Internet voting in federal, state and local elections?

Recently, I received an e-mail alerting me this year’s elections to the American Mathematical Society Committee and reminding me that I can vote by postal ballot or over the internet. I vote in these elections each year and voted over the internet in their elections in 2006 for the first time. 

I suspect that many organisations allow their members to vote over the internet in their internal elections and I wonder whether voting over the internet will become an option in Australian federal, state and local elections and referenda in the near future?

 Let’s assume that security issues associated with voting over the internet can be satisfactorily dealt with. (Yes, we have all heard about problems associated with electronic voting in the United States.) Would political parties then legislate to change the voting system so much as to allow people to vote away from a polling booth on a single Saturday? 

Many people already vote away from a polling booth on a single day via postal or absentee voting. However, most votes are still cast on polling day at the polling booths in each electorate. The dynamics of election advertising and vote counting would change if many people voted over the internet – there would probably be a marked drop in the numbers of people voting at polling booths, especially in the inner parts of capital cities, and there would no doubt be a window of a few weeks during which people could vote over the internet. This would make the impact of election advertising less predictable, especially if people were not able to change their vote once they had voted. 

People should have the greatest freedom in casting their votes and should be allowed to change their internet votes up until 6pm on election nights. This would make election campaigns more dynamics and interesting. Internet voting would make elections less prone to last minute surges, as many people would have voted in the weeks leading up to the election and not all of them would resubmit their vote on election day. 

I wonder how internet voting would affect the cost of holding an election?

12 Responses to Internet voting in federal, state and local elections?

  1. Caz says:

    Whoa, way off track!

    Absentee and postal voting are a necessity, and the Electoral Office accommodates such. (Notably, absentee and postal votes are generally the last to be counted, after all the other votes.)

    You’re not only suggesting electronic voting, you’re suggesting that people should be able to vote “whenever”, and that they could keep changing their vote. All of which is not only absurd logistically, but would be a security nightmare. Sorry, but you would not be able to prevent “sneak previews” of voting patterns, and subsequent “leaks” to the media, which would – surprise, surprise – lead to “voting surges” and gross distortions to what the collective may have intended.

    I don’t know what “voting surges” you think happen now in Australia, but basically they don’t; that’s a very American thing, and very much caused by the nature of their voluntary voting system. The only “surges” we have are people turning up to the polling booth at 5.55pm – I know, I’ve done it myself, but the timing of my vote didn’t influence my voting decision.

    Even in America, where electronic voting has been tried, people vote on the same day as everyone else, and they still have to go to a polling booth to caste their vote.

    If we ever have electronic voting here, and I believe we should have, some day, I would fully expect that there would still be a single voting day and people would still attend a booth to cast their vote electronically.

    I’m bemused why you would want people to be able to cast a vote and then keep changing their minds – are you painfully susceptible to political advertising? Political advertising, as with ALL advertising does not influence voting intentions. Advertising only serves to reinforce existing views (positive and negative); this is well established by research.

    In terms of “voting from home” – I believe that could be done, securely, but I don’t believe it will ever be permitted. A lot of people submitted their census forms via the Internet last year, and there were not IT security problems, but they still had to meet and greet someone at the door, and be given an access number.

    Voting is trickier, given that one of the requirements prior to voting is being sighted (even if “you” are pretending to be your friend) and having your details crossed off the electoral role, and asserting that you have not already voted at any other booth – all of which can only occur if you are registered in the first place. All of your suggestions would throw these crucial administrative requirements into chaos, not because of IT security, but because the integrity of the entire electorate and the entire process would be thrown into doubt.

    Voting electronically and voting over the Internet are entirely separate concepts; the former is likely to be used some day, but not the latter.

    The idea of people voting randomly and amending their votes all over the place in the lead up to a close date and time won’t ever happen. It wouldn’t make elections “dynamic” or “interesting” – you make it sound like a two-up game on ANZAC day. It isn’t.

  2. Sacha says:

    Caz, one can submit one’s tax return via the internet (and I did so this year). I imagine that the tax dept is interested in the security of this system – and the essential element of this is that one has a previous tax return.

    Currently, there is practically no ID needed when one votes – turning up “in person” hardly improves on this. All that is required is that one knows a few personal details of some voter and turn up to a voting station where that person hasn’t voted, and one can vote in their name. It wouldn’t be difficult to create systems in which a greater amount of ID security is required in order to vote than at present.

    I disagree with the contention that political advertising doesn’t influence voting intention. In a compulsory voting system, there would be no need to advertise if this was true. (The situation in a voluntary voting system would be different.) I imagine that the result of the 1993 federal election would have been different is there had have been no political advertising (remember the ALP ads about the “15% GST”?)

    Electronic voting is just a slightly different form of casting a ballot and allowing them to be counted. I havn’t heard any compelling reason why one should not be able to change one’s vote until 6pm election night. Why not have that chance?

  3. Caz says:

    Working backwards:

    – Once you’ve submitted your tax return or your household census data electronically (or even in hardcopy, for that matter), you’re done, you can’t ask for it back so you can re-think it. Can’t see why voting would be any different if conducted electronically, that is, make your decision, cast your vote, and you’re “locked in”. Voting has always been a “lock it in” system! Wouldn’t matter if you’re voting on a horse or a pollie, once it’s done, it’s done.

    – There wouldn’t be any logic or value-add to allowing electronic votes to be changed. It would be just as pointless as it would be if voters could wander around with their hardcopy slips and re-do them, or worse, submit them, then ask for it back (so much for anonymous voting!) so they could rub it out and start again.

    – Speaking of anonymous voting, that would be one aspect of electronic voting that would need to be carefully assessed, however, it has been done overseas, so it mustn’t be too much of a hurdle. Not sure how it is handled in the US, but I assume it’s no different to hardcopy voting, that is, electronic voting still requires attendance at a polling booth, and presumably the person only gets one shot submitting their electronic form, anonymously, just as you do with hardcopy sheets.

    – Studies truly have shown that advertising, most particularly political advertising, reinforces existing views – that in itself makes it inherently worthwhile for political parties. That is, at the very least they want to make sure they hold onto their existing support base, and in order to do that they advertise and advertise and advertise. They want to make sure the voter sticks to their decision right up to the second they are standing in the polling booth with a pencil in their hand.

    It would be wrong to negate or under-estimate the power of reinforcement. Eg, Telstra is a crap telco and a crap ISP and a crap mobile phone provider, YET they GAINED new customers in all three retail areas last financial year. They have the highest charges and the worst deals, yet they GAIN customers! How? Reinforcement of the Telstra brand, over and over. Telstra customers complain endlessly about the service, but they stick with the brand.

    – With political advertising it’s only slightly different. The reinforcement is of existing views, so an ALP supporter will have their positive view of their voting intention reinforced regardless of whether they see an ad for the ALP or an ad by the Liberals. A Liberal ad (regardless of content) will “confirm” their worst thoughts about the Libs. An ALP ad (regardless of content) will “confirm” their best thoughts about the ALP. And vice versa for Liberal supports, and minor parties. Apparently all people pretty much respond in that same manner.

    The effect of advertising in the US is exactly the same, but the dynamics of American politics and the nature of the advertising is entirely different to ours (stylistically, it’s far more in your face and brutal stuff), but that’s only partly based on the need to motivate people enough that they will go out and vote. Most of the “motivational” activity is, of necessity, done at a grass roots level, that is, getting people to register (not the same as our registration) in the first place is their first hurdle, and getting them to turn up to a polling booth is their second hurdle. You simply don’t try to achieve that via advertising; it has to be done on the ground, face to face, personal. I expect their advertising is intended to “motivate” the casting of a vote, but it doesn’t!

    There seems to be no inclination to go electronic here, which is interesting. Not from the public, and definitely not from the parties. Don’t see why everyone wouldn’t simply support it; can’t see any disadvantage to the parties.

    Come to think of it, hasn’t the ACT trialled electronic voting?

  4. Sacha says:

    The ACT has trialled it. Conceptually, it’s great – but as shown in the US, it’s probably useful to have physical votes for each electronic vote cast for recounting and checking purposes. Of course you need backup systems in case the system isn’t working (e.g. a blackout or there’s a problem with the computer).

    An advantage with electronic voting is that you can remove the donkey vote effect as the order of the candidates on the screen needn’t be fixed beforehand – the order of candidates could be randomised each time someone votes. The advantage of reducing the donkey vote has to be weighed up against the disadvantage that voters may wish to be guided by a political party as to how they vote. Having worked as a polling official for the electoral commission, I’ve seen that many voters follow a how-to-vote card in casting their vote.

    Another advantage with electronic voting is that, as soon as all the votes are in the system, the election result is known relatively quickly. If one candidate is to be elected, the result can be known almost instantaneously. If a number of candidates are to elected by proportional representation, it will take a bit longer, but it’s much shorter than counting it manually!

    Indeed, in casting a vote over the internet, the system could be such that the vote is cast and then unable to be changed, or it could be such that one could change the vote until the last minute. For the American Mathematical Society elections, you can log in and change your vote over several weeks until you’ve cast it. The system could be that you can cast it and then change your mind and recast your vote – but that would be a fundamental change to a core element of the electoral system. It’s up for debate whether that would be a desireable change or not. I wouldn’t support it.

  5. jchester says:

    Sacha, the donkey vote is easily avoided by printing x different versions of the ballot, where x is the combination of the orderings. I imagine you know which version of nPr or nCr spits out the answer 🙂

    As for internet voting generally, it’s OK for small organisations because the stakes are low. For government elections I do not endorse it. It is too easy to attack and to attractive not to be attacked. The Australian system is expensive but it is very very hard to game.

    More generally, I believe Bruce Schneier has stuff on why online voting is a bust.

  6. jchester says:

    “too attractive”.

  7. Sacha says:

    Jacques, at a moment’s thought, I think that you would need n! different ballots to remove donkey voting. n choices for the first position, then (n-1) choices for the second, and so on.

    n! gets very large very quickly! Say you have 20 candidates, how many different ballots would you need?

  8. David Barry says:

    If you just want to get rid of the donkey vote, then you only need n different ballot papers (ie, the Robson rotation system that they use in Tasmania). 1/n ballots have the order A,B,C,D, 1/n have B,C,D,A, etc.

  9. David Barry says:

    Oops. I mean N/n ballots have the order A,B,C,D etc., where N is the total number of ballots printed and n is the number of candidates.

  10. Sacha says:

    Hi David,

    While the Robson rotation method looks as if it removes the impact of the donkey vote, in a single-member system it doesn’t. Assume that the two major party candidates are in positions D and E on a ballot in which there are a total of 5 candidates. In the Robson rotation, the five different ballot orderings are:


    A donkey vote on any of the first four ballots results in a preferential vote going to D while a donkey vote on the fifth ballot results in a preferential vote going to E. While the Robson rotation lessens the impact of donkey voting, it doesn’t remove it in general.

  11. David Barry says:

    Ahhh, clever, I hadn’t thought of that.

  12. of course like your web site but you need to take a look at the spelling on several of
    your posts. Many of them are rife with spelling problems and I
    find it very troublesome to inform the reality however
    I will surely come back again.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: