The Spurious Linking Of "One Vote, One Value" With Territory Senator Numbers

After each election comes a new season in which the Joint Standing Committee on Electoral Matters receives submissions and considers proposals for changes to electoral law.  This JSCEM season has special significance because as well as a change of government in the lower house, there has been a serious shift to the left in the Senate.  Any ALP legislation that is supported by the Greens and ACT Senator David Pocock will have the numbers to pass.

There have been several media articles commenting about this, though it is not always clear to what extent the articles are reporting on what Labor wants, and to what extent they are reporting on what other actors would like Labor to do.  A common theme in these articles (here's the latest) is that a proposal for more ACT and NT Senators appears in the context of a discussion of "one vote, one value" (a principle to which Labor's policy platform included a general commitment without any specifics.)  The linkage of the issue to "one vote, one value" is spurious.  From a pure one vote, one value perspective, the proposal looks like an attempt to rig the Senate to favour the left.

This argument has been appearing - assisted by overly credulous reporting - in the form that it is not fair that the ACT, with a population of about "460,000", gets only two Senators, while Tasmania with "560,000" gets twelve.  (That's some curious rounding there since in fact if rounded to the nearest 10,000, the ACT has 450,000 and Tasmania has 570,000.) The problem with this argument is that comparing to Tasmania is cherrypicking.  Tasmania is the most outrageously over-represented State in the Senate and is an example of what not to do if one has the choice.  As it stands, the three largest states are underrepresented compared to the ACT and every state except Tasmania is underrepresented in comparison to the NT.  

A note that through this article I will be using "the left" to refer to Labor, the Greens, any similar parties and tentatively David Pocock, and "the right" to refer to the Coalition, UAP, One Nation, LDP etc.  These are relative terms based around the common appearance of left and right sides in Senate balance debates.  

The following table shows the number of Representatives and Senators per head of population for each state and nationwide (based on the 2021 Census results):

The ACT has 1.46 times the national average number of Senators per million people, and the Northern Territory has 2.72 times the national average.  These figures are not as high as Tasmania's (7.14 times the national average) but increasing the ACT's or the NT's share of the Senate would further increase the disproportionality of the Senate.  On a one-vote-one-value basis, voters in NSW, Queensland and Victoria would become more underrepresented, voters in the Territories would become more overrepresented and only voters in Western Australia, South Australia and Tasmania would move very slightly towards where they should be.  (Our system uses population as a proxy for voter numbers, but this doesn't have much impact on the conclusions.)

My own view is that the existing level of Senate malapportionment is wildly excessive and we're just lucky it hasn't bitten as hard as it might.  I don't get involved in that debate all that often, because Senate malapportionment was a price of Federation and has therefore been entrenched in a way that is in practice impossible to unpick.  Under Section 128 of the Constitution:

"No alteration diminishing the proportionate representation of any State in either House of the Parliament, or the minimum number of representatives of a State in the House of Representatives, or increasing, diminishing, or otherwise altering the limits of the State, or in any manner affecting the provisions of the Constitution in relation thereto, shall become law unless the majority of the electors voting in that State approve the proposed law."

There is a view about that this might in theory be circumvented by first having a referendum to remove this limit and then having a second referendum to make the Senate more proportional, but even if that view is valid it's hardly likely that the first stage would carry four States at the moment.  At least one State advantaged by the current system would need to be the proverbial turkey voting for Christmas.

Defenders of the current system who I encounter often confuse history with morality (just because the Senate system was designed to protect States does not mean that severe violations of one vote, one value are the right way to achieve such an aim). They also often think there is something special about small States that makes them worthy of representation to protect their interests when other more disadvantaged minorities do not have any special Senate rights.  And they don't have too much to say on the empirical question of whether Senate malapportionment actually achieves its aims.  The only example most of them can name is Brian Harradine, who also used his unrepresentative Senate power to pursue a religious morality agenda unrelated to representing a State.  (I suspect that an overabundance of Tasmanian, SA and WA Senators within party caucuses has had at least a modest impact in the less populated States' favour over time but assessments of that appear thin on the ground.)

The impact of increasing ACT Senate numbers

By and large, Senate malapportionment has had fortuitously little impact because the States in general vote close enough to alike.  This is a major difference with the USA, where averaging results across states produces a skew equivalent to over 2% 2PP, because small states tend to vote Republican.  There is not any consistent relationship between an Australian state's population and its  lean between the major parties, and over time which side would appear favoured by the Senate's method of effectively averaging results across states has varied from election to election.  For instance in 2001, 2010, 2016 and 2019 Labor's average Reps 2PP per state was lower than its actual 2PP (largely because of strong results in Victoria) but in 2007 and 2022 it was higher.  

Giving the ACT four Senators, however, would potentially distort the Senate in favour of the left in a way that the existing malapportionment hasn't done.  The reason is that the ACT is very left compared to Australia overall (by over 10 points 2PP in the Reps at the last nine elections in a row, including a 50-year high of 14.8 points higher in 2022) and the ACT would habitually return a 3-1 left-right result.  Had the current system been in play in the long term, I estimate the following results:

1990 2 Labor 1 Liberal 1 Democrat

1993 2 Labor 2 Liberal

1996 2 Labor 2 Liberal

1998 2 Labor 1 Liberal 1 Democrat

2001 2 Labor 1 Liberal 1 Democrat

2004 2 Labor 2 Liberal (just, very close to 2 Labor 1 Liberal 1 Green)

2007 2 Labor 1 Liberal 1 Green

2010 2 Labor 1 Liberal 1 Green

2013 2 Labor 1 Liberal 1 Green

2016 2 Labor 1 Liberal 1 Green

2019 2 Labor 1 Liberal 1 Green

2022 2 Labor 1 Liberal 1 Pocock 

If removing Pocock (the impetus for his campaign would probably not have existed in a four-seat system) then 2022 is 2 Labor 1 Liberal 1 Green.  The Democrats were overall more centre than left so from 1990-2004 the four-seat system would sometimes advantage the left by one seat and sometimes by none.  But from 2007 onwards it advantages the left by two seats, which would be 2.5% of an 80-seat Senate, an advantage exceeding the ACT's share of Australia's population (1.76%).  All these elections except 2022 returned a 1-1 major party result for the ACT, with 2022 returning an exceptional 1 Labor 1 Pocock result.

In the meantime, the NT with 4 Senators would have consistently returned a 2 right, 2 left result, usually in the form 2 CLP 2 Labor, though in 2022 it would have been 2 CLP 1 Labor 1 Green (and in 2010 close to that result as well).  

So while there's a claim that "two senators are considered more likely than one as each major party would likely gain one extra senator each from both territories", in the NT's case that isn't true with respect to 2022 (indeed the CLP would also only have just held off the Liberal Democrats) while in the ACT's case, 2022 is probably the only election since 2004 for which it is true.  With Andrew Conway's site available to test alternative scenarios, it is easy to see that such claims are on shaky ground for the last few elections.

With 4/80 Senators each, the ACT would have 8.82 Senators per million people and the NT would have 16.04, compared to a national figure of 3.10.  The ACT would be overrepresented by a factor of 2.85 times and the NT by 5.17.  They still wouldn't be as overrepresented as Tasmania, but the overrepresentation in the ACT's case would bite harder in terms of its consistent impact on results.  It might be argued that with David Pocock now elected, the left could get a consistent 2-seat legup in the ACT under the current system anyway, but it remains to be seen if that holds true.  

Raising the ACT to three Senators instead of four would probably be fairer, as it would typically result in a 2 left, 1 right split, which is a fair representation with a more or less fair impact on the balance of the Senate overall.  An argument that can be made for doing this is that originally made against giving the Territories just a single Senator - that having at least one representative of each major party from each Territory is desirable.  

However raising the NT to three Senators at the same time would also generally produce a 2 left, 1 right split there as well (2013 is a probable exception and 2001 a possible one) despite the NT being only mildly favourable to Labor on average. 

Giving the ACT an extra Senator while not giving the NT one is probably the step that would least unsettle the national picture, but it would change the number of Senators to odd, though valid arguments could be made in favour of that radical step.  It would mean that all else being equal, the numbers to block and pass motions would become the same.  

Note: In comments it's been pointed out that the Territories could be given two extra Senators alongside switching them to six-year terms.  In the case of the NT this would consistently result in 1-1 left-right splits.  In the case of the ACT the risk would be that more Pocock-style or Pocock-supported candidates emerged and that the Liberals might still not be able to win seats.

Background, and are more Senators allowed?

It seems a bit surprising nowadays but the creation of Territory Senators with voting rights was quite controversial, passing the High Court 4-3 in 1975.  In that case arguments were made that the power to create Territory Senators could be abused to flood the Senate without limit, but they were given short shrift by the Court - those judges examining this argument often held that the Constitution assumes Parliament will in general act reasonably and therefore an otherwise reasonable proposal shouldn't be struck down by a thin-end-of-the-wedge complaint.  The Court could always cross that bridge when it came to it.  

It would be surprising if an attempt to create more Territory Senators wasn't challenged, but I haven't yet seen anything clear regarding whether there would be any basis for the High Court to reject a modest increase if valid arguments were made for it.   Would the Court have any reservations if (as is currently the case) the argument being offered is simply and obviously false?  

It's interesting that the original choice of two Senators per Territory was designed to avoid the Territory results impacting on the national picture, as might be the case had the Territories been granted more Senators.  In 1975 there was immediately a prospect that this might not be the case when John Gorton ran for the ACT Senate as an independent, but Gorton fell well short.  Now, for the first time, it has happened, leaving the ACT without a Liberal Party representative.  

Through the 1980s, JSCEM discussed the issue frequently, and often favoured adopting some kind of formula to limit the number of Territory Senators.  None of them would have resulted in the ACT being allowed four Senators at present.  As noted in comments, Section 40(2) of the Commonwealth Electoral Act currently allows for extra Territory Senators once a Territory becomes entitled to a sixth House of Reps seat, which neither Territory is anywhere near doing.

District magnitude

It is correct to argue that the Territories have an inferior form of Senate proportional representation compared to the States, and that this is a form of discrimination against their voters.  All the States get to vote in Senate elections with either 6 or 12 seats at a time, ensuring that parties with substantial minority support will win seats and leaving little room for strategic voting arguments (although they are there if one looks closely enough).  Two seats just isn't ideal for a proportional election.  The recent ACT Senate race gave Labor and Greens voters a tricky decision as to whether they should vote 1 for their preferred party, or vote 1 for David Pocock.  The strategic arguments for voting for Pocock were (i) that he was more likely to beat Zed Seselja on preferences than the Greens were (and indeed the Greens would not have beaten Seselja had Pocock's team not run) and (ii) that bringing Labor's primary vote down to about a third (which the voters collectively managed almost exactly) would avoid leakage to the Liberals and exhaust on Labor's surplus. The result has been seen as a case of massive tactical voting, but it's hard to distinguish between that and voters shifting from other parties to Pocock because they actually preferred him.  

The number of ACT seats could be increased without further worsening the Senate's current parlous one vote, one value situation only by increasing the size of the Senate by 50% or more, allowing for a similar increase in the number of Territory Senators.  That would result in an increase in the number of Reps as well, based on the nexus which ties the number of State Senators to around half the number of Representatives.  Which brings me to ...

Reps malapportionment

The issue of malapportionment in the House of Representatives is also likely to come under some scrutiny.  There is very little malapportionment in the Reps these days, but what there is systematically favours Tasmania, and in practice at present also favours the NT and the ACT.  The Tasmanian malapportionment stems from the five-Representatives-per-state minimum that currently means the state gets five MHRs when population only entitles it to 3.33.  For the Territories, it comes from the lumpy nature of trying to represent a population accurately with one, two or three seats, which in the NT's case resulted in the formula being changed to keep it at two seats and protect it (for now) from under-representation.

As it happens, both the Tasmania and NT over-representations would be solved nicely on current numbers by increasing the size of the House by 50%.  This would leave about the following seat numbers: NSW 71, Victoria 58, Queensland 46, Western Australia 24, South Australia 16, Tasmania 5, ACT 4, NT 2.  (Total 226).  At the same time increasing the size of the Senate by the same amount would allow the ACT and NT to have three seats each at least without making the existing overrepresentation any worse, providing the Territories with a slightly better quality of proportional representation.   And given that the population has grown by 66% since the last substantial increase in the size of the Reps, there's an argument that a 50% increase is not excessive.  

However, increasing the number of politicians is generally not popular, and I suspect a smaller increase would be more palatable.  I do think an increase is overdue as many divisions have simply too many voters to allow for direct and well connected representation; that an increase would help address Reps malapportionment is another argument for it.

I hope this article will contribute to a lifting of the standard of the Senate numbers debate.  There are some valid arguments for increasing the ACT's Senator numbers, including ensuring representation for both major parties and improving the representativeness of Senate races.  However, there are also risks to the balance of the Senate.  To dress the issue up as a "one vote, one value" issue (perhaps in order to pretend that Labor has a mandate for it), when in fact it makes the Senate's "one vote, one value" problems worse, is not on and I won't put up with it.