Labor-Green agreement

Australia’s Labor and Green parties have reached a support agreement. The Greens won their first House of Representatives seat at the recent election. One seat, out of 150, on over 11% of first-preference votes.

One of the provisions of the agreement is that Green Senator Bob Brown will reintroduce as a Private Members Bill the Commonwealth Electoral (Above-the-Line Voting) Amendment Bill 2008. The Labor party “will consider” the bill. Among other provisions, this bill would allow voters who vote for a party ticket in Senate elections, rather than rank their preferences across all candidates running, to rank the parties in order of preference.

The agreement also includes several proposed reforms to parliamentary procedure, including guaranteeing minor parties the right to ask questions of the Prime Minister no later than the sixth question during Question Time. It further stipulates that the parties acknowledge that any of the Green’s policies for the 2010 election can be brought forward for discussion in parliament. Greens will receive Treasury briefings. There will be a “well resourced Climate Change Committee.”

All in all, a very fine agreement. There is just one catch: the Labor and Green parties remain short of a majority in the House by three seats. There are four independents, whose votes could still give the Coalition (of Liberals and Nationals) a majority if they choose to swing that direction.

The Labor and Green parties appear to have combined for over 49% of the first-preference votes, compared to around 44% for the Coalition. Yet Labor and Greens have just under 49% of the seats, despite the use of a “majoritarian” electoral system (and one that is often taken as a model here in the USA), and despite the fact that the electoral swing from Labor to the Greens was greater than that to the Coalition.

(All claims about the partisan breakdown of first-preference votes need to be taken cautiously until all votes are counted, but the pattern of swing is clear.)

19 thoughts on “Labor-Green agreement

  1. The Labor-Green alignment actually falls 2, not 3, short of a majority. People have tended to assume an absolute majority of 76 is required. The constitution is quite clear that an absolute majority (76) is only required for a constitutional amendment. That may become important if Wilkie, the urban independent, abstains as he has threatened to do. 75/1/74 would be enough to sustain a government, although organising the parliament would be hell.

  2. Is there any recent precedent for such an agreement in Australia?

    If Labor and the Greens are committing to reliably vote together until the next election, it will be hard for Abbott to promise the independent MPs any sort of the “stability” they say they want. In a year, when the new Senate sits, there would be no way for the Coalition to push through any legislation, barring a double-dissolution.

    However, if the agreement is reasonably likely to fall apart should Labor not form the government, or should political conditions change, then it may not at all affect the thinking of the independents.

    [T]his bill would allow voters who vote for a party ticket in Senate elections, rather than rank their preferences across all candidates running, to rank the parties in order of preference.

    By my reading, it would in fact replace the single above-the-line-vote with party rankings. It sounds like it would no longer be possible to vote with just a single mark, for better or for worse.

    There are four independents…

    Have there been any indications yet of whether Tony Crook will sit with the other Nationals or alone? I have no idea what his policies are like, is there any chance he would sit with Labor?

  3. [What Alan said, +] … or if a non-Govt MP is elected Speaker, and thus can vote only to break a tie, which won’t arise in a 75-74-1 situation. (Quaere whether the s 57 and s 128 requirements of an absolute majority impliedly override this rule? Or could we still have a Marquet Loop* at the federal level?)

    * Arose under West Aust State Const. Legislative Council members favoured an electoral reform bill 18-16. However, the President could not vote as there was no “tie”, so the votes were 17-16. Which meant the Bill fell short of an absolute majority as required, because it would have amended the State’s Constitution Act. So it failed.

    Mind you, WA is also the State that abandoned STV and reverted to MNTV for local council elections because the State’s shire clerks complained that preferential voting just too hard. So we’re not talking a laboratory of democracy here.

  4. @Tom
    s57 and s128 apply to very specific situations and are therefore exceptions to a general rule. I suggest the relevant provision is s40:

    Questions arising in the House of Representatives shall be determined by a majority of votes other than that of the Speaker. The Speaker shall not vote unless the numbers are equal; and then he shall have a casting vote.

    A number of states have had similar arrangements. Until recently the South Australian Labor government included a Nationals MP as a cabinet minister. The Tasmanian government is a Labor-Green coalition with Green ministers. Both agreements give the junior partner had the right to opt out of coalition discipline for specific issues. The South Australian agreement lasted 2 parliamentary terms. The Tasmanian agreement is only about 6 months old.

  5. Alan, the Senate ruled early in its history (this is mentioned somewhere in Quick & Garran) that sec 17’s “the Senate shall… choose a Senator to be the President of the Senate” impliedly overrode Sec 23’s “Questions arising in the Senate shall be determined by a majority of votes, and each Senator shall have one vote. The President shall in all cases be entitled to a vote; and when the votes are equal the question shall pass in the negative”. Which therefore allowed the Senate to draw lots to break a tie in a ballot for President.

  6. The Speaker does not vote in the House except in the event of a tied vote.
    I’ve always found this to be a strange rule:
    – when a party has one seat majority and wins the speakership, every following vote is a tie and depends on the speaker
    – when both parties have half the votes, the one winning the speakership (whatever the tie-braking rule for the elections of the speaker), loses the following votes by one vote.

    (Sweden had this situation in 1973, government and speaker stayed in power, tied votes were decided by lottery and in 1976 the seat toal became uneven)

  7. It’s no defence of a silly rule, but Australia did not have a party system when the constitution was written. In my view the rule in both houses should be that a motion needs an absolute majority to pass.

    In other news, Wilkie has gone with Labor so the numbers are now 74/3/73 or 74/4/72 depending on the somewhat erratic WA National.

  8. From Wilkie to Franklin… No, not the 1940 US presidential election, but the Tasmanian House of Assembly. Looks like the 1998 Cutback Amendment is now going to be rescinded, ie, the Apple Isle will be returning to seven-seaters (thus overtaking the ACT as Australia’s most optimal district magnitude).*

    Maybe someone looked at the federal result and twigged that small district magnitudes are no protection against being “held to ransom” by “fringe zealots”…

    * 6.999 (rounded off) is the optimum. 21, as for the NSW upper house, is too large.

  9. In the following situation:

    1) a government determining House has an even number of deputies,

    2) the government party or parties have exact exactly half the number of deputies, 50%, not 50% +1.

    3) the Speaker doesn’t vote except in case of a tie, and the government somehow provides for a speaker.

    In this case, a government simply needs 50% plus one for a majority, not 50%. The government either must convince an opposition deputy to agree to become speaker, which is functionally not too different from convincing an opposition MP to defect to the government, or must convince another party to join the governing coalition.

    In other words, I don’t see the difficulty with a legislature having an even number of members. Even number or odd numbers, you still need 50% +1 for control, 50% isn’t enough.

  10. My reading of Australian poltics is probably flawed, since I am not Australian, plus the situation now is unusually fluid.

    That said is the agreement really “very fine”? How realistic would it have been for the Greens to first reach an agreement with the independents/ National defectors? A sort of Green – country bloc, if it could have been managed, would have improved the bargaining power of both groups vs. the major parties. It also would have started the emergence of a rural-based party with a clear identity apart from the Liberals, plus improved the Greens’ image with non-urban voters.

  11. Actually, the sudden rash of uncontrolled (or hung, or balanced) parliaments elected single member districts cuts against both sides of the proportional representation debate.

    We have a clear demonstration that using single member districts are no guarantee of a majority for one party, and as another example in South Africa one party has had a majority for several consecutive elections using proportional representation.

    But on the other hand, single member districts can’t be said to shut out minor and third parties or keep them from being influential.

    At best, you can say that the use of single member districts, as opposed to multimember districts with proportional representation, will increase the odds of one party obtaining a majority, and reduce the number of third and minor party deputies. There are no guarantees.

    The argument shifts more to the ground of proportionality vs. local representation. And why the German or the Irish systems are not satisfactory compromises.

  12. [Note for the record that I spotted the India/ UK/ Canada/ Aust/ NZ hung parliaments before reading Dunleavy’s post a few days back… ie, did not plagiarise him without attribution.]

    One could say “Well, of course NZ has hung parliaments – they voted for PR in 1992-93”. Yet (a) parties can win absolute majorities under MMP (Germany 1957), and (b) at (at least) the first MMP election, no party won a majority of the district seats either – ie, presumably FPTP, too, would have produced a hung Huss of Rips in 1996. (It probably wouldn’t have had only 65 districts – although NZ had only 75 in the 1970s – but with 120 or so seats all from SMDs, it would have been even easier for small parties and Undupundints to win seats.)

    Enid Lakeman used to delight in contrasting Austria, Tasmania and Malta with Canada and France (although many British anti-PRists of the Spectator type seem to think France is a multi-party warning of the evils of PR).

    However, since 1982 (when she last published), Ireland and Austria have seen “hung parliaments” become the norm, and Tasmania has seen them become a possibility (Malta remains immune, it seems, while the loi Duverger has not yet kicked in in Canada, other than impelling the Prog Cons and Reform to merge).

    It does seem that absolute majorities (genuine or manufactured) are rarer under PR-List than under PR-STV, both because of –

    (a) the “carrot” that STV offers (different factions can stand a common team of candidates, and if their supporters really, really can’t stand their political bedmates, they can direct their preferences elsewhere, or let them exhaust. Unlike, say, Israel, they don’t need to break away and run their own list to make sure their votes don’t elect someone from an opposing faction)


    (b) the “stick” that STV offers – relatively high thresholds, based on small-magnitude regional/ local districts with no nationwide or large-region top-up. IOW, STV quotas are generally much higher than PR-List quotas, even without a superimposed threshold.

    Apart from the NSW Leg Council, an outlier with DM = 21, and a [now rare] Australian Senate double dissolution, with DM = 12, the maximum DM I know of under STV is 9 for Cambridge, Massachusetts and formerly the ACT Advisory Council. Otherwise, it’s 5, 6, or 7 apart from the occasional 3- or 4- in Ireland.

  13. There’s a good write-up on the South Australian supply and confidence agreements here. The principle that a minister can opt out of cabinet solidarity is also followed in the Tasmanian Labor-Green coalition.

  14. A few town councils in Ireland have 12 members, all elected at large, and the rest have 9. City and County Councils are divided into wards, but even there I think the magnitude is 5 or 6 per ward.

  15. I’d overlooked local councils. If we count them, then many in NSW (and possibly Victoria and SA too?) have as many as 12 seats at large, although most use 3-, 4- or 5-seat wards.

  16. Tom,

    I believe that the only council in Victoria with 12 councillors is the City of Geelong. All others have between five and nine. Some councils are elected from single-member wards, some from multi-member wards, some from combinations of both and some at large. It is a real mess. When the Shire of Nillumbik had its electoral representation review, my submission argued for one ward with STV, but the commissioners went for seven single-member wards on ‘community of interest” grounds.

    Here is my submission:

    There is no rational argument for single-councillor wards in any council in the state. Single-councillor wards do not so much allow as compel the representation of ‘geographic communities of interest’, while preventing the adequate representation of any other community of interest. Unsubdivided councils allow, but do not compel, the representation of ‘geographic communities of interest’ and of any other community of interest that sufficient voters identify with.

    I wish to support an unsubdivided council, to be elected by proportional representation. Having an unsubdivided council is more efficient because it removes the need for electoral redistributions as populations change in different parts of the municipality.

    Nillumbik has had wild swings between pro-conservation and pro-development majorities on its council when the division in the community is not so extreme. This is a result of single-councillor wards, which are intrinsically unrepresentative. The use of single-member electorates for Lower Houses of Parliament in most of Australia’s political jurisdictions assists in stability of government in our two-party system. Whatever the validity of this argument in that context, it is irrelevant to local councils, which are not dependent on one group having a permanent majority for their successful operations.

    I advised the Council at the last review not to adopt single-councillor wards as it could result in an unrepresentative council, which it did: the pro-conservation majority was chucked out and replaced with an overwhelming pro-development majority, which was in its turn chucked out and replaced with an overwhelming pro-conservation majority at the following election.

    It is not the continual changes of structure that are the problem in Nillumbik. It is the inability to get the structure right no matter how often it is changed.

    Proportional representation is a much fairer electoral system because it allows minority representation and ensures that larger groups are represented in proportion to their support, unlike single-member systems, which shut out minorities and distort the representation of larger groups. It also allows the voters to group themselves instead of placing them in geographic groups that they do not see as important to their voting decisions. It is true that the quota for election with a seven-councillor structure would be 12.5 percent of the vote, but that is an advantage: the election would be much more democratic.

    The argument for the preliminary preferred option is not convincing. The report states:

    ‘In supporting a single-councillor ward structure, a number of submitters argued that the Shire has a numerous “place-based” communities of interest which are better served by local councillors representing a specific constituency. Submitters contended that councillors in single-council wards are more involved, knowledgeable, identifiable, responsible and accountable to local communities. It was mentioned that there is little interdependence between the various townships, which each have their own heritage and identity.’

    ‘It was also stated that as there is little cultural or ethnic diversity across the Shire, and no obvious minority groups, single-councillor wards representing geographic communities of interest are appropriate for Nillumbik.’

    ‘The VEC recognises there are distinct communities of interest within the municipality, which were detailed in a number of submissions. The many townships appear to be quite independent, with their own sporting clubs….’

    In an unsubdivided council, if ‘geographic communities of interest’ are the most important factor in electors’ voting choices, then such electors will vote for candidates who they think best represent their geographic communities of interest. If the electors do not in fact particularly see their geographic community of interest as the most important factor in their voting decisions, an unsubdivided council will not force them to vote on that basis. An unsubdivided council allows the voters to create their own communities of interest by their voting decisions: if getting a Diamond Creek representative is really the most important thing for the voters of Diamond Creek, then an unsubdivided council will allow them to elect one. Equally, if the voters of Diamond Creek have little interest in where their councillor comes from but have a passionate interest in protecting the Green Wedge or in dismantling it, they can group themselves with other likeminded voters from all over the Shire and, if just over one in eight do so, they will elect such a representative. The key feature of an unsubdivided council elected by PR is that the voters group themselves; they are not grouped by where they live unless they want to be.

    If the voters in an unsubdivided council think that councillors from their ‘geographic community of interest’ would be ‘more involved, knowledgeable, identifiable, responsible and accountable to local communities’, they will elect them. If not, they will not be artificially grouped and forced to so, but will be able to elect someone else.

    The Nillumbik Guide for Submissions says:

    ‘Communities of interest are groups of people who share a range of common concerns or aspirations….The communities of interest to which people belong are often also an important part of self-identity.

    ‘Communities of interest may occur where people are linked with each other geographically (e.g. a town or a valley) or economically, such as where people work in mutually-dependant(sic) industries (e.g. fruit growers, transporters and canners). Communities of interest may also appear where people share a number of special needs (such as new immigrants, who may have little English, require assistance with housing and need help finding employment. Communities of interest may also include ethnic groups, retired people, the unemployed or many other groupings of people.

    ‘Communities of interest are important in electoral representation reviews when they have similar needs from their local government. In such cases, it is important to endeavour to ensure that communities of interest have the opportunity to be fairly represented on councils.’

    An unsubdivided council allows the communities of interest that the voters themselves identify with to be part of their voting decisions. A single-councillor ward structure forces voters to see their geographic community of interest as the only one that counts.

    The statement in the Guide that an unsubdivided council ‘May lead to significant communities of interest …being unrepresented’ will only be true if the voters in that municipality do not see those communities of interest as important when they vote.

    The statement that an unsubdivided council ‘May lead to significant…points of view being unrepresented’ is illogical because any significant point of view supported by 12.5 per cent of the voters will elect a councillor, whereas in a single-councillor ward structure, a point of view with the support of 49.9 per cent of the voters can still be excluded from the council.

    The other less positive features for an unsubdivided council have some weight, but not as much as the basic principle of representative democracy, which is to allow the voters to choose a body which best represents them in all their diverse opinions, which is what an unsubdivided council elected by PR does.

    The report also says:

    ‘It was asserted that political party involvement is more likely in multi-councillor wards, and that running for election in larger wards is more time-consuming and expensive, thus limiting the potential pool of applicants.’

    The one way that political parties can get into local government is if people vote for them: if they do so, it is thus a democratic result. If voters do not want political parties, they are free to vote for independent candidates – and if under this proposal no independent can get even 12.5 percent of the vote after preferences, it would be obvious that few voters really wanted an independent after all.

    The report also says:

    ‘It was also claimed that voters in larger wards would only have a superficial knowledge of candidates, and that councillors in broader wards would have difficulty in representing diverse and scattered communities. Some submitters wrote that urban areas could dominate in an unsubdivided structure to the detriment of the Shire’s rural areas. Additionally, it was suggest that councillors’ workload in multi-councillor wards would actually increase, as councillors are still expected to attend all meetings and have to travel further distances in the course of their duties.’

    It is also argued that multi-member wards reduce local representation. This is also false, for any voters who think that the most important thing is to elect someone who represents their locality can still vote for the local person. If not even 12.5 per cent choose to do so, they have obviously thought other issues more important, and again that is their democratic right. Voters may not wish to vote on the basis of whether they live in an urban or a rural area, but if they do vote on that basis, an unsubdivided council will allow them to elect an urban or rural person for every 12.5 per cent of the vote.

    An unsubdivided council would have the additional benefits in Nillumbik of improving continuity on the council and preventing undemocratic inflated majorities as it would prevent wild swings between the two sides of the ongoing Green Wedge debate.

    The preferred option contradicts the reasoning about ‘geographic communities of interest’ as it includes several townships in each of Bunji, Sugarloaf and Yellow Gum wards, while it divides St Andrews between Bunji and Sugarloaf Wards, Diamond Creek between Wetlands and Edendale Wards and Eltham between Sugarloaf, Butterfly and Wingrove Wards. It is not possible to establish wards based on townships without a much larger number of councillors. Nor is it legally possible to have the required number and meet the requirement for only ten per cent tolerance between ward voter numbers. The attempt to argue for single-councillor wards on the basis of the geographic interests of townships falls down when the number of townships far exceeds the legislative limit on the number of councillors.

    Paradoxically, the three-ward multi-councillor ward structure does a better job of joining geographic communities of interest than does the single-councillor ward structure.

    The report also says:

    ‘…it was stated that in singe-councillor wards responsibility cannot be “buck-passed” to other councillors.

    If a councillor in an unsubdivided council passes the buck, such a councillor can be voted out of office in the same way that a councillor from a single-councillor ward can.

    The report also says:

    ‘Many submissions…argued that a recommendation for single-councillor wards would be consistent with the VEC’s reasoning and recommendations in these two reviews, and that as all other metropolitan/rural fringe municipalities are subdivided, it would be consistent for the Shire of Nillumbik to be likewise divided into wards.

    My argument is that no council should be subdivided because proportional representation is the most representative voting system in the world. The fact that some like councils have been left with single-councillor wards is unfortunate for them. It is not a reason for Nillumbik to be stuck behind in the long process of advancing democracy.

    The report also says:

    ‘While the VEC’s preferred preliminary option is a single-councillor ward structure, it has also provided an unsubdivided structure for public comment. Such a structure may have the advantage of providing a Council which is more reflective of voter sentiment and significant minority groups, rather than allowing one group or another to dominate.’

    ‘An advantage of an unsubdivided municipality is that the structure can accommodate non-geographic communities of interest and interests that are broader than a single-councillor ward. This is because constituents from across the Shire, not just those who lie in a particular, can choose to vote for any candidate who they feel will effectively represent them. Additionally, candidates representing issues that affect substantial minorities have a better chance of election in an unsubdivided structure with the application of proportional representation.’

    It is an automatic consequence of an unsubdivided council that it will be ‘more reflective of voter sentiment and significant minority groups, rather than allowing one group or another to dominate’. That advantage is inherent in the proportional representation system of voting.

    An unsubdivided council elected by proportional representation gives the widest choice to the voters to decide on the issues on which they will vote. It lets them form their own groups. They can vote in accordance with the ‘geographic communities of interest’ if they wish, and if they do not do so, it cannot be that important after all. They can vote for non-party candidates if they wish, and if they do not do so, it cannot be that important after all. They can vote in accordance with the rural/urban divide if they wish, and if they do not do so, it cannot be that important after all. In essence, they can vote on the issues that they see as the most important and be confident, if one in eight voters see things the same way, they will have a councillor who will speak for them – something no small group can be confident of in any single-councillor ward structure. They can also be confident that the majority on council will be the majority that the voters have chosen, something that they cannot be confident of in any single-councillor ward structure.

    I wish to speak at the public meeting on 30 April.

  17. In NSW a council must have between 5 and 15 members and may be a single electorate or use a ward system.

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.