Welcome to the house of fun

Foreword: Most of the following was written before I noticed that pretty much everything in it had already been said by Antony Green.

The Victorian Electoral Commission has released its proposed boundaries for the reformed Legislative Council, the political implications of which are discussed by Paul Austin in The Age. Austin reckons it "one of the most remarkable own goals in Victoria’s political history", for two reasons. Firstly, Steve Bracks does not appear to appreciate the likely impact on his party’s fractious state branch of sending 25 sitting members chasing after 20 or so winnable seats. Secondly, he has ensured that Labor "will almost certainly never again have the Legislative Council". The first point is political which makes it Austin’s turn rather than mine, and I do not doubt for a moment that he is correct. But the second is psephological and here I flatter myself to imagine that my assessment might be worth something.

Previously, the Council consisted of 22 provinces each made up of four lower house districts, which were represented by two members elected at alternating elections and serving eight-year terms. Under the reforms, the chamber will consist of eight five-member regions covering 11 lower house districts each, elected by much the same method as the Senate and other mainland state upper houses. The quota for election will be 16.7 per cent, in common with five-seat regions for the Legislative Council in Western Australian and the Legislative Assembly in Tasmania and the Australian Capital Territory. Conveniently, no redistribution has been conducted for the lower house, so it can easily be inferred what outcome would have been produced by the results from the last election.

In carving up the pie, the commissioners have predictably aspired to create distinctly metropolitan and rural regions, for which there is room for five and three respectively (despite what the Nationals think – they proposed two non-metropolitan regions in the hope of concentrating their support). Two of the rural electorates largely draw themselves – Western Victoria, from the outskirts between Melbourne and Geelong out to the South Australian border, and Northern Victoria, from Mildura in the west to the line of the Yarra and Great Dividing Range in the east. Eastern Victoria untangles itself from Melbourne a little more messily, sneaking into the southern part of the Mornington Peninsula and absorbing two electorates (Monbulk and Evelyn) that the sprawl is beginning to catch up with.

Moving into the city, the Yarra is an obvious dividing line since it contains 23 districts to the north and 32 to the south, allowing for a neat division after one northern district (Eltham) is conceded to the south. The division of the two northern regions leaves the inner city seats of Melbourne and Richmond at the south-western extremity of the Northern Metropolitan region. South of the river, Southern Metropolitan consists of the Liberal-leaning areas nearest the city, South Eastern Metropolitan covers a more distant stretch of the bayside inland to Narre Warren, and Eastern Metropolitan takes in the electorally volatile mortgage belt from Box Hill west to Kilsyth and Eltham south to Ferntree Gully.

The voting system will be the same as that introduced for the New South Wales Legislative Council for the 2003 election, with voters able to number either a single box above-the-line (as per the Senate and the upper houses of Western Australia and South Australia) or a number of boxes below the line equal to the number of vacancies up for election (also the case for the Tasmanian lower house). One difference in relation to the below-the-line option is that New South Wales elects 15 members from a single statewide electorate whereas Victoria’s regions have five members, meaning below-the-line voters will need to number only a third as many boxes. Even so, the rate of above-the-line voting is unlikely to be dramatically lower than the 98 per cent recorded at the 2003 election in New South Wales. This leaves open the prospect of results determined by the parties’ registered preference tickets, such as that in the five-member Western Australian region of South Metropolitan where the Fremantle Hospital Support Group came within an ace of winning a seat at the recent state election from 1.3 per cent of the vote.

The following table presents the votes in each region based on the results of the 2002 election. Obviously this election was atypical, so I have massaged the data to produce a second set of results in which both parties score an equal share of the statewide major party vote. Another point to consider is that the adjusted figures on the right greatly underestimate the likely "others" vote, which ran at 2.7 per cent at the 2002 state election compared with 11.0 per cent in the Senate last year. The latter might be thought a more reliable guide as there will now be many more candidates, due to the bigger electorates and better prospects for success.

. 2002 RESULT 50/50 OUTCOME
Northern Metropolitan 56.8 23.3 17.0 2.9 53.1 27.0 17.0 2.9
Eastern Metropolitan 43.4 44.1 10.7 1.8 38.8 48.7 10.7 1.8
Southern Metropolitan 39.5 43.6 15.0 1.9 35.5 47.6 15.0 1.9
Western Metropolitan 62.3 25.3 9.8 2.6 58.3 29.3 9.8 2.6
South Eastern Metropolitan 54.4 34.7 8.9 2.0 50.2 38.9 8.9 2.0
Northern Victoria 38.8 48.9 8.9 3.4 34.3 53.5 8.9 3.4
Western Victoria 47.2 42.4 8.3 2.1 42.8 46.8 8.3 2.1
Eastern Victoria 41.1 47.5 10.1 1.3 36.8 51.8 10.1 1.3
Total 48.7 39.0 10.1 2.2 43.9 43.9 10.1 2.2

Northern Metropolitan. All parties’ submissions except Labor’s agreed that this region should contain the Greens’ two strongest divisions, Melbourne and Richmond, boosting their vote above 16.7 per cent quota and assuring them of a seat. Labor can be equally sure of three seats since its vote is above 50 per cent on every measure, with the leftover going to the Liberals.

Eastern Metropolitan. The various proposals for this region did not differ dramatically. Labor and Liberal can expect to win two seats each with the fifth going either to the Liberals or the Greens, or possibly Labor if there is a repeat of the 2002 landslide. For reasons explained in the conclusion below, this region looms as the one to watch if Labor is doing particularly well.

Southern Metropolitan. A similar story to Eastern Metropolitan, except that Labor’s chances of a third seat are non-existent rather than slight.

Western Metropolitan. Labor would have the latte left divisions of Melbourne and Richmond located here rather than Northern Metropolitan, although they clearly share a closer community of interest with Northcote and Brunswick than Footscray and Essendon. Labor would presumably like a thinner spread of the Greens vote, although if they had their way they might well win seats in both regions. Significantly, the Liberals wanted to be stronger in this region by having it extend far beyond Labor’s western suburbs heartland to Geelong. This would have given them a much better chance of winning a second seat, which would have loomed large in their scenarios for a potential upper houes majority. As things stand, it’s a certain three seats to Labor and one to Liberal, with Liberal and the Greens battling it out for fifth spot. P.S.: Antony Green wisely notes "the odd chance of a minor party winning Labor’s third seat on Liberal preferences". The Greens’ vote is low enough that a fringe player could conceivably get ahead of them, while Labor have easily enough votes for three seats but not enough for four. If Glenn Druery’s out there, he might care to give this one a bash.

South Eastern Metropolitan. All proposals had this as the least Greens-friendly region in the metropolitan area. Labor would usually win three seats and the Liberals two, although this would sometimes be reversed.

Northern Victoria. The quirkiest aspect of the Liberals’ submission was their recommendation that this region encroach upon the metropolitan area to take in Broadmeadows, this being a necessary consequence of their scheme for Western Metropolitan outlined above. This would have come at the expense of a certain third Coalition seat here, the victims of which have been the Nationals rather than the Liberals. As it stands, it’s hard to conceive of an outcome here other than two Labor, two Liberal and one Nationals. P.S.: Antony Green reckons the Liberals and Nationals would compete for the final seat, correctly calculating that a normal result would give the Liberals a substantial surplus over their second quota.

Western Victoria. The Liberals wanted this region to be stronger for the Coalition by including the locality of Bendigo rather than Geelong. Instead, the Coalition will have a fight on their hands to win a third seat. The Greens are not without a chance, but it will usually be a race to see if Labor can get enough preferences from them to pip the Coalition. The urban territory means that a third Coalition seat would be unlikely to go to the Nationals. P.S.: Antony Green says the Nationals are an outside chance if Labor polls poorly and gives them preferences ahead of the Liberals.

Eastern Victoria. The VEC’s proposal for this region is the same as Labor’s submission, while the Liberals’ differs by two seats. The Liberals’ proposal would have increased the chance of a third Coalition seat going to to the Nationals rather than the Liberals, although a Nationals win would still have been more likely. In any case, an outcome of three Coalition and two Labor is almost guaranteed.

Tally all that together, and a landslide could see Labor win an absolute majority of 21 seats, especially if they get the wind in their sails in Eastern Metropolitan. The Coalition faces a bigger challenge in that there are two regions where they can normally expect to win only one seat. Any scenario for 21 seats requires that they win two in either Northern Metropolitan or Western Metropolitan, the latter looking more likely. The Nationals can expect only two seats rather than their current four – as Paul Austin notes, this gives them little hope of maintaining the 10 seats required for party status, unless they can dislodge those pesky lower house independents (it should be noted that one-vote one-value legislation has left their Western Australian counterparts similarly placed). The Greens are likely to secure a permanent home in Northern Metropolitan, but beyond that their prospects are hard to call. They could conceivably win as many as five seats, but two would be more typical. Along with the odd independent or micro-party member, there is little doubt that they will hold the balance of power more often than not.

Author: William Bowe

William Bowe is a Perth-based election analyst and occasional teacher of political science. His blog, The Poll Bludger, has existed in one form or another since 2004, and is one of the most heavily trafficked websites on Australian politics.

9 comments on “Welcome to the house of fun”

  1. Hi Just thought I would mention that originally the ALP had proposed a division of the state into five provinces. The five x seven or five by nine option would have been far preferable then the current eight by five option. Why the VEC did not consider the five seat option I fail to understand? It would have meant an increase in the upper house by one member but could have been offset by a reduction from 88 to 75 in the lower house.

    The five seat option also facilitated a better division of Victoria with two rural seats (East and West divided along the Hume corridor and three inner urban electorates)

    Anyway this is all history and in typical VEC fashion they need to make there own mark. The various models imposed on their Local Government reviews is a patchwork quilt of various representative systems.

    My analysis of the new boundaries certainly shows it will be difficult if not impossible for the ALP to win control of both houses. John lenders is a man of high integrity and whilst he originally favored a five by seven option has delivered on the ALP promise of reform. Reform that is long overdue.

    When I worked for Evan Walker one of my main tasks was to consider issues related to the reform of the Upper House. Something had to be done. The ALP had its one and only chance before 1999 prior to the Nunawading by-election to reform the electoral system but personal power games by the likes of George Crawford and Jean Mclean put an end to any hope of reform at the time as they white anted the proposal of reform to save their seats which would have been abolished.

    In contrast to the ALP’s position the Liberal/National Party has a better then 50% chance of winning both houses.

    I have published elsewhere on this site and on my Melbourne City Council blog (www.melbournecitycoucnil.blogspot.com) a preliminary analysis of the eight new electorates based on the Victorian Senate Vote.

    Whilst I am more then aware of the potential differences in voting between the Senate and State parliaments the fact is that the Senate vote is the best and only source where all parties have run and under a similar system to the one that is about to be implemented. It shows a favorable outcome is available to the Liberal party who could secure 21 out of 40 seats. The ALP around 16-17 the Greens at 2-3 at best and the Nationals 2. The key is in the fact that each electorate returns 5 members with a quota of 16.67% the thresholds are such that only the ALP and the Liberal party secure a quota in their own right with all parties fighting for the last seat. In the inner west the ALP is assured of 3 seats with a possibility of obtaining 4 in all optehr seats they get 2 and could be opeged back to one in one seat. The Greens are strongest in the Northern and South Eastern mero seats but they are not a shoe in as tey fall below teh 16% mark in each case. We will know the potential outcome soon after the close fo above-the-line registration.

    With the VEC seeking to implement for the first time a computerised counting system we hope and trust they will provide copies of the detailed preference data for independent analysis. The system as it stands is already stacked against minor parties as the formula used to calculate the surplus vote transfer distorts the one vote one value principle. It is also further complicated by a system of segmentation designed to facilitate manual counting systems and to offset/limit the distortion of the surplus transfer formula (the current formula uses the number of ballot papers as opposed to the value of a ballot paper when calculating the transfer value. Votes that were allocated at different value are transferred on at an equal value disportionally from their original value – this favors major parties – which explains why no one has sort to address this anominally)

    Ideally there should be one trabsaction per candidate (Elimination or surplus) with all votes being transfered prorata and in equal proportion to the value of the vote in a single transcation.

    With the move to an elctonic ballot system the publication of the preference data file is the only way we can maintain an open and trasparent voting system. Without thsi information scutineers will not be able to verify the validity or correctness of teh vote as the counting is dependent on the quaility of data-entry which is open to human error. By making teh oprefernce data file available independent analysis, scrutiny and and verifiction of tee election results can be undertaken. Without teh system is wide oipen to abuse. I am pleased to have put this on the public record prior to the 2006 election. If teh VEC refuse to publish this information I will be taking them to VCAT to seek to have this data made public. The parliament is already aware of these issues and any close result will cause more concern then a fox in a chook shed.



  2. It should be printed out that the 2002 Sate election was favorable to the ALP. The swing of the pendulum is on the rebound and as such the vote for the ALP will most certainly not be as strong. Further not all parties ran in the upper house and the voting system was significantly different to the one about to be introduced. The system to be used is similar to the senate and will also have the above-the-line ballot paper. By Contrast the 2004 Senate vote in Victoria favored the liberal party. I have made allowance and taken into consideration in my analysis issues related to the preference deal between the ALP and Family First. Good preference deals will potentially effect two electorates. Finally it is wrong to assume that the greens will get close to or above a quota based on the lower house vote in Melbourne and Richmond. ion 2002 the Liberal party did not run strong campaigns in these seats giving the greens a better then normal chance. in 2006 the liberal party will need to campaign strong in all seats to maximise their upper-house vote and secure more public funding as such the battle will be on for the greens to top the Liberal party. in order for the greens to win the seats of Melbourne or Richmond the Liberal party would need to poll less then 20% which is unlikely this time round. Also favorable preference deals will be harder to negotiate.

    Melbourne has a better chance of an upset then Richmond. Richard Wynne is an experience candidate with a string local presence. the same can not be said for Bronwyn Pike. With an change in the demographics Melbourne could very well be in trouble and under attack from the greens with the ALP falling below 50% primary vote. It is easier to win a so called safe seat then a close marginal.

  3. Your 50.50 analysis above puts the ALP at 43% and the Greens at 10%. The 204 senate election the ALP received around 39% and the Greens around 8-9% state wide. The threshold for the Greens to win a senate seat was the Alp above 39.5% and the Liberal party below 43% As a result of the Alp’s poor result, at the bottom end of the polls, the the preference deals cut in and delivered Family first who picked up from all quarters the sixth seat. David Risstrom (Greens Senate Candidate) was aware of this fact two weeks prior to the ballot following the registration of above-the-line preference deals and recent polling.

  4. A question when you tabulated you table above what vote did you use? the lower house of the Upper House. You really can not apply the results of a single member electorate system to multi-member proportional representation system. The comparison is like oranges to apples. At least with the Senate system the comparison uses the same electoral system I guess its like comparing oranges with lemons grown within the same orchard. 🙂

  5. PS I wrote to Antony Green and he qualified the comments that were attributed to him by saying . (And I hope he does not mind me quoting him)

    “I am constantly misquoted on having predicted the Greens will win the balance of power. All I did was transfer the votes of the 2002 election to match the new provinces. If all the parties got the same votes in the same seats at the 2006 election, that’s what you would get. But they won’t, so it’s all matter of opinion.

    I personally don’t see why the 2004 Federal vote would be any better a guide to the LC result. It was Labor’s worst Federal result in Victoria for a decade. Labor’s primary vote will be better than that at the state election and the Liberal vote will be worse.”

    I agree with Antony Greens assessment and was also surprised by the misleading comments attributed to him.

    And yes the 2004 senate vote was the ALP’s worst but it does use the same system and also shows the bottom end of the ALP pendulum and the top end of the Liberal party support. More important it includes recent voting support of the other parties (Including the Greens who in 2002 did not stand candidates in all legislative council seat). Tis is te first multi-member Victorian state election all comparisons are valid in the absence of more reliable data.

    I would love it if you could debate further the issues related to the formula and process adopted in counting the vote as this does need review. It is only a matter of time until it becomes an issue. one vote – one value. One transaction per candidate based on the value of the vote not the number of ballot papers in each distribution. Too technical? I am happy to explain it in more detail.

  6. have re-published an extract of a paper I published in January this year you can find a copy here..


    I have not had a chance to edit this extract and trust that the version is up to date and relevant information is not missing.

    II sincerely believe that we should simplify the counting process and adopt a value based formula and a single transaction per candidate.

    Paper based formula distort the one vote one value principle and now we have a computer aided counting we can and should consider adopting a single transaction per candidate as opposed to the illogical segmentation process adopted by the VEC. (It would be great if we could also preferential vote above the line – but that’s a different issue – maybe when we have online voting this will be considered)

    There are other issues related to the scrutiny of the ballot.

    I refer readers to a paper published on “Save our Suburbs” by Ian Quick.


    * Comment published both here and in the other post on the Victorian Election http://blogs.crikey.com.au/pollbludger/401

  7. Note to the editor… I think it would be worth while if you could add to your above analysis a table showing the number of quotas as opposed to just the percentage of the vote. Integers would represent the number elected and the remainder would be distributed in the similar process to that which applies to a “normal” preferential election with the lowest value being eliminated and redistributed until a candidate obtains a full quota. There remains what is refer to a wasted quota the person or persons who miss out in being elected.

    Published below csv (Comma delimited) data.

    Your Analysis,,,,,,,,,,
    .,,2002 RESULT,,,,,50/50 OUTCOME,,,


    Northern Metropolitan,,56.8,23.3,17,2.9,,53.1,27,17,2.9
    Eastern Metropolitan,,43.4,44.1,10.7,1.8,,38.8,48.7,10.7,1.8
    Southern Metropolitan,,39.5,43.6,15,1.9,,35.5,47.6,15,1.9
    Western Metropolitan,,62.3,25.3,9.8,2.6,,58.3,29.3,9.8,2.6
    South Eastern Metropolitan,,54.4,34.7,8.9,2,,50.2,38.9,8.9,2
    Northern Victoria,,38.8,48.9,8.9,3.4,,34.3,53.5,8.9,3.4
    Western Victoria,,47.2,42.4,8.3,2.1,,42.8,46.8,8.3,2.1
    Eastern Victoria,,41.1,47.5,10.1,1.3,,36.8,51.8,10.1,1.3


    Number of Quotas allocated,,,,,,,,,,
    .,,2002 RESULT,,,,,50/50 OUTCOME,,,


    Northern Metropolitan,,3.408,1.398,1.02,0.174,,3.186,1.62,1.02,0.174
    Eastern Metropolitan,,2.604,2.646,0.642,0.108,,2.328,2.922,0.642,0.108
    Southern Metropolitan,,2.37,2.616,0.9,0.114,,2.13,2.856,0.9,0.114
    Western Metropolitan,,3.738,1.518,0.588,0.156,,3.498,1.758,0.588,0.156
    South Eastern Metropolitan,,3.264,2.082,0.534,0.12,,3.012,2.334,0.534,0.12
    Northern Victoria,,2.328,2.934,0.534,0.204,,2.058,3.21,0.534,0.204
    Western Victoria,,2.832,2.544,0.498,0.126,,2.568,2.808,0.498,0.126
    Eastern Victoria,,2.466,2.85,0.606,0.078,,2.208,3.108,0.606,0.078

    Number elected prior to eliminations,,,,,,,,,,

    .,,2002 RESULT,,,,,50/50 OUTCOME,,,


    Northern Metropolitan,,3,1,1,0,,3,1,1,0
    Eastern Metropolitan,,2,2,0,0,,2,2,0,0
    Southern Metropolitan,,2,2,0,0,,2,2,0,0
    Western Metropolitan,,3,1,0,0,,3,1,0,0
    South Eastern Metropolitan,,3,2,0,0,,3,2,0,0
    Northern Victoria,,2,2,0,0,,2,3,0,0
    Western Victoria,,2,2,0,0,,2,2,0,0
    Eastern Victoria,,2,2,0,0,,2,3,0,0


    Percentage of vote/quota be distributed,,,,,,,,,,

    .,,2002 RESULT,,,,,50/50 OUTCOME,,,


    Northern Metropolitan,,Quota filled,,,,,Quota filled,0.62,0.02,0.174
    Eastern Metropolitan,,0.604,0.646,0.642,0.108,,0.328,0.922,0.642,0.108
    Southern Metropolitan,,0.37,0.616,0.9,0.114,,0.13,0.856,0.9,0.114
    Western Metropolitan,,0.738,0.518,0.588,0.156,,0.498,0.758,0.588,0.156
    South Eastern Metropolitan,,Quota filled,,,,,Quota filled,,,
    Northern Victoria,,0.328,0.934,0.534,0.204,,Quota filled,,,
    Western Victoria,,0.832,0.544,0.498,0.126,,0.568,0.808,0.498,0.126
    Eastern Victoria,,0.466,0.85,0.606,0.078,,Quota filled,0.108,0.606,0.078

  8. “If Glenn Druery’s out there, he might care to give this one a bash.”

    Well HEY!
    Thanks for thinking about me,
    you never know I just might move back to VIC one of these days…..

    best regards

Comments are closed.