Breaking up the Senate? Why it’s such a bad idea

The Australian reported on Sunday about an idea from Queensland LNP senator James McGrath that would see the Senate broken up so that senators represent “provinces” within each state. While details are scarce, this proposal would see the end of proportional representation in the Senate, would likely wipe out all minor parties and would see one-party rule in the Senate, largely replicating the results of the House of Representatives.

The idea is obviously self-serving and is very unlikely to go anywhere, but it’s such a bad idea that I think it’s worth explaining why it’s so bad. I also had a go at modelling what Senate elections would look like under such a system.

Firstly I’m going to define the proposal. The journalist seems confused about the idea, suggesting that legislation would not be needed to “forcibly change the make-up of the Senate”, but then quotes Senator McGrath who specifically proposes “breaking the states into provinces and having two senators elected across six provinces within each state”. So I’ll assume the proposal would be to draw up six “provinces” in each state, with each province electing one senator per election for a six-year term, with two senators representing each province.

I don’t believe such a proposal would be unconstitutional, but it would most definitely require legislation.

So this would firstly end the system of proportional representation used to elect the Senate since 1949. It would likely result in no more Greens, independents or other minor parties winning seats in the Senate. While a handful of crossbenchers have won seats in the House of Representatives, most Senate provinces would be larger and there would be a smaller number of districts. Even in inner-city Melbourne I don’t believe the Greens would be in a position to win a province which would cover more than six House electorates.

While this would require a certain number of senators to represent specific geographic regions, it wouldn’t necessarily improve the voting power of rural voters. But it would help the Nationals, who have a concentrated vote and benefit from single-member electorates to win a disproportionate number of seats.

Nationals who have been suggesting this change have been full of talk about “representation” and “diversity” but such a change would strike a significant blow to diversity in most ways. Proportional voting systems tend to elect more women, more members of ethnic minorities, and of course representatives of a broader range of parties. Indigenous Australians are significantly more well-represented in the Senate, and we are much closer to gender parity in the Senate.

Of course rural voters already have the option to vote for rural candidates if they wish, and they often do. A change to the provinces would reduce their options, and mean that their vote would be less likely to matter, potentially living in a safe Liberal or Nationals seat. The Nationals choose to reduce the choice for rural voters by running a joint ticket with the Liberal Party, but the choice is still there.

It’s worth noting that the Senate provinces, particularly in the big east coast states, would be massive and populous. When I made an attempt at splitting New South Wales into six provinces, I ended up with three provinces covering Sydney, plus a fourth stretching from the Central Coast halfway up the north coast, with a fifth province having to stretch from Campbelltown and Sutherland past the ACT to the Victorian border. The sixth, the province likely to elect Nationals senators, would likely stretch from Tweed to Albury. I don’t know how a senator could be any more capable of representing regional voters in that province then they are now in being one of twelve representing the whole state.

I decided to have a go at estimating the impact on past elections. My model is very simple. I split up House of Representatives electorates dating back to the 2004 federal election into six provinces per state. Except in the case of New South Wales from 2010 to 2013 and Queensland since 2010, the number of seats in each state has never divided evenly into six. Where seats needed to be split between provinces I simply distributed the votes of each seat evenly. Like I said, it’s a simple model. But it gives you a sense of how real House of Representatives votes would translate into these massive electorates.

I also assumed that the two senators each for the Northern Territory and the ACT would effectively be elected by the two electorates these territories were each covered by (at least until the last election in the ACT). So that means I assumed Labor would win both ACT senate seats, and that Labor would win the second NT senate seat in 2004, 2007, 2016 and 2019.

I also assumed that no minor parties or independents would win, and thus based the results solely on the two-party-preferred vote.

The result looks quite a lot like the House of Representatives, but with some important differences, mainly down to the whims of arbitrary single-member electorates and the weighting of votes that favours the three smaller states. The overlap of Senate terms also means that election results had impacts over longer periods.

The Coalition did very well in 2004, and would have likely won a large majority in the 2004-2007 term. Labor took back ten Coalition seats in 2007, producing a perfect tie in the Senate. That would have been interesting.

The 2010 election results were much closer, but combined with the 2007 Labor landslide produced a large Labor majority in the Senate, which was replaced by a large Coalition majority following the 2013 election.

The Coalition’s majority was cut to the smallest possible size after the 2016 election (ignoring how a double dissolution would change things – see below), but the most interesting result was after the recent election. The LNP would have won a majority of seats in 2019, but combined with Labor’s 8-seat majority in the 2016 results, Labor would have gained a slim majority in the Senate, facing off against a majority for the Coalition in the House.

This system will often produce majorities for the governing parties, which many people would see as a negative. But it can also produce solid majorities for the opposition, which would be difficult to resolve.

As for a double dissolution, it would be a complete mess. Let’s assume that each province would elect two representatives under the voting system used at the moment to elect senators in the territories. Almost every province would produce a result of one each for the major parties, and there would be very few regions where a seat would even be in play.

Two-seat electorates under proportional representation are dreadful. Unless they are very slanted, both major parties would win one seat each. You’d need a major party to drop below one third of the vote for there to be potential for someone else to win. It’s called “stasis” and can be just as bad, or worse, than the safe seats that exist under the single-member voting system.

Across 36 provinces in six elections (216 individual elections), the two-party-preferred vote for one of the major parties exceeded two-thirds of the total vote in just three cases: Labor cracked 70% in the inner-city Melbourne province in 2007 and 2010, and the Coalition exceeded 67% in the southern rural province in South Australia in 2004. Even in the large North Sydney province (which needs to include parts of Parramatta or Blacktown to meet the population quota) the Liberal Party polled over 63% just once over these six elections.

So the most likely outcome at a double dissolution would be a tie, with a handful of very lopsided provinces creating the potential for one side to win two seats and thus win a majority. Hardly a sensible way to decide an election.

The proportional representation system we have in the Senate actually works pretty well, since the elimination of group voting tickets. There is meaningful representation of a variety of minor parties, but the quota is high enough that you don’t have a massive array of microparties. Some would argue for abolishing above-the-line voting to increase the importance of individual candidates, but the Senate does a pretty good job of representing diverse interests.

The Nationals can whine about the Senate being a “chamber of objection”, but that reflects the fact that no major party has won a majority in a federal election for many decades, so the proportional system does not give the government (elected with the support of a minority of voters) complete power. The one thing holding back the representativeness of the Senate is the malapportionment between states which was a necessary part of the deal under federation. So far the proportionality is pretty good despite this malapportionment, but it would be a lot worse if the Senate was elected through single-member electorates.

The proportional system we have is a feature, not a bug. It was first proposed under the first electoral bill in 1902, after being used to elect Tasmania’s senators in 1901. It was subsequently legislated in 1949, and over more than half a century has produced a wide array of different Senates, but generally has worked pretty well at counterbalancing the House of Representatives while still allowing stable government. While there are minor tweaks I would make (such as my next post), it is not the part of our federal electoral system that is most in need of a fix.

One final side note: NSW Nationals senator Perin Davey sensibly dismisses the idea to change the Senate voting system but instead focuses his attention on the House of Representatives. She mourns the increasingly large size of rural electorates as they become a relatively smaller part of Australia’s total population, and suggests a cap on the size of electorates.

It used to be quite common to have different population quotas for urban and rural electorates, but has been mostly eliminated, most recently for the Western Australian Legislative Assembly about a decade ago. Severe malapportionment remains in place for the WA Legislative Council, and there is a small amount for very large electorates in Queensland and Western Australia.

While it is understandable to be concerned about the size of electorates, and I don’t see any problem with granting extra transport, staff and office resources to MPs representing those areas, we really shouldn’t be modifying population requirements. Ultimately any move to shrink the population requirements for rural seats reduces the value of the votes of other Australian voters, many of whom have their own challenges. Why not also allow for less populous electorates in Western Sydney areas with higher levels of poverty, where people may need more help from their MP in dealing with government services?

If you are concerned about the growing size of electorates there is only one solution: increase the size of Parliament. This doesn’t require any constitutional change. The Parliament can simply legislate to increase the number of senators per state to 14, which would add roughly 24 more seats to the House of Representatives, since the House quota is related to the size of the Senate. This would lock in the third ACT seat and second NT seat, and also add numerous extra electorates to every state.

The Parliament was last expanded in 1984, so it’s about time.

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