Wednesday, December 13, 2017

Senate Section 44: The Term Lengths Issue Is Back

Once upon a time, a young chap in the Senate discovered that he was a dual New Zealand citizen and resigned.  Back in those quaint, far-off days (it was actually July this year), the fact that he was the holder of a six-year term was one of the most intriguing things about the situation.  With Ludlam's resignation merely the start of a citizenship issue that has now claimed eight MPs, with two more currently referred and questions about many others, the term lengths issue has been on the sidelines.  The High Court following Re Canavan simply appointed Jordon Steele-John to Ludlam's vacant place and it was assumed that that was all, perhaps because there wasn't an alternative.  But it turns out that was all because nobody suggested otherwise, and following a mention of the question by amicus curiae ("friend of the court") Geoffrey Kennett in the Fiona Nash case, the issue is back.

Firstly, although Steele-John is listed on Senate documents as having a term expiring in 2022, the WA Greens conducted a preselection (which he won) on the assumption that he would be facing the voters again in 2019.  Secondly, the issue has now been brought to the High Court's attention in cases dealing with the replacement of Senators Nash, Parry and Lambie (who all had six-year terms) and in the Lambie case it impacts on the future party makeup of the Senate.

As previously noted ("Majors Stitch Up Senate Term Lengths, Film At Eleven"), the Senate allocates Senators to short and long terms following a double dissolution.  It can do this on whatever basis it likes (including even a random or arbitrary basis), but has in practice always used the order of election method.  Currently, the order of election method means that the first six Senators elected in each state receive long terms and the next six receive the short terms.  The order of election method is a crude solution that can have very arbitrary results and could be quite unfair in some situations.  Nothing in the Constitution specifies what happens to the term lengths if someone is disqualified by the High Court, and one reason for that is that when the Constitution was written, the voting systems of the time did not allow for the reuse of the original ballot papers to determine a new Senator.

The cases of Ludlam, Nash, Lambie and Parry all have the feature that a Senator originally elected in the first six is set to be replaced by a Senator who is not elected in the first six in the special count to replace them.  I can identify at least the following four views on what the High Court and/or Senate can do about this:

1. As Section 13 of the Constitution only provides for the setting of terms as a one-stop process after each double dissolution, there is no power to alter terms beyond that.  The Court can only appoint new Senators to replace former Senators on a one-for-one basis, the new Senator inherits the term of the Senator they are replacing and the Senate can do nothing about this.

2. As the Senate's resolution on term lengths assigned those elected in the first six places to long terms and the rest to short terms, this resolution automatically by analogy applies to any replacements, with the order found in the special count determining who gets long and short terms, and neither the Senate nor the High Court has any power to do anything different.

3. By analogy with the procedure in Section 13, since the Senate sets the term lengths of the original Senators, it can also set the term lengths of a replacement Senator, and if this results in a Senator acquiring a different term length to the original Senator, then the Senate can also alter the term length of another Senator in order to make the numbers of Senators holding each term length even.

4. As the special count procedure overwrites the original election for a state with a new result, the original allocation for that state is void and the Senate reallocates terms for all the Senators from that state.  (Thanks to Andrew James on Twitter for suggesting this one).

Variants on these themes are also possible.  For instance, here's what I'll call the too-much-coffee variant of solution 4: since the original allocation of Senators might have been based on an agreement that a certain package-deal outcome across various states was fair, any disqualification allows the Senate to revisit the terms of every Senator, including those from other states.

Lambie replacement as a test case

The Nash and Parry replacements (ditto for Ludlam) only involve a new Senator potentially jumping into a six-year term from below Senators from the same ticket who keep three year terms.  This is slightly complicated in the Nash case by the presence of the two Coalition parties on the same ticket.  In the Ludlam and Parry cases I can't see much dispute that if it is legally possible for the Senate to rearrange the terms so that those at the pointy end of the ticket get the longer ones, then this is the right thing to do.  In the Nash case there is only the added wrinkle that since she was a National, perhaps the other National on the ticket (John Williams) might be said (at least by Nationals) to be best selected for the six-year term.

However the Lambie situation is where the problem becomes more important.  Lambie's ticket polled over a quota in primary votes, but included a massive below-the-line personal vote for Lambie, almost half of which failed to flow to her ticket #2, the mayor of Devonport Steven Martin.  In the special count to fill the Lambie and Parry vacancies, the order of election of the top six changes from:

Abetz (Lib), Urquhart (ALP), Whish-Wilson (Green), Lambie (JLN), Parry (Lib), Polley (ALP)


Abetz (Lib), Urquhart (ALP), Whish-Wilson (Green), Duniam (Lib), Polley (ALP), Bushby (Lib)

(with Lambie's replacement now elected ninth).

So if the order-of-election method is applied (either imposed by the Court or chosen by the Senate) then the Lambie Network is reduced from a six-year term to a three-year term, and the Liberal Party gains an extra six year term.  Firstly, the Liberals won only four of twelve Senate places in Tasmania to Labor's five, so they do not deserve to get three long terms to Labor's two.  Secondly, this gives the Liberals a more or less automatic seat gain in Tasmania at the next half-Senate election, at which they would be defending only one seat (Colbeck).

In my view, this is a strong argument against solution 2.  The Senate did decide to impose a particular principle in allocating terms, but one could hardly say it did that blindly on general principles and without awareness of what the consequences of that decision would be.  Indeed one would hope it would not have used the order-of-election method had the order-of-election method resulted in something massively favouring one party or another.  So it cannot be assumed that the Senate necessarily wanted the order of election method reused in a case where it might alter the balance of the Senate.  (One might also more flippantly say it's wrong to hold the current Senate to the will of the original Senators when so many of them have left!)

However, while the Court might decide solution 1 is the only legally viable option, it doesn't have much going for it on the fairness front apart from that.  Jacqui Lambie won a six-year term by virtue of personal votes that pushed her party over a quota.  Without those personal votes, no other lead candidate for her party would have done that.  She also would seemingly have been elected, based on the votes cast, had the election been for just six Senators (see the Section 282 count) although, as the stitch-up article pointed out, that method isn't perfect either.  It's not clear whether Lambie's replacement would win the Section 282 advisory count if one was held (Martin might still beat Lisa Singh, but it would be close).

Lambie's #2 and #3 candidates are both being challenged by One Nation, who are claiming that both are ineligible.  Should both be disqualified then One Nation, a party not even elected at the original election, would according to solution 1 inherit a six year term.

Solutions 3 and 4 are more flexible, but amplify a problem with the existing method of allocation.  It's simply not right that the Senate allocates terms itself (as is currently done) since this opens the door to partisan manipulation of terms to try to entrench an advantage.  On that basis, it's a further concern if the Senate is able to vote on altering terms every time someone is disqualified in the years after a double dissolution.  It's probably even more likely that partisan considerations would influence such a vote when it was held mid-term in the heat of battle over various issues of the day.  Probably, this isn't the Court's problem, but rather the Senate's.

Anyway those trying to guess which way the court might jump may enjoy reading the transcripts regarding Nash and Parry/Lambie.  The biggest hint we apparently get is that Nettle J describes the idea that the Court should decide who is elected and the Senate can reset the terms if it wants to (Solution 3 or 4, presumably) as "an attractive proposition", and it might well be that everyone is going to agree and leave it at that.  This might then leave the question of whether Steele-John is a special case because of having been elected "in place of" Ludlam, as opposed to just being declared elected.

Meanwhile Tasmania won't be getting any new Senators until at least the Nash term lengths matter is resolved.  The next step in resolving that is a directions hearing on 22 December to discuss whether the Nash term lengths matter can be resolved by a single judge or requires referral to the full bench.  Even if that leads to a quick decision (eg that Molan and by implication Colbeck can be seated and the Senate decides how long for) the Martin matter will need to go to the full bench in the week of 29 January.

I have been meaning to write about Section 44 in the Reps as well, but this article is long enough as it is, so I will leave that to some other time.  Comments about S 44 and the Reps are welcome on this article too.

Update Dec 22: Molan has been declared elected but with Gageler J apparently not addressing the term lengths issue.  Presumably the Senate will be left to sort it out and then anyone who doesn't like the outcome (on which they might have trouble establishing standing) could in theory challenge.  In this case it is clear that Senators Fierravanti-Wells and Molan agreed that the former should get the six year term and the latter three years. We will have to wait to see what kind of precedent this sets for the Colbeck/Martin case and whether it is handled the same way (given that in that case agreement between the Liberals and JLN might not be so forthcoming!)


  1. "with the order found in the special count determining who gets long and short terms"

    Surely that would be an odd thing for the High Court to find. It has been the order in the ordinary count that determines who gets long and short terms, no?

  2. The argument would be that the original "ordinary count" no longer exists. Indeed that is how the AEC website portrays matters, writing over the original counts with the special counts whenever a special count is held.

  3. Disclaimer: I am not a Lawyer.

    The wording of section 13 of the Constitution would tend to favour options 1 or 2 as it uses the dissolution of the Senate, rather than the election of the Senators, as the event which the sitting that decides the allocations and thus makes arguing Senators being newly elected and thus needing new allocations near impossible. I also think that 2 is more sound, especially given that the motion to allocate the terms did not specifically mention Senators by name, only mentioning their listec order of election.

    The wording of section 13 separates the division into terms by state, make the far too much coffee variant of 4 a distictly unviable view. It also sounds a bit too political for the High Court.

    The division of term allocation into individual states is so strong thatit is likely that, were the election of Senators in different states to be preeceeded by different 1st of July they would get different term starting years.

    The High Court might use this as an oportunity to rule whether Senate terms set after a DD are individual or by state as the first and third paragraphs of section 13 contradict eachother on he matter with their variant numberd 3rd person possesive pronouns (even though the plural 3rd person pronouns can be used as the common singular, instead of the gendered singular possessive pronoun used).

    If the NSW Senate term allocations end up coming up again, I wonder if their will be a push to extend Senator Rhiannon`s political career?

  4. Great analysis. Tassie voters think for themselves. Half only voted for Lambie but not her running mate!.

  5. Tasmania is trained to comnsider individual candidates by Hare-Clark and has the lowest population to senator ratio meaning that there are fewer candidates, making choosing easier. The whole of Senate election also increased the likelihood of a Lambie surplus, potentially increasing the desire to preference away from the rest of Lambie`s ticket as ATL voting could get them elected.

    The save Lisa Singh and save Richard Colbeck BTL campaigns probably also have drawn votes away from the rest of Lambie`s ticket by giving people further reason to vote BTL. 5.5% of Lambie`s suplus went to Singh. Colbeck got 1.7% of Lambie`s surplus.

  6. There is another interesting wrinkle around Lambie and term length. Assuming Steve Martin replaces Jackie Lambie on the same 6 year term, and Lambie stands again at the next senate election... JLN would have two senate seats for a term. Both having been won on her name, neither of which would (likely) have been won with a party running mate heading the JLN ticket.

    Don't think its a reason to favour one option over another for resolving issue. But it is an irony.

    1. Very good point there. Provided Lambie can actually win at a half-Senate election (which is probable) her party would have benefited from her own failure to check her citizenship.

    2. I was saying something similar to anyone who would listen as soon as Ludlam, with his significant personal vote, stepped down. Currently looks like Ludlam is done with politics, but we're a while from a half-senate yet, so it could who knows.

    3. This comment has been removed by the author.

    4. The WA Greens are not a one personality party like JLN (or NXT/SA Best or PUP). The Greens have won with Rachel Siewert in 2004 (before Ludlam was in the Senate) 2010 and in 2016 (as second on the ticket to Ludlam). From the item on 7.30 about Steele-John (with an interview), he looks like a decent politician who could develop a personal following before he faces re-election in 2021-2022.

      (Sorry about the omission needing the delete and repost.)

    5. The Greens have a long history of doing quite well in WA, and Rachel Siewert is quite high profile there. If The Greens chose to preselect Ludlam instead I don't think his personal vote would be a significant factor in the decision.

  7. I've been waiting for one of these one-person-brand parties to take this to the logical conclusion, by having their headliner run every three years. They would resign ahead of the election and be replaced (casual vacancy) by a trusted offsider, then run on a "Six More Years" slogan. If they get elected, they've now got two Senators. If they fall short, the offsider could step aside (casual vacancy again) to allow the headliner back in. In fact if they got the timing right, they could make it that the casual vacancy isn't filled until after the election and the headliner could fill their own vacancy if they fail to win a new term.

    The main risk with the strategy would be a backlash from the obvious cynicism involved. But I still think the net chances of ending up with two seats would be improved.

    1. That's quite sneaky. Another risk is the state parliament that has to endorse the replacement for the casual vacancy might get sick of it and decide they will no longer endorse anyone who has previously resigned unnecessarily. (There's a claim that the Greiner government was making noises that if Irina Dunn resigned to allow Robert Wood to resume his seat, they might not endorse him.) The state parliament is required to only endorse a nominee approved by the party, but is not required to endorse any specific nominee, which in the John Devereux case led to a brief standoff.

    2. Section 15 has no mechanism for insuring that state parliaments actually fill Senate casual vacancies. The John Devereaux stand-off was only resolved by the election following the Australia Card Double Dissolution to throw off Joh for Canberra. It could have lasted until 30/6/88 if there had been no DD, because that is when Grimes`s term ended but had he held a 1991 expiring term, it could have lasted until after the 1989 state election. Countback, with Senate President issued writs (with the governor and or governor-general as back up), would have been a far better sollution for the prevention the reccurance of the casual vacancy issues of 1975.

      Such countback would also prevent the use of the temporary executive appointment of casual vacancy fillers, when state parliaments are out of session, to apoint Albert Field type candidates before their party can expell them.

    3. I think Gray would have stopped the silly games eventually in the Devereux case, but then again given it was Gray, perhaps he wouldn't have.

      Re Dunn/Wood I have found material indicating that the Greiner government was *not* intending to get in the way of Wood's reappointment (despite Dunn's concern that this might be the case).

      Apologies for the delay in clearing the above comment - somehow I either missed or never got notifications for a small number of them.