The Electoral Matters Committee of the Victorian parliament has launched an inquiry into the structure of the Legislative Council.
The inquiry is not seeking submissions on group voting tickets, but the inquiry is related to the push to abolish GVTs because such a change would likely stop a number of small parties from winning seats. A change in the regional structure could allow for a broader range of parties to win.
The previous inquiry into the 2022 state election produced a report that endorsed the abolition of group voting tickets, but also opened the door for further discussion about reform to the current structure of eight regions of five members each. While the Labor, Liberal and Greens members of the committee endorsed the abolition of GVTs whether or not there was reform to the regions, the minor parties in the upper house have indicated that a package deal of reform to the regions and group voting tickets would be more acceptable.
I previously analysed the likely outcomes under the current region structure but without group voting tickets. A magnitude of 5 creates a reasonably high barrier to entry, and a number of small parties have been able to get around that solely due to group voting tickets. A continuation of the current region structure but without group voting tickets would see a narrower range of parties winning seats – not just Labor, Liberal and Greens, but other parties would need to do quite well to win a seat.
The committee has published a discussion paper presenting a number of options that could change the structure. It includes a single statewide electorate, either all elected at once or split over two elections with members serving eight-year terms ala New South Wales. It also suggests 4 regions of 10, 7 regions of 7, or merging the five metropolitan regions while leaving the three regional regions intact.
They also propose something I had not considered previously, where the current structure would remain intact but top-up seats would be added for parties with a certain share of the vote.
I will be making a submission, and I encourage others to do the same. I won’t try and make my full argument here, but I wanted to make a few general points.
Firstly, I think you need to consider the purpose of the 5-member regions. A low magnitude tends to limit party fragmentation and lead to a chamber with a smaller number of parties. That seems important for a house of government, where low-magnitude PR (the electoral sweet spot) can ensure that governments are held accountable and don’t have to include too many different parties. You could argue that this has value for an upper house (so governments don’t have to wrangle a bunch of separate individual MLCs but instead can have a few parties work together) but I think the value is much less clear.
The other value of the regions is to ensure local representation, and I think it fails on that front. They cover vast areas and there isn’t that much community of interest for each region. And that would be even more true if they were made even larger. If you were to remove any local representation by having a single statewide electorate, the evidence suggests regional areas with a strong identity would still have strong representation.
So I don’t think 4×10 and 7×7 really do much for increasing proportionality, and provide poor local representation. I also think the idea of having 5-member regions in rural parts of the state and a 25-member region in the city is also unacceptable, since it creates very different contests in two parts of the state – in the city, minor parties will be much stronger, but in the bush the major parties will remain dominant.
I do think there is a good case for a single statewide electorate. But if you were to make that switch, I don’t think the single transferable vote system would work for such a system. I explain my logic in this post, but in short: preferences play an important role in low-magnitude elections, but as the magnitude increases, the preferences make less of a difference, but the burden of the preferences (both for the voter and the administrators) increases. A simple list PR system using D’Hondt or Saint-Lague would be vastly more simple to administer, and much easier for voters to use. And the seat outcome wouldn’t be much difference. Indeed D’Hondt slightly favours the larger parties, which could allay concerns about the very low threshold you’d get from a 40-magnitude district.