Voters in the Australian Capital Territory go to the polls on October 18 for the seventh time since self-government was established in 1989, delivering a verdict on a majority government for the first time. Labor won nine of the 17 seats at the October 2004 election against seven for the Liberals and one for the Greens. Such was the unpopularity of self-government when it was first introduced that eight of the 17 members elected in 1989 represented minor groupings committed to its abolition. Their shifting sympathies produced two changes of government in the course of the first term: Labor’s Rosemary Follett held the reins from May to December 1989 and again from June 1991, with Trevor Kaine heading a Liberal administration in the interim. Three opponents of self-government held the balance of power after the 1992 election, sustaining Follett’s minority government throughout the ensuing term.
At the 1995 election the bar for minor party candidates was raised with the abolition of the shambolic modified d’Hondt system of Territory-wide proportional representation. It was replaced with the present Tasmanian-style Hare-Clark system based on three multi-member regions, two of five members (Brindabella and Ginninderra) and one of seven (Molonglo). As in Tasmania, the Robson system of rotating ballot paper order means candidates are forced to compete against their party colleagues. The first such election produced a Liberal minority government of seven members, headed by Kate Carnell, which was sustained with the support of two independents. Carnell achieved a status quo result at the 1998 election, before resigning in 2000 to head off a no-confidence motion resulting from an unfavourable auditor’s report into the revamp of Bruce Stadium. Her successor Gary Humphries led the government to defeat in 2001, moving a year later to the Senate where he has remained ever since.
The change of government after the 2001 election followed Labor’s gain of two seats at the expense of Liberal-leaning independents, with the Liberals retaining their existing seven seats. Re-elected Greens member Kerrie Tucker had ruled out supporting the Liberals, meaning Labor’s eight seats were enough to ensure a comfortable hold on government without having to rely on the one Democrats member. New Chief Minister Jon Stanhope introduced four-year terms effective from the 2004 election (held one week after the October 9 federal election), at which Labor gained the one extra seat required for majority government at the expense of the collapsing Democrats. The vote for both major parties increased, Labor from 41.7 per cent to 46.8 per cent and Liberal from 31.6 per cent to 34.8 per cent, but the Liberals remained on seven seats with the Greens on one.
There are several reasons to think Labor will struggle to match that feat on October 18, most notably an electoral cycle that has lately shown its force in Northern Territory, Western Australia and various state Newspoll surveys. Stanhope’s government has accumulated a heavy burden of baggage in the last four years, the low-point coming with a horror 2006/07 budget that sought to balance the Territory’s tottering finances by closing 39 schools (nearly a quarter of the total). The ACT edition of the Daily Telegraph reacted by describing the Chief Minister as an economic vandal who headed a disgraced government, beneath the front-page headline: STANHOPE-LESS. A large part of Stanhope’s political strategy consisted of needling the Howard government over terrorism laws and civil unions for same-sex couples, which lost much of its utility after Kevin Rudd came to power.
However, Labor’s woes have been matched by an eerily familiar story of leadership disarray among the Liberals. There have been two leadership changes in the current term: from Brendan Smyth to Bill Stefaniak in May 2006, and from Stefaniak to the 31-year-old incumbent Zed Seselja in December 2007. Both Stefaniak and Seselja were compromise leaders of a kind, Stefaniak taking the reins because Smyth and rival Richard Mulcahy each mustered three party room votes out of seven. Stefaniak’s departure followed a dramatic week that began with Mulcahy being stood down from his portfolios pending a federal tribunal inquiry into his past activities as executive director of the Australian Hotels Association. Mulcahy did not take this lying down, claiming to know of unspecified allegations against Stefaniak and Smyth and calling on them to join him on the back bench. The party room reacted by unanimously voting to expel Mulcahy, who will run in Molonglo at the head of his own grouping. Shortly afterwards Stefaniak and his deputy Jacqui Burke announced they would clear the air by standing aside, and Seselja and Smyth were elected in their place without opposition.
Posts on each of three electorates will follow over the coming weeks. Antony Green offers plenty to keep you going on with in the interim.
90 comments on “ACT election: October 18”
William, it is about the ACT election and hwo the system works. Chances are the system will most certainly effect the outcome of the election as was the case in Queensland.
Antony Greens explanation above is only partially correct. Under a proportional STV count the vote should be proportioned on the value of the vote and not the number of ballot papers. The system is distorted when votes at a fractional value are transferred at the same value as votes that hold a full value. It is further complicated by the distribution of preferences allocated to excluded candidates. Under the current system the votes are dealt from the bottom of the pack. A very arbitrary system which was designed to facilitate a manual count. With the use ofcomputertechniologty we should be reviewing this method used. The variation in t6en count can effect the outcome of the electio0n
To answer the question above then rules of the count are spelt out in legislation.
Itep, yes they are set out in legislation. At the primary vote stage, it is easy, simply transferring a proportion of votes equal to the total above the quota. A candidate elected at a later count becomes more complex. Hare-Clark as used in the ACT looks at the last bundle of votes received that put the candidate over a quota to determine which votes go on as preferences. So if a candidate received 200 votes that put them 100 votes over quota, only these 200 votes would be examined to determine the 100 votes to be distributed as preferences.
The Senate system looks at all ballot papers held by the candidate at the point when they reached the quota, not just the last received. There are problems with the current Senate system in terms of which votes end up being transferred as preferences, something that democracy@work and I actually agree on, though you might not have noticed.
Beyond these simple answers, there is a degree of complexity in the counting process that should only be discussed between consenting adults.
Antony at 50. You comments are misleading. the vote most certainly does increase in value disporportional to the vote. The reason and example was given above instead of being transferred at 0.5 the method used by the ACT increases the value of the vote to 0.55 as a result of then exclusions. Whilst it does not increase above one it most certainly does increase disporportionally in value as it captures the value of the exhausted votes. try again. Try counting the Queensland election as though only 7 candidates had stood. And tell-us your findings.
democracy@53 – read the clause I mentioned. It doesn’t increase in value.
Its only complex because the system was designed to assist a manual count. teamcomplexity having been built into the system in the process.ideally there should only be surplus distributions based on the value of the vote.Excluded candidate’s votes should be transferred as though that candidate had not stood. IE the second preference becomes a first -reference. A reiterative count is the only way that can be achieved. If you counted the Queens land vote with only seven candidates you might begin to understand the effect that dealing from the bottom of the deck has. The ACT and Tasmanian Last bundle system was an attempt to mitigate the distortion. In the process they have introduced further distortions. Distortions that are no longer required now that we can computerised counting.
It increases proportionally.
If you have a 1000 votes and a surplus of 200 votes what is the transfer`value? 0.2
If 500 of those votes exaughsted (IE optional preferential) the transfer value increases to 0.4 b(That twice the value that they would have been tranfered at normally) disportionally to the original value.
read the clause, here’s the link. http://www.legislation.act.gov.au/a/1992-71/current/rtf/1992-71.rtf
If the formula would cause the transfer value to increase, the transfer value stays the same.
That is crystal clear
Maybe someone could set up a psephological blog dedicated to insignificant minor differences in counting methods so the rest of us can get on with more significant issues. I’m all for talking about electoral reform and improving the system, but can we talk about something bigger than fractions and increasing values of fractional votes?
WB, surely this qualifies as “boring” under your new moderation rules?
I agree with Bryan, it is not an issue generally. Both side has their positive It is like arguing that Hare-clark/Full Preference/Optional Preference system has effect on who get voted in, all of them are fair and valid systems
Last in last out is as good as proportional, why should votes the elected one person be used to elect another person later on?
Read the clause
How a fractional transfer value is calculated
The fractional transfer value is calculated using the following formula:
number of surplus votes (Dividedby)
total number of ballot papers with *** further preferences shown***
Ir if 1000 ballot papers prepresent a surplus of 200 votes they would have a value of 0.2. If 500 of those ballot papers exhaughst they value would increase to 0.4
(Or: divide the number of surplus votes by the total number of ballot papers with further preferences shown.)
Last bundle is not proportional it is`segmented and there is not real justification for rational for the segmentation. It is pot luck and we might as well go back to random selection or adopt a party list system., Again count the Queensland senate as though only seven candidates are left standing redistribute all the votes and tell me that the Greens were not elected. EACH VOTE MUST BE TREATED IN THE SAME WAY AND SAME MANNER. It is not a roulette wheel and should not be decided on chance or distortion in the way the vote is counted. Adopt either Meek (A approximation non linear counting method a reiterative linear count system) Anything else is just excuses for maintaining a flawed system.
Ben @ 59 it is so insignificant that the distortion produced a 50,000 vote error inj queensland and elected the wrong candidate. It happens more often then you think. Maybe 5-10% of the time we have only recently been able to detect it as they now publish the detailed preference data. but hey 50,000 votes is not much. Near enough is good enough hey? I guess it depends if your the one that wins or loses.
D@W, you’ve now posted 21 comments on this subject – I think that will do for now.
Blow the fact sheet, read the Act. And sorry to quote C code, but
static bool update_vote_value(struct ballot *ballot, void *unused)
/* Ballot vote value may only DECREASE */
if (fraction_greater(ballot->vote_value, new_vote_value))
ballot->vote_value = new_vote_value;
for_each_ballot(ballots, update_vote_value, NULL);
Implementing the Act.
[…sorry to quote C code, but…]
And at the 65th comment, nerdy chicks started swooning throughout the Pollbludger Yooniverse 😀
I should be banned for quote C code, but it was taken from the ACT Electoral Commission’s open source counting code.
I think open source might be the wrong term. It’s publicly available anyway.
Well, as long as you don’t start quoting FORTRAN 🙂
Then again, if you did it may not be too dissimilar to some late night posts you find around the US election thread! 😀
Getting serious for a moment though – it’s interesting that the powers that be release that stuff publically. If only Diebold machines and their equivalents were so transparent.
Why they moved away from paper to begin with across the pond makes me shake my head.
Has there ever been any serious push in Australia that you can remember to move across to electronic voting?
Not as long as I know members of the Joint Standing Committee on Electoral Matters they’re won’t be.
oops “there won’t be.”
There has yet to be a serious push for electronic voting in Australia. It is likely in the future that voting in overseas posts, and in multi-electorate booths like Sydney Town Hall or Brisbane City Hall, could be conducted by an intranet based system. The main reason would be to cut the huge administrative load caused by having to have copies of every possible ballot paper available, and then have to reconcile each electorate seperately. Central city early voting centres could also use this sort of system.
The main thing that pushed mechanical and later electronic voting in the US was the number of elections conducted on the same day. There can be dozens of elections and plebiscites. President, House, sometime Senate, state upper and lower houses, Governor, statewide offices, local government, local statutory officials, judges, plus initiative referendums. They all get printed on the same ballot which is very difficult to count. That’s why they invented those odd punch card systems.
There is testing being done on OCR counting equipment, but if the cost of the equipment costs more than keeping the polling staff back an extra hour to do the count manually, there isn’t that much to gain. The OCR equipment would more likely be useful when the votes are check-counted in the RO’s office after being returned from polling places.
The ACT uses computer voting for pre-poll and some absent votes, usually 5-10% of the total vote. They release this count not long after the close of polls, well before the manual counts of the Robson-Rotation ballots are reported. The ACT EC did a major report on rolling the system out to all booths, but the cost was horrific and the only way to keep the cost down was to conduct the election at a smaller number of polling places over a period of several days.
When the ACT EC releases the pre-poll count, they also do a full distribution of preferences on it. In the two weeks after polling day, all ballots are data entered. Twice a day they do the count with full distribution of preferences. It can be fun in close counts as candidates appear to be elected and later unelected with the progressive counts. Of course, the only official count is the final one.
The ACT EC report on its use of electronic voting can be found at:
I don’t see anything wrong with the pencil and ballot paper, and votes counted by hand in front of scrutineers. People know this system and trust it, and it’s very hard to rort.
Excellent – thank you Antony.
And keep up the good work Dr Carr – may you howl early and often in outrage to any member of the JSCEM that even looks like having a random thought about implementing electronic voting.
That is inded my attention, Mr Marsupial, whether you are being sarcistic or not, which I have given up trying to determine.
I’m deadly serious Adam.
And you’re right – I need to use more smileys and things at times.
Adam, good to find something we agree on. I’m quite supportive of using electronic voting for rare situations – for the blind for example, and for Antarctic voters. However, the idea of using it as the main system in a federal or state election is just mad. I had a chat the other night with a computer programmer who was all excited about the idea. I just could not get through to him that the benefits were small. The risk may be small as well, but the consequences of things going wrong are so diaboldical that the idea is terrible.
This discussion re electronic voting is approporiate to the current thread – Elections ACT offers electronic voting for pre-poll voters and at each of the major polling booths in Civic, Woden, Belconnen, Tuggeranong and Gungahlin.
Does anyone here have any knowledge of the system used and its means of ballot security and vote registration/transmission?
Australia is lucky to have an electoral system in which the people have total confidence. In 1998 the ALP won a majority of the 2-party vote, but failed to win enough seats and Howard was re-elected. Were there riots in the streets? Did Beazley “refuse to accept the result”? Did it all finish up in court? No, we just accepted that these things sometimes happen. No-one doubted the integrity of the system. Electronic voting may be marginally quicker, but who cares? We don’t *need* to know the results 10 minutes after the polls close. Against that, electronic voting is very expensive, will be alienating for many many voters, will cause a higher informality rate, and will be vulnerable to rorting, crashes and hacking. There is no computer system in the world that cannot crash and cannot be hacked. I say, let’s stick with what works and what people trust – the Great Australian Pencil.
Besides which, scrutineering is fun because you get to go nyah nyah nyah at the Liberal scrutes when Labor wins, as it always does.
In ’98, Howard should have done a Steele Hall and resigned and formally advised the G-G to appoint Beazley as PM pending a fresh election
And Hawke should have done the same thing in 1990 when Peacock polled a majority of the 2PV but still lost? I don’t think so. Our electoral system is based on winning seats, not on aggregate votes.
Steele Hall recognised that getting a majority of seats without a majority of the vote was unworkable – Hawke and Howard should have as well
Steele Hall didn’t do that, Mary. He governed for two years after the 1968 election, during which time he introduced electoral reform. An election was then held after the government was defeated in parliament over an unrelated matter.
Besides which, what reason is there to think that a new election would have produced a different result?
I agree with Adam about electronic voting. There are minor benefits and minor risks, but they aren’t worth it.
The only system that could work would be if a person voted on a computer then had their vote printed out in front of them and they then put them in a box old-style. This would solve the problem of accidental informal voting and incomprehensible voting, and would allow for a much faster count (either by counting the electronic version with paper as a backup, or by using a computer to count the printed ballots), but wouldn’t help much with paper. You need a paper trail.
And Mary, you can’t expect John Howard to not claim victory if he wins. A victory is a victory. It’s a much bigger problem than that. You can’t fix electoral reform by having a new election whenever the result is more grievously out of whack then we are willing to tolerate.
Obviously the legitimacy of an electoral system based on single-member seats depends on the seats having approximately equal numbers of voters. In SA under the “Playmander” the state constitution said that there must be 2 country seats for every Adelaide seat. Since by the 1960s Adelaide had twice the population of the rest of the state, that meant that country seats had on average 25% the enrolment of city seats. Labor won the 1962 election by a wide margin in terms of votes, but not in seats. They only won in 1965 because suburban sprawl beyond the metro Adelaide boundary created new “country” seats which Labor won. Once Playford retired the Libs stopped trying to defend the malapportionment and Hall carried his election reform. Now the only seriously malapportioned chamber left in Australia is the WA Leg Council – thanks to the Greens.
Re the 1998 result – the majority of seats v majority of the vote thingo wouldn’t be such a big deal except that every now and again there will be a Howard apologist claiming a popular mandate from the 1998 election for the introduction of the GST in 2000.
Actually Adam, electronic voting would mean fewer informal votes, not more. Electronic voting would allow the computer to tell people if their vote was informal, so they could have another go if it was an accident, not deliberate.
However, all your other arguments are quite valid, and easily outweigh the “we won’t have to wait for the results” argument.
Comments are closed.