by getting rid of the computers.
Seriously though, I like the fact that when I’m voting, I’m doing it the same way our ancestors did who fought and sometimes died for such a right. There should be some room for continuity and tradition when it comes to the physical act of being able to choose our leaders, an achievement up there with clean water supplies, the discovery of fermentation and the ability to control our bowels when it comes to human civilisation.
Besides, when it comes to that moment of truth in the booth, there’s nothing like a pencil and sheet of paper for expressing your true feelings about the candidates on offer.
Maybe IT voting systems should provide an “informal” option. 140 characters or less.
Chris@72: “There must be some serious privacy issues in the smaller booths as I can imagine in some places nearly everyone voting the same way. And I had thought that if you vote outside of your electorate area they were transported to your electorate rather than counted at that booth. I also thought there was some centralisation of senate vote counts.”
As I understand it, there is a lower limit on the size of polling booths for economic, operational and privacy reasons, but it would rarely happen that even a small booth would vote entirely one way. Where this does happen, at many aboriginal polling booths for example, the voters do not seem to be too concerned that their collective will is transparent.
If you vote outside your electorate, you have to cast a declaration vote, and the envelope containing your ballot paper is transported the next day along with all others, to their home destinations, during the state-wide “declaration vote exchange”, which is done under highly secure protocols.
And yes, senate counting is centralised, but the first count of above the line votes is done on the night in the polling booth….enough from me.
I’m biting the bullet and am currently working on a submission on the topic of electronic voting.
When I have a draft, I’ll post a link to it for comment.
Both. Depending on the election, votes get counted once at the booths on the night (which is how we get to know who will form government), then transported to the headquarters and counted again twice. More if there’s disputes.
Another thing that’s nice about paper systems is the role of scrutineers. Here you have mututally suspicious parties with a vested interest in sniffing out the merest hint of corruption by the other side. Pen and paper systems, sign-offs, tagged ballot boxes etc are easily verified by any literate individual. Contrariwise, verifying the safety and integrity will require very high levels of expertise. And all those who would qualify as such experts are against the idea.
I lump them together because it’s a case of dumb and dumber. Electronic voting is a bad idea; internet voting just magnifies the problem.
The count on the night is indicative only. The next two counts, held at AEC state headquarters, are the actual counts. When a seat turns out to be marginal, each of the parties sends in its most experienced scrutineers, and the AEC its most experienced officers. They argue much more aggressively about what is formal, what isn’t and so on. Consequently this will cause the tally to change.
Whether someone has trialled it is quite beside the point. It’s still a dumb idea.
They did so in 2007. I support helping the disabled to vote, but that’s a compromise. For the mainstream, hell no.
I posted a copy of my submission over at Troppo.
Well you’ve convinced me Robert (and others). Not that I necessarily thought electronic voting was a great idea anyway – I’m quite a fan of veriifiable marks on paper that a whole bunch of different people can all easily see and interpret at the same time. But I hadn’t really been overly exercised by the issue either way previously.
I shall henceforth raise myself to the state of vociferous bah-humbugging of the concept when and if required in the future.
Thats a bug, not a feature. A rather serious bug when it comes to very close elections.
It’s an inevitability: people will argue about the vote whether it’s on paper or on disk. At least with paper you don’t need expertise to argue, and the ability to settle it in court is well understood and accepted by all parties. Nobody can construct a plausible attack on the present system that is within two bull’s roars and a cooee of the cheapness and low risk at which any proposed electronic or internet system can be subverted.
Electronic voting is a solution in search of a problem. Beyond extending franchise to people with disabilities, I struggle to find any convincing reason for its use in Australia. The only people who would benefit would be the makers of the machines. There is no issue with the speed of the count or electoral fraud in Australia so why flush hundreds of millions of dollars and public confidence in the system down the toilet to save a couple of hours on election night?
Nabakov@74
Bah, too much flashy modern claptrap for my liking. Bring back the pebbles!
Also, can anyone name a major IT project that came in under budget and on time while delivering total data security? I spent an hour on the phone today with NAB Visa over a fucked up online purchase. Care to put your exercise of the voting franchise through such hoops?
“Bring back the pebbles!”
At least you can fling ‘em at the scrutinisers.
While it is theoretically possible to tamper with electronic systems it is remarkably difficult to perpetrate a crime and not be detected. I once worked on a gambling IT project and we programmers were required to try to think of ways we could cheat the system and get away with it. The problem was always “getting away with it”. This is the same with electronic voting. It may appear to be easy but making a difference is actually hard. In the end we decided that the best way was to beat the system was to follow the bets of a person with inside knowledge – so we made sure that telephone operators didn’t actually know who they were talking to.
I see no reason why we shouldn’t be given the choice. If I trust an electronic system why shouldn’t I be allowed to use electronic voting. If I don’t then let me use a paper form.
We would find that if we allowed electronic voting then we would get probably about 40% of the people using it in the first year and each election it would get more as people became more familiar with it and found that they could trust it.
The greatest failing with the current system are things like people voting for the wrong person because they have not changed their address or people not voting because they did not fill in all the paper work on time or properly or people misunderstanding and voting for a person they did not intend to vote for.
Kevin, election systems have fairly unique requirements that limit the amount of auditing you can do. Furthermore, you have to assume very well-resourced and motivated attackers. It’s practically tradition for the CIA to interfere in elections, and they also have a long tradition of electronic sabotage.
Finally, you have to take into account the public confidence issues. You not only have to convince yourself or your dev team it’s fair, you have to convince my mother, who still thinks computers are some kind of scary voodoo.
Jacques Chester@77: “Both. Depending on the election, votes get counted once at the booths on the night (which is how we get to know who will form government), then transported to the headquarters and counted again twice.”
If I can just add to this: The indicative throw of preferences on the night, immediately after the booths close at 6 pm, is indeed the means by which we know almost within the hour which side is likely to form government.
(The ballots are then counted again later over the next few days, to confirm the indicative count, and adding in the declaration votes, which may or may not change the result, and perhaps again, if a recount is required. Yes, there are occasionally small changes to the totals on successive recounts, sometimes by the Court of Disputed Returns, but this is not usually not fatal.)
Electronic voting will not substantially improve on this, so what is the point? I look forward every election to that hour or two when the booth results are coming into the AEC central tally and the whole nation is sitting on the edge of its seat waiting for the call. Its better than the Melbourne Cup.
And on a related issue, there is no need these days for the National Tally Room, with centralised, computerised tallying of the results by the AEC in its own facilities. Direct tally feeds to interested parties, like the media and the political HQs, happen almost instantaneously from the AEC. The National Tally Room is, these days, an expensive farce.
Greenhorns in the media will tell you that the “atmosphere” at the Tally Room is unparalleled. Rubbish. That was in the “old days”. These days, only a few hundred members of the public can get inside, to stand behind ropes on naked concrete in a drafty hall, doing what? Actually, craning their necks to look at a the few TV screens mounted up high to watch Kerry O’Brien et al call the result.
There are hundreds of reporters sitting in front of their computer terminals, with the backdrop an old fashioned tally board where serfs behind put up cardboard numbers that are way behind the computer feeds, and a large enclosed temporary construction to house the media stars like Kerry O’Brien “broadcasting from the National Tally Room”. They could just as easily be sitting in their studios at work.
One more observation on the first count on the night. The two major political parties are getting close to killing the goose that laid the golden egg – the indicative throw of preferences that gives us the early result on the night.
By insisting on intervening in the distribution of postal vote applications in the weeks leading up to polling day, which should be done by the AEC but which has been infiltrated by the parties through legislative change, and blanketing marginal electorates with official postal vote applications (funnelled back through party electorate offices before being onforwarded to the AEC), there is a gradual shift happening towards postal voting over voting on the day.
This will eventually mean more and more declaration votes, more time to count, less reliable indicative throws at the close of polling, and less confidence in the call on the night.
Why do the parties want to take over the distribution of postal votes applications from the AEC? In order to find out how you are likely to vote (whether you send your application back to the ALP or the Lib electorate office), giving them the edge in targetting you for persuasion through direct mail. Another brilliant american invention….
I think its one of those things that needs to be kept around so things can be seen to be done even if its not really relevant anymore – a bit of public confidence PR.
Sorry for the double negative in my previous post, that should read “…this is usually not fatal”.
Chris@87: you are in agreement with the media parasites, who have their theatrical backdrop built for them at huge public expense, for one night, including the minions behind to put up the cardboard numbers just like 60-70 years ago, when nobody can read them at camera distance and they are way behind the computer feed anyway, and including several hundred miles of cabling and terminals so a bunch of overpaid media boys and girls can be flown in and all sit together reading the results as they come in on their terminals, not from the tally board, when this could be better done in their offices or at home…..and with the old boy pollies, who still can’t use email (why do we have electorate officers and massive postage allowances, whoops that’s just gone, fuck) and want to pretend that computerised information flows don’t happen and that telephoning booth results in from Kalgoorlie is still more reliable than…..
That’s OK in the end, its just public money for a meaningless circus, but what about the Big Lie that is being fed to voterland, who still squint at the screen trying to read those numbers on the Tally Board. Who is going to tell them that those cardboard numbers are way behind the running electronic tally along the bottom of their TV screens or the bobble heads telling us about the latest swing in woop woop…
There are a number of serious concern related to electronic voting. Not the least is the security and access to the recorded preference data files, voting statistics and election results.
The electoral commissions have spent millions of dollars developing computerised counting systems, with each State and Federal Election commission duplicating resources and expenses. Public money wasted with no oversight as to the expenditure and security of the data. Elections are no longer open and transparent.
Whilst there are some advantages in electronic voting, the use of a transcribed data-entry process for single member electorates can not be justified. Without access to the recorded data-files and ballot statitistics you can not properly review or scrutinise an electronic count. The use of electronic counting has come at the expense of open and transparency.
Access and publication of preference data and polling place statistics MUST be readily available to scrutineers and subject to public scrutiny and review.
It took the AEC over three months to provide copies of the preference data files pertaining to the 2007 senate election.
Analysis of the data provided had highlighted problems with the way in Senate votes are counted. The Greens should have own the sixth senate spot in Queensland.
The AEC with millions of dollars and numerous resources at their disposal was unable to verify the facts. Preferring instead to turn a blind eye. ask no question and you need not giver answers.
All that was required to demonstrate the flaw in the system was to recount the ballot excluding all candidates except the last final seven candidates (3 ALP, 3 Liberal/NO and 1 Green). No other change in the system is required.the greens Larisa Waters should have been elected. The system and method used in distributing the vote denied her the right of representation. Other flaws in that way we count the vote had also been highlights. Analysis of the Victorian Senate results showed that ALP David Feeney could have lost out to a distortion in the calculation of the Surplus transfer Value which would have given the Greens a 7,000 vote bonus allowing them to unfairly win the final Victorian Senate seat.
Victorian Electoral Commission
The Victorian Electoral Commission refused to provide copies of the preference data-files and statistical voting information for the 2006 State elections. The information was only partially provided following an FOI application and again a three months delay.
The Victorian Electoral Commission failed to provide information on number of postal and pre-poll votes issued prior to the 2006 State Election.
The VEC failed to provide copies of the preference data-files pertaining to the Western metropolitan primary count (Count A).
Votes went missing in Western metropolitan region between count A and Count B Where the total number of votes recorded on each count changed along with an unexplained change in the result.
The Victorian Chief Electoral Commissioner gave evidence to the Victorian Parliament that the data files had been destroyed and as such could no longer be subjected to Parliamentary review. They were also not provided to scrutineers. Access to this vital information would have allowed scrutineers to analysis the data and the quality of the data-entry process.
The Victorian Electoral Commissioner confirmed in hi evidence that the VEC had accessed recorded data-files pertaining to the electronic voting data without scrutineers being present.
The fact that information was not secure and data-files had been deleted raises a number of serious concerns about electronic counting and security of the electronic ballot. It is extraordinary that the VEC had deleted copies of the data files pertaining to the primary cont of Western Metropolitican and that they failed to make backup copies of this data. It costs 100,000’s of thousands of dollars to data-enter the preference vote data and the VEC does not make backups? Either way you look at it the VEC appears to have been engaged in a cover-up exercise to avoid public review. If they deleted the data as claimed the system is clearly at fault. If they failed to make backups they are possible negligent. if backup files do exist then they have deliberately mislead the parliament. either way the Chief Commissioner and his deputies are at fault and should resign.
Access to the primary data-file would have allowed proper analysis and review of the data-entry quality.
A request had been made prior to the close of the polls for this information yet the VEC failed to secure the data files and or make them available to the Victorian Parliament.
The total number of votes recorded between Count A and Count B had changed as did the result of the election. Had the VEC provided polling place statistical details as to the number of ballot papers issued as requested it would have been possible to ascertain more accurately where these votes went. None of the data provided by the Victorian electoral commission tallied. The polling place results did not match the number of votes counted. The number of votes recorded for the lower house did not match the number recorded for the upper-house. The number of total votes between count A and Count B did not match. There was up to 500 vote discrepancy.
In the 2008 Victorian Local Government elections the VEC refused to undertake a preliminary sorting of the ballot papers into primary votes prior to the data-entry tabulation process making it virtually impossible to scrutinise the data entry and counting of the ballot paper.
Whilst the VEC provided access to the preference data files for the City of Melbourne they refused to do likewise for other municipalities. Many who had cause to complain about the lack of transparency in the computerised counting process.
The use of a data-entry computerised count for single member electorates provided no benefit and came at the expense of maintaining an open and transparent counting of the ballot.
We are spending Millions of dollars duplicating resources and system development. Money wasted.
The Victorian Electoral Commission went as far as harassing individuals who had valid cause to raise concerns and complain about the conduct of the elections intimidating witnesses who gave evidence to the parliamentary inquiry.
As a result complaints have been made to the Victorian State Ombudsman who was prevented into looking into the veracity and details of the complaints. The Ombudsman has no jurisdiction over the VEC and as such can not undertake any review of complaints. Why?
re Tally board. Yes the board is a farce but the central tally room is more then just a media stunt. It allows access to a host of information and various commentators all from one location. Sure if you want access to information the Internet is a better resource.
My main criticism in the 2007 Election was that the AEC failed to provide election night feeds showing polling booth returns for the Senate. The only reason Senate data was not provided was because the ABC and Antony Green was not interested in Senate count data. There fore it was not available as part of the data feed. The AEC could have readily provided senate data as a separate feed. Just because the ABC is not interested in the Senate Count does not mean that this information should not be published. The AEC is not a media organisation. Election ballot statistics plays an important roll in the scutiny of the ballot and the ability to effectively monitor the progress of the count.
I would also have liked a breakdown of the registered postal, pre-polling and absentee voting statistics so that the number of outstanding votes could have been ascertained and monitored more closely. IE no surprises and missing votes as occurred in the 2006 Victorian State election
This coupled with the publication of the senate preference data file as and when information is available would greatly assist in the scrutiny of the electronic count.
having just had a look at the paper i’d like to know what people on this thread think of the ideas in chapter 7 – that provide for changes to the enrolment process to allow electronic enrolment and also the possibility of ‘automatic enrolment’ something that would address the 500 000 or so under 25s who aren’t currently on the rolls
alexinbangkok – I think some forms of automatic enrolling and update could work well. Eg getting people to tick a box if they would (subject to age) be eligible to vote when applying for a medicare card or drivers licence and then automatically adding them when they are old enough would reduce the number of young people not being enrolled. Similarly for updating it would be convenient if you could just tick a box to have your enrollment information updated when updating your medicare card address or drivers licence address.