Senate reform – questions that need answering


Two weeks ago I wrote a piece for the Guardian putting the case for reform to the system of Senate group voting tickets, in response to a number of pieces arguing against reform on the basis of who is for the reform, or who supposedly will be favoured.

Today I want to write about a few points I didn’t get to cover in my post, and respond to stories about politicians (primarily in Labor and the Greens) expressing hesitation about supporting the reform seemingly due to concerns about the partisan impact on their parties, and the proposal of a (very bad) idea to keep group voting tickets but impose a minimum vote threshold of 4% for any candidate to be elected.

Antony Green has previously conducted an analysis of how the proposed system of above-the-line preferencing and optional preferential voting would likely change the make-up of the Senate, assuming no major change in voter behaviour. His predictions would have seen Labor and Coalition each hold two additional seats in the current Senate. The Greens would hold one less seat, with a number of other crossbenchers missing out. Nick Xenophon would have likely been joined by his #2 candidate under the proposed system.

It seems some people in Labor and the Greens seem to think that their respective party could be disadvantaged by these reforms. The problem with making predictions about how the system would work (as acknowledged by Antony Green) is that a change in system will change the behaviour of parties, candidates and voters. The changed system would likely result in fewer parties running, and over time this could see different voting blocks developing. A smaller ballot paper would likely have hurt the Liberal Democratic Party, who won in NSW on the back of substantial confusion over their name.

While it looks like the Coalition would have been the chief beneficiary of no group voting tickets in 2013, at other times the reforms would have more likely benefited Labor. This has been explained by Kevin Bonham.

Others have opposed these reforms because they have been pleased with the impact of Ricky Muir and others on the Senate crossbench. Like with any lottery, it’s possible for it to produce a positive result, but that doesn’t mean it is fair or that it can be predicted to produce another result you like in the future.

Ultimately, you need to make your judgement about Senate reform on its impact on the system as a whole, and whether it is fair, not whether it advantages your side, or politicians that you like. That may seem naive, but in the medium term it’s the only sensible approach to electoral reform.

We have seen the impact of short-term thinking in New South Wales and Queensland, where past Labor governments introduced optional preferential voting, in part to benefit from division between Liberal and National candidates (and Queensland Labor went further in benefiting from vote-splitting between the Coalition and One Nation in 1998, encouraging voters to ‘Just Vote 1’).

Since then, the Liberal and National parties have merged in Queensland and gotten better at avoiding three-cornered-contests in other states, and One Nation has disappeared, while the Greens now take a much larger part of the left vote, and those same ‘Just Vote 1’ messages first used by Peter Beattie were used by Campbell Newman for the opposite effect earlier this year.

In addition to us not being able to make any certain or long-term predictions about the partisan impact of a reform, there are deeper issues with the current Senate system, beyond which candidates might win.

The group voting ticket system, with full preferences and almost-complete control of those preferences by parties, tends to produce results where the exact order of elimination of candidates is critical to the result, and small changes in the order can produce dramatically different results. We saw this in Western Australia in 2013, when a gap of 14 votes (or was it 12 votes?) changing who won two Senate seats.

This problem hasn’t gone away, and if we continue down the current trend of increasing numbers of parties and candidates, it will continue to get worse. We may well see another very close contest at the next election, resulting in another recount and more pressure on the AEC to go above and beyond its normal responsibilities.

So this isn’t just about who wins and who loses, this is about having a democratic process which is understood, respected, and not seen as producing arbitrary results which have little to no relationship to the votes cast.

Finally, a number of media outlets have reported that some in the Greens including new leader Richard Di Natale are open to an alternative reform, which would keep the current system but exclude from election any candidate from a party that polled less than 4%.

This is a terrible idea, and would be worthy of the criticism that minor parties have directed at the reform process so far. It would be a bald-faced case of the major parties using their position to entrench their power at the expense of small parties, and would not fix the problem.

Preferences are not only abused by small parties – they can be abused by larger parties. In 1984, the major parties effectively used the new group voting ticket system to starve Peter Garrett (then of the Nuclear Disarmament Party) of preferences, and have often directed preferences away from the Greens.

A threshold system would be particularly bad for the Greens, who would still be vulnerable to losing Labor preferences.

Using thresholds would also completely fail to deal with the lack of transparency in the current system, and allow backroom deals to continue to be made, with results that are very hard to predict.

While the above-the-line voting system is unlikely to elect anyone with less than 4% of the vote, it keeps the option open, as down the track parties get better at encouraging their voters to mark preferences. It’s also completely out of keeping with the Australian system to impose an arbitrary threshold, where 4.1% makes you eligible to win and 3.9% means you can’t win.

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


  1. Thresholds won’t stop small parties from running, but it may change their behaviour. Without being able to make deals to get themselves elected, the focus for preference negotiations from small parties might shift towards demanding policy commitments from the larger parties in return for their preferences. I’m not sure the larger parties will like this.

    Since the larger parties can still benefit from preference harvesting, it may also encourage them to support the creation of ‘front’ micro-parties intended to channel preferences towards them. So it doesn’t fix the problem, just changes who can benefit from gaming the system.

    Other than the unique circumstances of the SA 2013 situation, where, let’s face it, the voters preferred Xenophon to the Greens so get over it, I would think the Greens are more likely to be disadvantaged by thresholds as opposed to abolishing group voting tickets. A major party whose surplus is behind the Greens primary vote could still overtake them on a preference harvest.

    I’d also tend to think that thresholds would benefit the Coalition more than Labor. There seem to be very few small parties that generally preference Labor ahead of both the Greens and the Coalition, but there are plenty of small right-wing parties who preference the Coalition ahead of both Labor and the Greens, whose preference flows to the Coalition would most often be far weaker if left to the voters to decide.

    So aside from not fixing the problems and still delivering election outcomes that don’t reflect the genuine will of voters, I’d tend to think thresholds may well not be in the interests of Labor and the Greens anyway, unless they have a plan to game the system through setting up a myriad of fake micro-parties, and as Ben says, it’s impossible to predict over the medium-term what party a reform will benefit. Safer to go with a reform that benefits the voters.

    Unfortunately it’s in the nature of all established political parties to be motivated by their own self-interest to some extent. I hope in this case that aside from being bad, unecessary policy that doesn’t address the real problem, Labor and the Greens also realise that it may not actually be in their self-interest either.

  2. My understanding of a threshold system was

    a) It’s based on your Senate PRIMARY vote, not the vote with preferences.

    b) If you don’t poll the threshold primary vote, your preferences don’t even count.

    So, doing “dodgy preference deals” with micro-parties, or setting up a harvesting system, is irrelevant. These parties will all be excluded and their preferences will go nowhere.

    I don’t mind this type of threshold system in principle. “Genuine” minor parties like the Greens or Xenophon would easily clear 4%. Palmer United (or some equivalent) would surely clear it in Qld and WA at least. Maybe Family First in SA or Vic. Possibly the LDP.

    The only parties who would suffer are the all bullshit preference-harvester parties, and their very reason to exist would be taken away from them if their preferences were excluded.

    That doesn’t sound unfair to me.

  3. I have not seen any mention of the idea of a threshhold for having preferences passed on in the current (2015) discussion. If it has been revived I would be very interested to see quotes or links to that effect because it is a grossly undemocratic and presumably unconstitutional idea.

    A threshhold for having preferences distributed was floated as an option (not a formal proposal) in the Liberal Party submission to JSCEM. In that case the threshhold was proposed to be set at 1.4%, a figure that seemed cherry-picked to exclude left micros but not right religious micros. As far as I’m aware it was a thought bubble that has not resurfaced since.

  4. Arguments about ‘vote splitting’ and ‘wasted votes’ are already nonsensically overstated in Australia as a way of suppressing new voices. Imagine if it became a dominant theme of each of the 3 established parties’ campaigns and you have an electoral system which enables the established parties to de-legitimise other parties and independents.

  5. One of the issues that goes hand-in-hand with the Senate voting is the registration of the political parties. If the present recommendations about registration of non-parliamentary political parties are adopted, then it will effectively remove most of the smaller political parties from the scene. The problem with Senate voting is then resolved!
    I think that this removal of the smaller non-parliamentary political parties is “the elephant in the room”.

  6. I’m going to do a blog post tomorrow about party registration.

    As I understand the proposal, the number of members required to register a party will rise to 1500 members, but it will be possible to register federally for particular states, with a pro-rated membership threshold. Roughly, there’s 10 members per seat, so in NSW you’d need approximately 470 members, and in Tasmania roughly 50 members.

    So if you’re in NSW there would be approximately the same requirements as now (possibly with the addition of requiring more explicit recognition of membership, which I think is a good thing) but if you want to register a party in a small state it will become substantially easier.

    This will help reduce the effect of a party not just running in the state where they are based, but once they are registered running everywhere else to swap preferences, which should shrink the ballot paper.

    Having said that, I still don’t believe that party registration and candidate nomination rules will do much, because they aren’t the cause of the problem. All they mean is that poor legitimate parties get knocked out and rich people face little extra burden.

  7. The key design principle for reforms of Senate voting should be about transparency of party decisions and accountability and maximum power to individual voters to make informed decisions. Anything else simply opens up a democratic deficit

  8. Hi Ben and tallyroom,

    Given your comments on the lottery what are your thoughts on the citizens jury replacement of the senate as proposed by the new democracy foundation research.

    Politically a lot more unpredictable, however a lot more effective at a representative group of people in the senate looking over the facts.

    I look forward to hearing your thoughts.

  9. Hi Craig,

    I think using juries as a deliberative body can be a useful process in democracies.

    I think they are particularly useful in the context of achieving electoral reform where the elected legislators have an inherent conflict of interest (ala the Citizens Assembly in BC last decade).

    I also saw a Citizens Jury work very effectively in discussing infrastructure funding and rates levels for Marrickville Council recently.

    I am sceptical of their value to actually replace electoral bodies.

    I would also argue that the Senate bears some similarities to a lottery in that people throw in something (in this case, their preferences) and someone pulls out the lot, but I don’t think it’s entirely random or done in a fair way. People are capable of expressing real preferences, and those should decide who wins.

  10. Good article.

    I had some sympathy to the 4% threshold, but looking closer it’s a shocker. Its entirely possible that parties with similar platforms will cannibalize each other’s vote meaning you get three genuine parties on 4 minus x% and each can’t be elected, but their preferences can be harvested to parties lower down the chain and with different platforms.

    I still think though that we live in a funny time where we look to a voting fix for house which is intentionally fundamentally undemocratic in how it works.

    Changing the Senate numbers to reflect actual voters and one term elections would be a start for democracy.

  11. Assessing how democratic the Senate is needs to take account of its role in balancing out the less democratic features of the House of Reps rather than looking at it in isolation

  12. Upper houses are (normally) inherently undemocratic and usually something that the right rather than the left benefits from.

    The only reason I don’t call for its immediate dissolution is that almost coincidentally it has developed some more helpful features that removing it would need to consider.

    But I shake my head when I hear left wingers complaining that QLD doesn’t have an upper house…

    Making Senate seats one term and twelve seats elected at each election is about the biggest and most progressive change that I could imagine actually passing at present.

  13. What is “inherently undemocratic” about upper houses?

    The use of proportional representation in all mainland upper houses makes them substantially more democratic than their respective lower houses. While the malapportionment of the Senate is unfortunate, that doesn’t apply to Victoria, NSW or South Australia.

    I do agree that having MLCs and Senators serving longer terms than MPs in the lower house is unfortunate and should be abolished, on balance the use of a proportional system still makes them more democratic.

    And the Victorian and Western Australian Legislative Councils demonstrates that it isn’t necessary for an upper house to have six- or eight-year terms.

  14. All around the world upper houses are used restrict popularly elected governments.
    It’s great that the Senate uses proportional representation, but that’s not a world wide rule.

    As it is the Australian Senate has a tick for proportional representation, but several crosses for State’s receiving a fixed number of seats not by population, the high quota, the split terms and probably a few other issues that I haven’t thought of.

    As it is many on the left (and the general population) probably consider these good things because at the moment stopping Abbott or Howard in the Senate has been the electoral goal. I also think that is true enough that many Australian voters like a bob each way.

    That’s all very good, but it’s not to say the system is fair or democratic. Moving toward an ideal fair system would be one in which one upper (or lower house) vote anywhere in Australia was worth the same as one vote anywhere else.
    Also one where if you are getting 10% of the vote you end up with something like that many % (of total) seats and so on.

    I don’t think wholesale voting reform is possible. Equal terms with a full election (and hence a much lower quota) is probably too much to hope for.

    Just looking at the upper house results it would probably make too much of a progressive improvement.

  15. None of that is “inherent” to an upper house, though. What do you think of the Victorian Legislative Council.

    Those arguing for an upper house in Queensland presumably support one elected proportionally, with one-member-one-vote and four-year terms.

    You also ignore the fact that the democratic processes to elect lower houses (and thus governments) are also flawed, in producing majority governments off a minority of the vote.

  16. Dictionary definition of inherent gives
    vested in (someone) as a right or privilege.
    “the president’s inherent foreign affairs power”

    Perhaps traditional , historically or usually would be more helpful. Happy to use any of those terms instead.

    Explicitly the Senate in Australia was setup as with concessions for State rights. Some senators still refer to that as it’s role even though the party system has limited it. The upper house in QLD was appointed by the Governor! The house of lords was setup for nobility with lifetime appointments. Pinochet rangled a life time spot in the Chilean Senate explicitly so as to be able to limit the democratic process. Other upper houses haven’t even had elections or universal suffrage!

    Now in almost all these cases the system has been drastically improved. That doesn’t change that the right or privilege of the upper houses was to limit the democracy of the masses. Usually the lower house at least had universal suffrage and no lifetime appointments despite it’s democratic flaws.

    So I think it’s fair to write that the usual role of upper houses has been in limiting democracy even if democratic reforms have potentially made them more democratic than some of their lower houses.

    As it is I’d probably be happy with the Senate if it’s flaws were fixed. I’d also like to see lower house reform.

  17. If a popularly elected upper house restricts a popularly elected lower house then what that often means is that a bill is acceptable to the representatives elected by one method is not acceptable to those elected by another. There could be many bills that would pass a popularly elected upper house that would be blocked by the lower house if sent there, but we less often hear about those. Assuming (which I do) that it is more important to prevent governments making bad decisions than to ensure they can always make good ones, I think the idea that the people elect a government in two chambers and a bill needs to pass both is fine.

    Specific upper houses have major issues though. Group ticket voting is a problem in some. Tasmania’s has excessive power and inadequately understood, overly restricted elections. The Senate is grossly malapportioned. NSW has terms that are too long. (etc)

Comments are closed.