Western Australia is a physically vast (larger than Texas and Alaska combined) Australian state unsurprisingly located in the Western part of the Australian continent. Three-quarters of its 2.6 million people live in the capital city of Perth. Only 8% of the state’s population live outside the south-western corner.
On the 11th of March, the entirety of the Legislative Assembly (State House) and Legislative Council (State Senate) is up for election. Australian Governors are not elected and have only ceremonial power under most circumstances.
Currently, the Liberal Party and National Party of Western Australia control both houses of parliament after a comfortable reelection in 2013. The Liberal party is basically a conservative centre-right party representing the interests of business. The Nationals represent primarily rural electorates and have a focus on the interests of agriculture.
However recent polling (Newspoll, Labor 54%-46%) suggests that the Labor Party, the main centre-left party, will likely win the popular vote comfortably next month. This diary introduces a model (EMWA2017) to estimate which seats are most likely to be won by each party in the legislative assembly. Western Australian state elections for the legislative assembly operate under a system of both compulsory voting and compulsory preferential voting (instant run-off).
The state has been redistributed (redistricted) since the 2013 election. As Australian redistributions are non-partisan this shouldn’t have much impact on the election’s outcome.
As malapportionment in the Legislative Council gives one-third of the seats to the 8% of the population living outside the south-west it is unlikely Labor will be able to win control of that house.
Polling also suggests that the resurgent One Nation (a very Trumpish party) is likely to win perhaps 10-15% of the popular vote. While this probably won’t be enough to win Assembly seats it could help or harm the governing parties enough to throw the model out in the rural seats to some degree. I can’t do much about this other than note the possibility as there is no historical data to work with. Finally, this model is calibrated for left-right match-ups. In seats where the Liberals and Nationals are competing with each other (Kalgoorlie, Pilbara, etc ) the model just awards the seat to the incumbent party.
The table in this diary is fairly straightforward; the first column contains the names of each seat colour coded to the party currently holding them (Red=Labor, Blue=Liberals, Green=Nationals). Hillarys is held by an independent but as he was elected as a Liberal I’ve left his seat blue pending an election win without the party label.The percentage in the middle column represents EMWA2017’s estimate of the likelihood of either the Liberal or National party winning the seat.
If the public polling is accurate Labor are almost certain to win Government with a median of 37 seats in the 59 seat parliament.
Rather than reinvent the wheel, if you are interested in detailed seat profiles I’d suggest heading over to Ben Raue’s Tallyroom.