New South Wales redistribution: take two

The New South Wales boundaries have now been finalised as well. Geographically dramatic changes have been made to the large electorates in the west after the original proposal had Parkes occupying the entire north-western quarter of the state. It has now traded in more than two-thirds of its total area as originally proposed for the Wellington and Mid-West Regional shires to the east of Dubbo. The state’s north-western vastness will instead be divided between Calare and Farrer, the latter of which loses the Murrumbidgee shire to Riverina. All affected electorates are safe for the Coalition except independent MP Peter Andren’s seat of Calare, whose centre of gravity has moved still further from his home base of Orange.

Elsewhere, a small amount of rejigging has been done around the junction of Paterson, Newcastle and Hunter; changes have been made to the boundary between Parramatta and Reid after the original redistribution deprived the former of the Parramatta town centre; and various adjustments have been made affecting the boundaries of Wentworth, Kingsford-Smith and Sydney. The comments thread of the previous entry contains much productive discussion of the likely effect of these changes.

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.

180 comments on “New South Wales redistribution: take two”

Comments Page 1 of 4
1 2 4
  1. from memory: Wentworth, North Sydney, Parramatta, Werriwa, Macquarie, Hunter, Robertson, Newcastle, Cowper, Richmond, New England, Eden-Monaro; Melbourne, Melbourne Ports, Kooyong, Flinders, Gippsland, Indi, Ballarat, Bendigo, Corio, Corangamite, Wannon; Brisbane, Moreton, Wide Bay, Capricornia, Herbert, Kennedy, Maranoa; Perth, Swan, Fremantle, Kalgoorlie. If you count the 1903 seats in SA and Tas: Adelaide, Boothby, Hindmarsh, Barker, Grey, Wakefield; Denison, Bass, Franklin

  2. It is indeed deplorable that Gwydir has been abolished. Parkes could quite easily have been called Gwydir, and Calare renamed Parkes. Calare is not a federation seat (created 1906) and its name has no particular significance.

  3. Except Gwydir is a geographic name for a region far from Dubbo and Mudgee at the southern end of the new electorate of Parkes. Maintaining the name Gwydir might maintain a nice historical link, but it really doesn’t describe where the electorate is, and more particularly, its centre of population.

    Calare is indigenous word that is not as easily identified with a specific region. Parkes was a Federation seat, like Cowper, Robertson and Martin, chosen to recognise the great 19th century NSW Premiers. Like Martin, Parkes was abolished in the 1960s. As Parkes is recognised as the father of Federation (even if that is an overly simplistic view of what Parkes was trying to achieve in pushing Federatioon in 1890), it has some worth in being retained in place of Gwydir.

  4. Gwydir is named for the Gwydir River, which flows from Armidale to Collarenibri, right across what is now to be the seat of Parkes. This seat has boundaries very similar to those of Gwydir over the years, and should have retained the name

    The federation seat of Parkes was in inner Sydney. This Parkes was created only in 1984, and has no particular regional identity

    Calare (which should be pronounced Kalari) is an indigenous name for the Lachlan River, which flows from Gunning to Balranald, and which enters the new seat of Calare only at its very southern edge.

    It would thus have been perfectly logical to abolish Calare and call this seat Parkes, while preserving the name Gwydir for the Dubbo-Narrabri seat.

    I believe the Act actually requires the commissioners NOT to abolish federation seats if they can avoid it, and here they clearly could have avoided it.

  5. Gwydir technically ceases to exist on the day the current Parliament is dissolved doesn’t it?

    Wonder if Anderson will say anything meaningfull about it in his farewell speech

  6. Members cease to be members EITHER when the Governor-General dissolves the House OR when its term expires (which has only happened once, in 1910). Anderon will cease to be Member for Gwydir at that time. I suppose you could say that Gwydir will thus cease to exist, but an electorate is not a legal entity.

  7. Gwydir holds great identity significance to rural folk [I happen to know quite a few and was up there watching the tv opinions at the time of the first discussion of it being abolished] – It doesn’t matter that Gwydir might not be as important as Parkes- but it’s an identity for the rural people to say they’re from the Gwydir electorate. Abolishing it does loose a sense of both history and identity. Australia has reached past the 20 million mark and keeps expanding at a great rate… the parliament should therefore be expanded to say 152 seats… leaving all in tact and creating two new seats in QLD where growth is at its maximum.

  8. Expanding the House to 152 seats would not be a bad idea, Gwydir would have been retained in those circumstances and Queensland would have gained another electorate in addition to Flynn which was abolished.

    The next redistribution will see most likely an Victorian or another South Australian electorate abolished to make yet another Queensland one.

  9. Politics_Obsessed, the size of the House of Reps is (plus minus a small number) pretty much determined by the total number of senators. You can’t just arbitrarily expand the house of reps from 150 to 152 members (except by adding members to the territories, which wouldn’t be done at present). The number of seats in each state is determined by the relative populations of each state, and so Qld will have one extra member, and NSW will have one fewer.

    If you want to increase the size of the House of Reps, you could add an extra senator to each state, (to 13 senators) which would increase the size of the senate to 82 (78 state senators + 4 territory senators), and the House of Reps would then “start with” 156 state members + 4 territory members (currently there are 146 state + 4 territory members). The extra HR members might possibly be allocated to the states as follows: NSW 4, Qld: 3, Vic: 3, SA 1, WA 1 giving a total of 158 state + 4 territory members = 162 members. (I think that Tasmania has effectively 2 extra members than its population would warrant.)

    The point is that the process of allocating seats to states does not have a huge amount of flexibility under current laws, once the number of senators from each state is fixed.

  10. I’ve read that in the US, there are squeals for the size of the House of Representatives to be increased each time the 435 seats are reallocated amongst the 50 states by members of congress in states which lose seats. Of course, those congressmen might well feel that their district is the one that will be abolished by the state legislature, which could be a well-grounded fear given that that US legislatures often draw boundaries to benefit or disadvantage particular members.

  11. You couldn’t add one Senator to each state, you need two to keep the half-senates intact. You can fiddle with the Territories though. Or a create a new State. Only original states are required to have equal representation in the Senate.

  12. I wondered about this – if half the Senate needs to be elected each time, you might be able to elect 7 senators from 3 states and 6 senators from the remaining states in one election, and the next election swap the numbers elected around.

    If Darryl’s right, then you’d have to add at least 24 members to the House of Reps.

  13. I found these relevant sections of the Constitution:

    9. The Parliament of the Commonwealth may make laws prescribing the method of choosing senators, but so that the method shall be uniform for all the States. Subject to any such law, the Parliament of each State may make laws prescribing the method of choosing the senators for that State.
    The Parliament of a State may make laws for determining the times and places of elections of senators for the State.

    13. As soon as may be after the Senate first meets, and after each first meeting of the Senate following a dissolution thereof, the Senate shall divide the senators chosen for each State into two classes, as nearly equal in number as practicable; and the places of the senators of the first class shall become vacant at the expiration of three years, and the places of those of the second class at the expiration of six years, from the beginning of their term of service; and afterwards the places of senators shall become vacant at the expiration of six years from the beginning of their term of service.
    The election to fill vacant places shall be made in the year at the expiration of which within one year before the places are to become vacant.
    For the purposes of this section the term of service of a senator shall be taken to begin on the first day of July following the day of his election, except in the cases of the first election and of the election next after any dissolution of the Senate, when it shall be taken to begin on the first day of July preceding the day of his election.

    14. Whenever the number of senators for a State is increased or diminished, the Parliament of the Commonwealth may make such provision for the vacating of the places of senators for the State as it deems necessary to maintain regularity in the rotation.

    My reading of this suggests that you could have an odd number of senators for each state, it’s up to the Senate to “…divide the senators chosen for each State into two classes, as nearly equal in number as practicable;” which might be a 7/6 or a 6/7 split for 13 senators.

  14. Section 9 might require that the same number of senators are elected each time. In that case, you could elect 7 senators from each state in one election, and then 6 senators from each state the next election, and so on. This might seem a bit odd, but it looks possible.

  15. Antony: I didn’t say Armidale was ON the Gwydir, I was making the general point that the river flows through the electorate now to be called Parkes. The source of the Gwydir is actually near Uralla about 20k south-west of Armidale.

  16. The next redistribution is mre likely to abolish a seat from South Australia rather than Victoria. If I remember correctly Victoria has well over 27 quota’s last time. However would WA be close to gaining another seat?

  17. Peter, look at the numbers. Victoria’s last three quotas have been 36.76 (1999), 36.65 (2003) and now 36.54 (2005). That 37th seat is a goner.

    South Australia on the other hand: 11.6 (1999) to 11.4 (2003) to 11.2 (2005) is in little danger of losing a seat in the next redistribution.

    Similarly, there’s little chance of Western Australia gaining a seat next time on current trends. 14.52 (1999), 14.50 (2003), and now 14.60 (2005). [Sidenote: Was WA in danger of losing it’s 15th seat in the last redistribution? Or would they have simply have fudged it with the MoE argument as for NT?]

  18. As for talk about increasing the size of parliament. There must be a serious chance of the next ALP govt (whenever that might be) legislating an extra two senators per state. Like the Hawke govt did in its first term. Think of it as the future legacy of the Coalition’s current Senate majority.

    And back to the original topic, I saw little problem with dropping the name Gwydir (federation seat or not) on the proposed boundaries, as it was clearly the seat truly ‘abolished’. Now I think that prize goes to Calare. And as Adam has insisted, the augmented commission should have altered the names accordingly.

  19. Just a little personal aside in response to Sacha’s comment concerning the 435-seat ceiling to the US House of Representatives. Americans call it “re-districting” when the seats are reallocated. I lived in Connecticut for 3 years, at a time when there was a national re-districting in the pipeline. There was a census, and the state of Connecticut was terrified of losing a seat, so the censors made darn sure to try and count EVERYONE in the state. Even I was asked to fill out a census form – and I wasn’t even an American citizen or resident! (Alas, this was all to no avail; Connecticut lost a seat.) The US re-districting system penalizes not only states that lose population, but also states that don’t grow fast enough to maintain the same or greater percentage of the total US population. It’s a tough system.

    Back to the Oz HoR… I would prefer a constitutional change by which the number of Senate seats is decoupled from the number of HoR seats AND the number of HoR of seats was vastly expanded, with a quota for all mainland seats in line with the current quota for overrepresented Tassie. Of course, this would require a referendum and bipartisan support… and we know how EASY that would be. Back to my pipe dreams…

  20. ah cheers sacha that just slipped my mind at the time of posting…

    Well 4 senators for each territorry wouldn’t seem too bad. Then again… creating another state [say Nt becoming like North Australia] would be a bit premature [Isn’t there someone where stating a state must have at least 5 electorates in the House of Reps… that would greatly over-represent the territory].

    It’s ironic that this distribution raises so many questions about the state-fed structure. Simply abolishing one seat in one state and then putting it in another state – wouldnt it be easy to transverse across the state lines? That could open a whole ‘nother can of worms though but the feds can easily see through state boundaries.

    But I think the better thing to do with expanding population is to create a new state -that be the State of Sydney and the rest of the state of NSW remains intact. [One look at shows that ‘New England’ ie Northern part of NSW almost became a state in the 1960’s which would have avoided this wider distribution problem … well more delayed it.] Creating this new state would allow 12 new Senators without being quite dis-proportiante compared to making more senators in territories and tassie and wa for example. It could be argued that it be better to break a full house of power on the states ie the chances of Labor {2002 to current} or Liberal/Nats controlling them all. It could also be argued that the NSW state would be less dominated with Syd issues and can focus on their own issues. Creating a state of syd could also see the rest of nsw have a chance not to have ridiciously high taxes. And in one sense – it’s a step foward to the ‘regions’ proposed in the ABC QLD Election forums and the fact some overseas countries have states based on cities.

    What does everyone think would be the best long-term solution?

  21. There was a referendum on breaking the ‘nexus’ between house and senate in 1967 (same day as the Aboriginal powers referendum actually) which was rejected.

    Something tells me that if a proposal for more politicians was put to the Australian people …… it would fail also!

  22. I don’t dispute that the AEC is an independent entity, but it is strange how often independents get shafted in these sort of major redistributions. If anything, this final version of the new boundaries seems even worse for Andren than the first version was. It seemed like a similar happened with Pauline Hanson when Oxley got cut in half in the 1998 redistribution (going so far as to inexplicably draw the boundary right through the middle of Ipswich).

    As far as I know, in determining redistributions there’s also no criteria about whether sitting members are retiring and the like, but it seems to be a bit more than coincidental that the seat that is being abolished is of a retiring member.

    These are just gut feelings rather than anything based on extensive research, so it may be that someone can correct me with more solid evidence.

  23. Only original states need to have the current 12 senators. New states can be allocated different numbers of senators. You can’t have seats crossing state boundaries without a change to the constitution.

    If people are concerned about very large seats, the easiest way is deal with this is to (a) include some guiding principle in the criteria to make seats not geographically large if possible, (b) increase the 10% margin when drawing seats, (c) weight seats (as in Qld) but that won’t necessarily work and I’m very much against weighting, (d) increase the number of seats by increasing the number of senators.

    The thing is, there aren’t many people in western NSW, so you’ll always have very large seats there.

  24. lol so true blackburnpseph – as much as one hates the ‘nexus’ at least it does keep some proportion into parties power and balance in the two houses….

    Just read from ABC that Lawerence Springborg aka Bumpkin ‘Borg has stepped down as Opposition leader. No surprise really, just a great pity. I would put Bumpkin ‘Borg in the same catergory as John Brogden – former Liberal NSW Leader. Despite the Nationals colourful selection in Premiers in the past… both ‘Borg and Brodgen came across as decent fellows and has some great potenial for being leaders in their respective states. It’s a pity the NSW Right ousted Brogden for Debnam [who? – definetly wont be voting for him as much as I’m not so in favour with Premier Dilemma] and that Bruce Flegg’s compaign stuffed up the Nationals in Qld. {Some conspiracy theories have been released about whether this was on purpose or not…. it’s possible the libs didnt want a nats premier and delayed any term in govt. until they hold more seats.}

    I think Bumpkin ‘Borg still has it in him to be the next Premier [assuming Bligh doesn’t take over] and would in fact be better than any Liberal Premier as they lack anything to offer. I could see him as in fact being the last Nat Premier for Qld. perhaps maybe a later leadership challenge towards the next election after some family time?

  25. A couple of points:

    tijawi – That “tough system” is little different to Australia’s. For instance, NSW hasn’t lost a seat because of population decline in absolute terms, only population decline relative to the rest of the country. The only difference between us and the US is that the US HoR has a fixed number of seats (435) whereas the Australian HoR can vary slightly in size from term to term (e.g. 148 in 1998, 150 in 2001) because of a different allocation method.

    Contrary to what Darryl and PO seem to be suggesting (unless I’m misinterpreting), the number of senators for the territories has no implications for the size of the HoR. None. The size of the HoR is based of the 72 Senators from the six states. The six states have been allocatied a total of 146 HoR seats, slightly larger than double (2*72=144) due to Tas’s consitutional minimum and other roundoffs. The seats for territories come extra to that, based on the ratio of their population to 1/144 of the six-state population, bringing the total size up to 150. You can see the detail of this in the links I posted above (viewable once that post becomes approved for moderation).

  26. Paul, a glance at the figures makes this more clear. Tasmania’s quota works out at 3.5; so the apple isle accounts for 1.5 of the 2 seat roundoff.

    Or to put it another way, there is more rounding up (Vic, Qld, WA, Tas) than there is rounding down (NSW, SA).

  27. David Walsh: yes, I do recall that WA came extremely close to losing the extra seat it gained in 1998 at the previous redistribution.

    Dave S: As you may now be aware, Lawrence Springborg claimed victory in Bundaberg today immediately before announcing he was stepping aside as leader. Their candidate leads by 441 votes and is clearly out of the woods.

    It doesn’t appear that any significant progress has been made in the count for Chatsworth – still no absentee votes counted. By my reckoning, Caltabiano will need to get nearly 60 per cent of the 2PP absentee vote to win, compared with his 48.7 per cent on polling booth votes and 55.2 per cent on pre-polls and postals.

  28. I think breaking the nexus, even if it was possible, would also cause major problems regarding joint sittings of Parliament. Basically governments could massively increase the number of MHRs without a response for the Senate, thus overwhelming the number of Senators when it comes to joint sittings following double dissolutions.

    Also, double the number of Senators is 152. Does anyone know why we only elect 150 MPs? I know that the constitution only states that they should aim for the HoR to ROUGHLY have double the number of Senators, but why 150 and not 152?

  29. David Walsh: disregard my earlier comment, I did not read your comment in its full context. Antony Green provides an answer to your question here:

    Ben Raue: the final number of seats is an arbitrary consequence of the number of states that get rounded up or down after their populations are divided by the quota. Take the 2003 determination for example:

    If WA had just slightly fewer people, its quota determination of 14.5049 could have fallen below the half-way point without significantly boosting the figure for other states. In that case WA would have had 14 seats and the overall total would have been 148.

  30. i am not a political genius like you guys but could someone tell me that who would benefit more from the final qld and nsw redistributions Labor or looked to me that contrary to malcolm mackeras’s original prediction i think that the coalition would be the clear winner in the nsw redistribution. please feel free to leave any comments you like

  31. G’day all! I’ve got a question:
    Have the boundaries of Berowra been altered at all?
    For those who don’t know, it’s a safe Liberal seat on Sydney’s North Shore, where I live. Phillip Ruddock is the local member.
    The ALP hasn’t got a prayer of ever winning here, so any slight redistribution wouldn’t make much difference.

  32. That’s what I thought. i.e. There’s much less room for fiddling when it comes to the states as opposed to the territories. WA’s 2003 determination could have so easily have been 14.49 instead of 14.50. In which case the state’s # of seats would have gone from 14 to 15 to 14 to 15 in as many redistributions. It would’ve been rather anamolous to see WA lose a seat for what was probably nothing more than a statistical quirk. Especially when this was ostensibly what they were allowing for with the Northern Territory exception.

    And my previous post should answered Ben’s question. Remember, the nexus is based on the 72 Senators of six states; not the 76 Senators in total.

  33. David Walsh,

    My point was just about enlarging the Senate and you’re right. Senators for the territories are not counted to determine quota for Divisions in States (s46 of the Commonwealth Electoral Act). But this isn’t the constitutional nexus which is not based only on State Senators. Section 24 of the Constitution says, in part: “The House of Representatives shall be composed of members directly chosen by the people of the Commonwealth, and the number of such members shall be, as nearly as practicable, twice the number of senators.”

    On the effects of senate enlargement, it’s not obvious that enlargement is in the interests of the ALP, but I’d be interested in hearing why you think it is.


  34. Hi. To Sacha and the suggestion of relaxing the 10% rule. This was a hard won principle that was first introduced as a 20% rule by the Whitlam government because the relatively smaller sizes of rural seats was giving the conservatives an advantage. But it has had the disappointing effect of resulting in excessively frequent and significant boundary changes, meaning not only that “federation” seats get abolished but that the average voter starts losing track of their own electorate.

    There is an alternative – that the size of the seat doesn’t matter but that each member in the Reps has a vote equal to the number of valid votes cast in the last election in their electorate. It can mean less redistributios, more reasonable sizes for rural electorates. But it is a fairly radical idea….but not hard to implement with modern technology (though it makes the concept of pairs difficult and so perhaps that would need to be replaced by a limited proxy system).

    Finally, on the subject of names – which is the only Prime Minister up to and including Holt not to have a Federal electorate named after him?

  35. Evan, three changes have been made to Berowra.

    It has lost 9,982 voters to Bradfield:

    “The committee decided to move the current Berowra-Bradfield boundary west to the Pacific Highway and north to Ku-ring-gai Chase Road … These changes moved electors from the localities of Asquith, Hornsby and Waitara into Bradfield”.

    It has lost 1,782 voters to Bennelong:

    “The committee decided to extend the current boundary west to Silverwater Road thereby transferring electors from Ermington into (Bennelong)”.

    It has gained 12,887 voters from Mitchell:

    “The committee decided to move the current boundary at Pennant Hills Road and Castle Hill Road west to include West Pennant Hills … Additional electors were obtained from Kenthurst and Annangrove by moving the existing boundary to the locality boundaries between Kenthurst, Marayla and Annangrove.”

    The following map shows both the old boundaries (in red) and the new ones (in blue). It actually shows the boundaries as originally proposed before the amendments which have just been made, but these do not affect Berowra or the electorates adjoining it. Warning: it is a big file and will take a while to load.

    Link to map

  36. Daryl, I’m not a constitutional expert, suffice to say the interpretation of it is the most relevant thing.

    An enlargement of the Senate advantages the ALP insofar as it disadvantages the Coalition. Over the last two elections, the Coalition has comfortably managed to win at least 3 out of 6 Senators in every state. Their current Senate majority was achieved by getting a rare 4th Senator up in Queensland in 2004.

    Now I believe that 4th Qld senator to be a fluke. But not the consistency of getting 3 out of 6 Senators up twice in every state. With fully half the numbers in the Senate the Coaltion can frustrate any future Labor govt. (No legislation can pass with 38 no votes.) Plus Labor wouldn’t want to leave in place a system where their opponents, with a bit of luck, can achieve a Senate majority in government.

    Now it’s true that if Labor is winning the House of Reps then the Coalition will have a harder time of gaining half the Senate seats. But it’s still very possible. With three quotas being only 43% after preferences. (Consider for instance 2001, where the Coalition got up three Senators in Tasmania, even though Labor won all five Tasmanian seats.)

    The way to circumvent this? Increase the Senate to 14 seats per state. To gain half those seats, the Coalition has to gain 4 out of 7 seats on 6 out of 12 occasions. 7 out of 12 occasions for a majority. Four quotas out of seven is 50%; a much tougher ask. You could argue that 4 out of 7 is easier than the recent gain of 4 out of 6 in Queensland, which is true. But that’s a feat they’ve achieved just once, not 6 or 7 times over. Plus the smaller quotas give minor parties a better chance to win seats, rather than just being preference distributors.

  37. Kim Howard, the best way to determine who benefits most from the redistribution – and the basis of Malcolm Mackerras’s conclusion, with which I concur – is to compare the sizes of the uniform swings Labor would need to win the election, before and after. Of course, swings are never uniform – they vary significantly from seat to seat, and particularly from state to state – but this is still the best method available.

    Going on the 2004 results, Labor would have needed to have won all the seats with margins of 4.4 per cent or less to gain a one-seat majority. Interestingly, the “median seat” with the margin of 4.4 per cent was John Howard’s electorate of Bennelong – so in the event of a uniform swing, it would have been his seat that decided the result.

    The redistribution has made a number of Coalition seats more marginal, so the uniform swing Labor needs is now 3.3 per cent, and the median seat is now Eden-Monaro. Bennelong is now the next seat along, with Labor needing a swing of 4.0 per cent to win. So the short answer to your question is that the redistribution has given Labor a quite substantial boost of 1.1 per cent. However, this is only fair – the two-party preferred score at the last election was 52.7-47.3, so under a perfectly fair system Labor would only require a swing of 2.7 per cent. This did not stop Liberal Party federal director Brian Loughnane from complaining that the redistribution was unfair, for which he was rightly ridiculed by those in the know.

    BTW, everything I have just said is based on the boundaries originally proposed in July, rather than the amended final outcome that has just been released. However, I do not believe this will change anything I have just said. All post-redistribution figures are based on the calculations of Malcolm Mackerras, which can be viewed here.

  38. Response to Andrew Bartlett from 11.57am:

    I would suggest that the tendency for independents to get shafted is due to the fact that they usually represent country seats – and population trends being what they are, most redistributions move representation from the country to the city. For instance, Ted Mack would never have had anything to complain about if he’d stuck around. No, that doesn’t explain the Pauline Hanson situation – perhaps she was just unlucky. She also erred in running for Blair rather than Oxley (or, better yet, the Senate).

  39. OT: 581 absent votes have now been counted in Chatsworth, and they’re actually running 53-47 against Caltabiano. There should be about 1500 more to come, but it’s almost impossible to see him winning from here.

  40. mackerras’s pendulum in qld was a shocker. he said beattie would win 45 seats but beattie won 60.i don’t know how good is his federal 2007 pendulum.i reckon that the latest redistributions announced in nsw was not as bad as originally predicted for the coalition and in qld it wasn’t that bad anyway.i looked at the newspoll results from 1980 onwards and figured out something. there is a pattern since 1983 that the leader who dominates his opponent as prefered PM ( howard average of 54 to beazley’s 26) never lost an election. even though i don’t like the guy, i must say that i can’t see howard being beaten in 2007 bar a disaster or so.
    rudd i don’t think can hack any better but gillard may have a chance .

Comments are closed.

Comments Page 1 of 4
1 2 4