Electronic voting should be a hot issue in Ireland. It’s the core of our democracy, and the government wants to convert it to electonic form. But the government doesn’t seem to have fully considered whether it’s necessary or desirable, or whether there might be a better way to improve things.
It’s easy to criticise the electronic voting proposals for Ireland. They aren’t voter-focused. The only people who benefit from introducing the proposed electronic voting system are the bureaucrats who are employed to count the votes. The system is more hassle to use than a simple paper ballot. The software has not been reviewed to any great degree, because source code is not and apparently will not be available. There will be no physical trace of the vote to allow a review or audit, and voters will be right to wonder whether their vote might just be ‘gobbled up’ by the system, or whether the details might make their way into some secret database. The fact that there will be just a few voting machines in each polling station will probably result in queues and time pressure at peak times on voting day.
On the other hand, the Irish voting system as it stands is unwieldy and tedious to administer. Although simple in principle, the practice of actually counting the votes and dealing with possible spoiled votes is pretty complicated. Counts lasting 10 or more days are not unknown. And for technical reasons, the voting is more random than what it should be under the manual counting system.
What the bureaucrats are gaining in efficiency, the voter is losing in convenience and transparency.
I think that there is an alternative ‘third way’ to deal with the voting problems that we have in Ireland that will deal with the inefficiency and the technical problems, but at the same time will maintain or improve the voters’ experience.
My proposal is that we deal with the problems by (a) computerising the counting to improve efficiency and eliminate technical problems and (b) improving the layout of the ballot to deal with the majority of problems that result in accidentally spoiled votes.
We can computerise the counting without computerising the actual voting. This can be done by allowing voters to vote on paper ballots as they have always done, then keying every vote into a computer in the counting centre. to ensure every vote was recorded correctly, each vote would be keyed twice by different election workers. This sounds time-consuming, but could actually be completed as quickly, if not quicker than a conventional vote. In a typical constituency with 35000 votes cast, and an average keying and verification time of 45 seconds/vote, the process would take 438 man-hours. With a team of 50 counters (which according to press reports is the usual size of a counting team for a constituency) the votes could be dealt with in less than 7 hours. Problems with ambiguous ballot papers would be dealt with during this time. This could be done using off-the-shelf software already owned by the public service and without any special equipment being required.
The system would be open and transparent. The original votes would be preserved in their original form for review as necessary. The entire database of votes (which would never include any identifying information) could be made availabe for public scrutiny.
Because the votes would then be in electronic form, this would eliminate the technical problems currently associated with Dail elections. At the moment, the outcome of vote transfers when a candidate is elected depends on which votes happen to be at the top of the elected candidate’s piles. An electronic system allows all the votes to be counted and fairly and proportionally transferred.
There is plenty scope for improving the layout of the ballot paper to reduce the number of spoiled votes. The blank areas on the ballot should be blacked out to reduce the risk of voters writing somewhere where they should not. Space should be allowed between the candidates’ names to reduce the likelihood of a pen-mark crossing between two candidates’ areas. To allow these improvements, space should be freed up by removing the candidates’ addresses. This information is available to voters elsewhere, and is hardly ever relevant to their voting decision.
Making these simple improvements in the voting system would vastly modernise it without comprimising it. It would make counting operations simpler and more efficient. It would not require us to entrust our electoral system to a private company in a foreign country. It would be simple and transparent.
Of course, these improvements should not be the end of the road. We should introduce other improvements, with the emphasis on making voting easier for voters.
For example, we should consider having an online electronic voting register. This would allow voters to vote at a polling station that was convenient to them, rather than being forced to vote at the station nearest their home.
Rather than using technology to simply reduce the hassle of operating our democracy, we should be using technology to enhance and enrich democracy by encouraging more people to participate. We shouldn’t be afraid to change, but we should make sure the changes are for voters’ benefit, not bureaucrats.
Agreed, this is a better system; OCR counting. It’s trivial for a “verification” group to verify the numbers using the physical ballots afterwards, too.
BTW 1 core complaint that e-voting purports to solve, is falling voter turnout figures; IMO adding electronic voting kiosks to the *existing* voting stations won’t make a difference there at all, as the voters still need to get to the polling booth on time on 1 day as before.
One idea I’ve heard, is using SMS to vote. It’s occurred to me that this could be made reasonably safe, by making it a subscription-required service; ie. you have to “subscribe” in advance, possibly by post or email, to register your phone number as connected with your name and address; then when you vote, it *calls you back* afterwards, with an automated menu system, and you must confirm (by e.g. pressing “1”) that you wished to make that vote. This protects against impersonation (if an attacker can fake the caller ID data through technical means), and also protects against attackers voting with stolen phones (if the phone was not registered beforehand).
However, stolen phones that were already registered can still be used; and it does not protect against voter intimidation (ie. your hired muscle menaces someone until they take out their phone and votes for you) — which is one of those things that was happening not too long ago in the North, after all.
In my opinion, the dangers of electronic voting
are ludicrously exaggerated.
Obviously everything possible should be done
to make e-voting secure and reliable,
and in particular I would strongly favour
an open source solution, by which I mean
that the source code of all programs used
should be made publicly available.
The steady fall in the percentage voting
in almost every democracy
is a far more serious matter
than the extremely improbable ocurrence
of computer fraud in e-voting.
Electronic voting — including voting by phone
or by television — offers the possibility
in the long term not just of reversing
the fall in voting numbers,
but also of extending voting
to a far wider range of issues.
This has never been possible
since the introduction of democracy
in ancient Greece, until now.
An SMS-based vote would not be a secret ballot, because you would be able to disclose your vote to another person before you sent it. This leaves the system wide open to vote-buying and extortion.
I wrote about this issue the other day on my own blog. My main thesis is that anyone can verify a piece of paper. By introducing computerised voting you move the process into the hands of a technological elite. I’m not sure this is something you want to do without instituting some form of checks and balances.
electronic voting IS a hot issue here in my local area. a local university (Johns Hopkins) released a study a short while ago on the system that the state (Maryland) committed to buying (at a cost of some $50+ million) and implementing state-wide by 2004. this has produced some back-tracking by the state, ordering a third party review of the system. even the neighbors are into the issue now, with Virginia making some ripples on the issue.
this should give you some interesting places to go with the subject. happy hunting…
There are about 60,000 windows viruses. And you are going to use a ms product?
There IS no security.
AND even if the source code of the voting software is revealed, the attacker can manipulate the results using ms backdoors. Why do you think the government of China, the USA, etc wants to see bits of the windows source code. To get re elected prehaps?
IN FACT,wouldnt a great us software company might help out its friends in ireland with votes in return for favours?
U have got to have have hardcopy of the votes.
Otherwise it cannot be reviewed if something goes wrong. Another point is a referendum. U have to have one on something like this. FF tried to change the election before to the good old english system. and the electorate rejected it.
AND changing to electronic voting IS changing the system. Lots of people diliberately spoil their votes. INDEED thats why they go to vote. Remember dustin in dublin in the Mary Robinson election?
He beet Austin Curry in some places. And believe it or not that is important!
Another flaw is the none of the above option. Where is it? In a 5 seater if all the candidates are crap or corrupt, you should be able to elect an empty seat. In a propper system, this gives a better measure of political eptitude than anything.
Anyway, where is the referendum?
Electronic voting is fine but …
the system proposed for use in Ireland is not fine at all …
the list of problems begins with the philosophy the politicians and beaufrocrats seem to be more interested in administrative convenience than in electoral transparency and accuracy.
see * http://www.evoting.cs.may.ie *
and * http://www.evoting.cs.may.ie/report *
and * http://www.spectrum.ieee.org/WEBONLY/publicfeature/oct02/evot.html *
for some specifics of electronic voting issues.
Meanwhile we should ‘if in doubt – do nowt!