The upcoming Western Australian state election will feature quite a change to WA elections, with the upper house being elected at large as a single 37-seat election.
This has prompted a lot of uncertainty about what might play out. But actually there is quite a lot we can know about what is likely to happen. In particular, how well the various parts of the state will be represented in the new upper house.
The old upper house had a severe malapportionment in favour of the non-metropolitan parts of the state. The state was split into six regions, with each region electing six members to the upper house. But the three non-metropolitan regions only make up about a quarter of the state’s population, yet chose half of the Council.
The new upper house didn’t simply redistribute the regions to equalise the population (listen to last Friday’s podcast for an explanation of why that wasn’t an option) but also abolished any local representation in the electoral structure.
This creates the potential for concerns about whether there will still be representation of particular parts of the state once the electoral boundaries are wiped away.
Indeed Liberal leader Libby Mettam has been playing up these concerns, yesterday saying that the regions had been “disenfranchised” by the changes, and promising to restore the old regime.
But at this point we actually know a lot about what is going to happen, and the non-metropolitan parts of Western Australia look set to have quite a large share of upper house representation – not half of the chamber, but quite a bit more than their proportion of the state’s population.
I identified candidates who seem to have a good chance of winning election, and identified which old upper house region they are connected to. It was easy for incumbent MLCs, but otherwise I used data on whether candidates are sitting councillors, or which areas the candidates contested at previous elections.
If you look at the tickets for the three biggest parties, there are some interesting trends.
Labor’s ticket is very city-dominated, barring two regional candidates in the top three. The Liberal Party appears to have used the old upper house regions for their preselections, producing a ticket with three regional candidates out of the top six. I haven’t listed the Nationals here – the top two are definitely regional, and I assume all are.
The Greens’ regional representation depends a lot on their vote. Their top three are all from Perth, but the next three in the marginal seats are in regional areas.
I came up with two different scenarios for vote shares, and thus a certain number of seats per party. The first scenario was based on the 2017 upper house election. The two-party-preferred polling right now is similar to 2017. The second scenario is based on recent polling, with the “others” vote split in a similar pattern to the 2021 election. The former was better for the Nationals and One Nation, while the latter is better for the Greens.
In both scenarios, the result produces a split of 24 metro MLCs and 13 non-metro MLCs, which equates to 35%. This is less than the 50% of the old system, but still a lot higher than the non-metropolitan share of the state’s population.
The ALP’s seat split is the most city-dominated, but it’s not actually far off the proportional share. In the first scenario, the city makes up 80% of seats. It makes up 75% in the second scenario. In reality, Perth contains about 76% of voters.
The Greens have a very different regional split depending on whether they win three or five seats.
The Liberal Party comes close to 50% non-metropolitan representation, and the Nationals are completely non-metropolitan.
The other parties have a range, but generally non-metropolitan areas are over-represented.
This strikes me as completely consistent with analysis I did about the NSW Legislative Council, which is the closest comparison to the new WA upper house.
And this all seems perfectly fine to me.
Not all voters treat geographic location as the most important factor when deciding their vote. It makes sense that some parts of the state where voters have more concern for this factor would tend to have more local representatives. Voters elsewhere can choose to prioritise other factors. The new WA upper house allows for regional representation, if the voters choose to prioritise it and parties respond to that. But it doesn’t lock it in and force it on voters. None of this is bad.
You should not represent acres or sheep but people. If the libs want more people from non metro areas they need to either poll better or preselect people from regional areas higher on their party ticket.
If looking at non metro representation then look at both houses of parliament