The built-in coup


As yesterday’s abortive coup played out, Nick Bryant from the BBC penned a fascinating piece laying out the differences in political cultures between Australia and the United Kingdom when it comes to leadership coups. Australia’s culture of sudden, brutal and frequent internal leadership changes is not at all seen in British politics.

Bryant, however, missed the primary reason why Australian politics has so many leadership coups compared to countries that appear to have similar political cultures. Australia is the only Western English-speaking democracy where the choice of political leaders is made solely by that party’s members of parliament.

There is nothing in Australia’s electoral system that requires political parties to choose their leadership through a vote of their caucus or party room. British and Canadian politics both also use the Westminster system, typically with a single party forming a government in the lower house. Yet no significant political party in either country gives sole discretion to their parliamentarians to choose the party’s leaders.

The British Labour Party gives only one-third of the vote in its ballots to its MPs (with the other two thirds split between grassroots members of the party and members of affiliated unions), while the Conservative Party’s MPs winnow the field of contenders down to the top two, who then face a ballot of the party’s membership. The Liberal Democrats and the various Canadian parties use other methods – either ballots of every member or a vote of delegates at a party convention.

Of course, the United States likewise involves large numbers of voters in choosing their presidential candidates. It is a different political system and is usually not discussed when talking about leadership change, but it is much closer to the UK or Canadian model than Australia.

There are numerous advantages to broadening the electoral base for leadership elections. It gives a chance for the party to see leadership candidates in action, and it actually gives a chance for a broader group of people to have a say (instead of MPs having to read polls or make their best guesses about the voters’ preferences). It also revitalises the grassroots membership of a party, giving them a real reason to participate, and some real power.

When you hold a ballot of the membership, it’s simply not possible to have so many leadership challenges, and they couldn’t be described as ‘coups’. It ends the ridiculous game-playing we saw yesterday: where potential candidates can choose to hold off and not nominate if they aren’t happy. A proper leadership election doesn’t happen every day – and it forces anyone seriously considering a tilt to throw their hat in the ring.

It wouldn’t completely eliminate leadership changes – but it would bring some stability. You would be very unlikely to see scenarios where a number of leadership elections take place in quick succession (as seen with the federal Liberal party during the Rudd government).

These elections are far more decisive than a vote of a small number of MPs who can switch sides a few months later. If Gillard and Rudd had faced off in a ballot of party members it would have decided the issue once and for all. Indeed, Gillard may not have challenged Rudd prior to the 2010 election if it required her to face party members and make her arguments. That may have made them think more about the negative impacts of deposing a sitting Prime Minister in his first term.

Liked it? Take a second to support the Tally Room on Patreon!


  1. The counter-argument is that a party’s MPs tend to be more in tune with the electorate (more “moderate” or “centrist”) than the party’s membership, which tends more to the extreme. Certainly MPs feel the hot breath of the wider electorate down their necks more keenly than party members. This means that a ballot of the party membership is more likely to elect a leader who’ll take strong policy positions that enthuse the party’s base, but lose the support of the wider community. I believe we saw this play out in the demise of the Australian Democrats.

    The UK Labour party seems to acknowledge this – one third of the votes going to MPs certainly gives them outsize influence in the result, even if they don’t completely control it as is typically the case here.

  2. It’s worth noting that New Zealand Labour adopted direct election at its Congress last year. It changed to a variant of the British Labour model with 40/40/20. Australia is now the only English-speaking Westminster country with no major party having direct election of the party leader.

    There a clause in the national party rules that stops the ALP adopting direct election which I outline here. I’m not sure when it was introduced but it is likely that it occurred after Jack Lang as he was the last (and to my knowledge the only) Labor leader elected by the party, not the caucus.

    I have a feeling that the rule will be removed at the next National Conference to allow state branches to adopt it if they wish to. If it is removed, Tasmanian ALP is likely to be the first to adopt a direct election model.

  3. @kme, there are some values in having MPs participate, clearly. I didn’t say they shouldn’t have any say.

    I quite like the idea of them getting a proportion of the vote, but not 100%. It works quite well for NZ Labour and UK Labour to split three ways – the Greens or another party without affiliated unions could give 60% or 66% to the members and the rest to the MPs.

    It still ensures that cooler heads prevail, time is taken and arguments need to be made out in the open rather than behind closed doors.

    I would also argue that, if you actually gave members more of a say in deciding their party’s leadership (and thus the direction) you would get more people and a broader cross-section of the community joining.

    This mainly applies to the ALP and the Liberals. I don’t actually think the membership of the Greens is particularly more left-wing than the MPs. Maybe this is partly because all Senators (except in Tasmania) are already preselected by a vote of the membership, which when you pair it with the party’s more participatory decision-making does actually give members more ability to make a meaningful contribution.

  4. For the last British Labour leadership vote:

    – The vote of one MP was worth the votes of nearly 608 party members and 12,915 affiliated members.
    – The vote of one party member was worth the votes of 21 affiliated members.
    – An MP’s vote was worth 0.12 per cent of the total electorate, a party member’s vote was worth 0.0002 per cent and an affiliated member’s vote is worth 0.00000943 per cent.

    Ed Miliband did not win the MP or the rank-and-file electoral college but won the affiliates electoral college by a large enough margin that allowed him to beat his brother.

  5. One thing you missed was looking at the impact of union endorsement of candidates on the affiliates vote.

    In all but one case, the candidate endorsed by a union received the most votes from that union’s members.

    Personally I think it shows that it is win the interests of affiliated unions to support party reform to directly elect the party leader.

  6. The NSW case saw rank & file devotion to Lang retain him as leader when he was electoral liability. Interesting that when unions overthrew Lang machine they restored election of leader to caucus= a voluntary surrender because they thought caucus control more likely to lead to electable party.

  7. Something that deserves to be remembered is that the Australian Democrats did have a direct election of the parliamentary leader by the membership (with no say by the parliamentary wing, as I understand).

    This didn’t stop them from having a revolving door of leaders between 1998 and 2008: Kernot, Lees, Stott-Despoja, Greg (acting), Bartlett, Allison. That said, only Meg Lees was deposed in what you could classify as a “coup”, with all other cases being ones where the leader voluntarily stepped down.

  8. The Australian Democrats had a system where the rank and file members could spill the Parliamentary Leadership, with a certain number of signatures, and chose a new one. They got the one they wanted and the one the public wanted, but unfortunately she had very little support amongst the MPs she was required to ‘lead’. Bye bye Democrats.

  9. Greg, that’s a misleading summary of the Democrats situation.

    The Democrats are the exception to the rule and are the one example that self-interested politicians pull out every time to justify keeping this power to themselves.

    The Australian Democrats were around for over twenty years and, as Bambul said, almost every leadership change was achieved amicably.

    I can list dozens (literally) of examples of parties being destabilised by leaders not being supported by their MPs despite winning a party room vote. None of those problems have been experienced by political parties in the United Kingdom and Canada, which have had much more widespread and longstanding experience of party leaders being elected by the membership.

    If you remember correctly, Greg, Natasha Stott Despoja had the support of three of the party’s Senators and four were opposed to her towards the end of her leadership. I don’t know who Vicki Bourne (the Senator who lost in 2001) would have voted, but let’s just say she would’ve supported Meg Lees. That meant Lees had the support of six MPs and Stott Despoja had the support of three. Do you seriously believe that a party that was so deeply divided wouldn’t have suffered from destabilisation anyway?

    And after Lees left, the party was split down the middle, 4-3. Such a situation would be deeply destabilising for any party, regardless of how the leader is elected.

    It also ignores the many other reasons the Democrats came to an end, including the lack of a committed base, the defection of Cheryl Kernot, the decision to support the GST (a decision supported by a majority of MPs but not by the membership), the rise of the Greens and the shift of the ALP to the right.


    There is no evidence to suggest that the large number of leaders the Democrats experienced in their last decade had anything to do with how they were elected.

    Kernot left because of her defection to the ALP. Greig was never actually elected as leader. Bartlett and Allison both lasted a full term, which was reasonably impressive considering the complete collapse in the party at that point.

    So really the only real leadership instability was around Stott Despoja and Lees. Considering the many, many, many examples of leadership instability in major parties in Australia in the last few years, that hardly is evidence of direct election leading to instability.

  10. It is hard to see Gillard’s new cabinet will be any good simply because of the numbers. If we take into account for arguments sake that the labor caucus contains about as much talent as the Coalition caucus does and 45 Labor caucus members are ineligible for cabinet level positions because of supporting Rudd you basically have to give whoever is left a position. I am not saying that all back-backbenchers are incompetent but you better hope that they all are talented because Gillard is running out of choices. Every one of her supporters, regardless of how smart of dumb is going to be in cabinet, against Tonny’s shadow cabinet which is drawn from a slightly bigger caucus, if you include the Senate, and which has had 3 years to familiarize themselves with there respective portfolios. To say the least it does not look good.

  11. You led your response with an attack on my motivations. Not really a lot of room for disagreeing with Ben on Ben’s blog I see.

    The more interesting fact would be how many of those Democrats’ leadership changes happened as a result of the membership calling the spill. Ive got no idea. The only one I know of was the Lees>Stott-Despoja one. Some might argue that the underpinning of the Democrats’ vote, certainly in their later years, was based largely on their leader’s personality. Plot their highly volatile rise and fall against the changing leaders and you’ll see what I mean. The point is, when the Democrats true existential crisis moment arrived, their mechanism threw up a leader that the majority of the members wanted and the majority of the MPs wouldn’t follow. Did the members know of care that that was the case? Coz it sealed the party’s doom. Sheit, Natasha’s media adviser even wrote a book about, which I thought explained it all rather well.

    And in those overseas party examples people refer to, in some of those cases the Leader they elect is the leader of the whole party, that is the membership arm and the parliamentary arm. NZ Greens being the closest example.

    You are not the only one to have ever thought about this Ben.

  12. Sorry if you didn’t like what I said Greg, I didn’t mean to single you out, but in my experience the Democrats example is used very selectively while ignoring the many more examples of direct election working perfectly well for parties, and ignoring all of the other problems the Democrats had. However I was probably too harsh in my comments.

    I think distinguishing between someone being leader of the “parliamentary party” versus the leader of the party is a matter of semantics – the public doesn’t know the difference. The very definition of a party’s leader is the person who leads them in Parliament and in terms of elections – there are always other people in charge of the administrative wing of the party.

    I don’t think that calling a leader “Parliamentary leader” gives them an exception from their responsibilities as a whole. While they may not have specific authority to direct the administrative wing of the party (as is the case with yourself and other Greens leaders around Australia), calling someone “leader” (along with the media attention and sometimes parliamentary resources that accompany the title) give them a huge amount of moral authority and sway over the direction of the party. I think if someone is given that authority they should be given it by their membership, not just by MPs.

    This can include giving the MPs greater say than the average member (say 50% of the vote goes to MPs), but I don’t think they should have sole authority.

    I still fundamentally disagree with the concept that the election of Stott Despoja had anything to do with the Democrats collapsing. The deep internal divisions within their Senate party room (which were expressed in fights over leadership, but still would have been if the leadership had been chosen another way) were combined with structural flaws and political problems.

  13. The title, and recognition by the parliament, doesn’t lead to me getting any extra parliamentary resources, so you can relax about that part!

    What about my example from Kiwiland? Two pairs of party (not just parliamentary) co-leaders since ~1996, one of those four people, my mate Rod, died in office. Replaced by a guy who wasnt in parliament, but later got put on the party list? Pretty successful model by your various criteria?

  14. Well said Ben. It is an interesting observation regarding the manner of election of party/parliamentary leaders in Australia, and equally shines a light on the continued dominance of parliamentary parties over the lay party. Now, this doesn’t mean that the lay party would necessarily do any better (the endorsement of a variety of Republicans for the Senators at the last US election demonstrates this!), but equally the assumption is that the party is ONLY about winning elections. I’ve heard it mentioned that the only reason the Greens did so badly in WA was that they didn’t have a “Leader” – which is frankly idiotic and entirely ignores the nature of the local campaigns and the political context of the general campaign. The party’s supporters are the core of what the party is and stands for (and here I am broadening out from just the card-carrying members), so why shouldn’t they have a say in who their spokesperson/leader is?

  15. Greg, I think the issue is more about where people can come into that leadership and who does the electing. A Co-Leader model (which is now used by a number of European Green parties) is different from the standard Australian model of “Leader”. I wouldn’t have a problem with a co-Leader arrangement, voted on by the party, which could include any/all of the MPs. Heck, if we think the MPs are worthy of being our representatives then we should probably be also voting for them as party leaders. But I do have some issues with the notion that ONLY an MP can be a party leader – we might consider Nat Bennett in the UK, for instance, who is not an MP yet is Leader of the Green Party of England & Wales, voted on by the party.

  16. Greg, I know you don’t get any extra resources, but others do. I don’t think there’s anything wrong with it, but I’m just saying that it’s silly to pretend that a “parliamentary leader” doesn’t have a significant role in the management of the party outside Parliament. I don’t think there is any practical meaning in defining someone as leader of only the parliamentary party.

    I think the NZ Greens are quite good. I’m not sure about precisely how they make their decision, but I believe the decision was made by the party’s conference. When both Metiria Turei and Russell Norman were elected, there was a period of months where the party’s membership and MPs had time to consider the candidates and make a decision about where they stood.

    I also like having co-leaders, although I recognise that it would be silly to have co-leaders with a small number of MPs (any less than 5). It also allows you to take the time to make a considered decision, and to sometimes have a leader who is not an MP in the short term (although I’m less enthusiastic about that than Stewart).

  17. As others have mentioned, Australian Labor is now one of the few parties of its type in the world that hasn’t experimented with some kind of leadership election or primary. Why they refuse to do so is just baffling – because the system can work to engage people, and it can produce winners. This is the 21st century, after all.

    British Labour has an equal electoral college between, members, MPs, and affiliated societies (unions). After the last general election, the party threw open its doors and allowed any new member who joined the party a vote in their leadership election. the result? 50,000 new members of the British Labour party in 6 months.

    New Zealand Labor have wisely adopted a similar system.

    In Canada, the New Democratic Party held a leadership election of all party members after the death of Jack Layton.

    The French socialist party recently experimented with a primary of all Socialist Party supporters for their presidential candidate. Supporters paid One Euro to participate in the election, and signed a pledge expressing their support for the values of the socialist party. . Over 2 million people participated over two weeks. The Socialists didn’t just have their coffers filled, but the process produced Francois Hollande, their first successful socialist party candidate since 1988.

    Even the British conservatives have a moderate form of membership chosen leader, where MPs narrow it down to the top two candidates, who are then put to all party members. This system produced David Cameron, who went on to be Prime Minister.

    Australian Labor simply has to move away from this idea that leaders are chosen exclusively by MPs, who are in turn beholden to by union officials and factional leaders. It’s legitimate to expect that MPs would have some say in who leads, them, because they do have to work with them. But the idea that they are always the most knowledgeable people in who is or isn’t a good leader is absurd.

    In practise, many MPs are factionally appointed sheep who look to others to tell them what to do. I know for a fact that one union secretary was calling “his people” in parliament last week to vote for their candidate of choice. It’s a 19th century system that continues to exist in the 21st century, not for any good reason, but for plenty of bad ones.

    Here’s a stat to illustrate the point: Between 1935 and 2001, the shortest time that any Labor leader served (excluding Frank Forde’s 3 weeks) was Paul Keating’s tenure, at 4 years and 3 months. After him, Beazley served for 5 years 9 months as opposition leader.

    Before Keating: Hawke (nearly 9 years), Hayden (5 years), Whitlam (11 years), Calwell (7 years). Evatt (9 years), Chifley (6 years), Forde (a few weeks), Curtin (10 years).

    Since 2001?

    Crean (2 years), Latham (1 year 2 months), Beazley (1 year 11 months), Rudd (3 years 6 months) Gillard (2 years 9 months).

Comments are closed.