The upper house nobody modelled

Victoria's Parliament has two chambers. The Legislative Assembly is 88 single-member districts where voters number every candidate and preferences are distributed until one reaches 45 seats, a majority; the outcome is decided by a dozen seats that turn on a few points. The Legislative Council is 40 members elected five at a time from eight multi-member regions by proportional representation. A majority there is 21 of 40, and the government of the day rarely holds it.¹

The Council can no longer block supply — that power went in the 2003 reform — but it retains the power to block and amend every other bill. It is the chamber where minor parties and crossbenchers concentrate, and the place a new government's programme can be held.

What group voting tickets did

From the 2006 state election, an upper-house voter could mark a single ‘1’ above the line for one party. Every preference after that was allocated by the party's lodged preference order — its group voting ticket, applied to a single above-the-line mark — which the voter never saw. The voter surrendered control of their own preferences to the party they numbered first.¹

That system produced a market. Preference brokers assembled ideologically unrelated microparties into reciprocal tickets: Glenn Druery reportedly charged firms around $5,000 to join a ticket block and around $55,000 if a candidate won a seat. Victoria was the last jurisdiction in the country still using the system. New South Wales abolished it in 1999, the Senate in 2016, South Australia in 2017 and Western Australia in 2021.¹

The result was that the leading candidate in a region could be beaten by a candidate polling a fraction of their vote. In 2018 in Eastern Metropolitan, the Greens led on 9.00% and lost the seat to Transport Matters on 0.62%. In South-Eastern Metropolitan the same year, the Liberals led on 12.30% and lost to the Liberal Democrats on 0.84%. Antony Green's measure for this — leading candidate's vote divided by trailing winner's vote — reached 14.52 and 14.64 in those two contests. Without group voting tickets it rarely exceeds 1.5.²

Defeated leader ratio, 2018 Eastern Metropolitan: 14.52²

What replaced it

The Electoral Amendment (Miscellaneous Matters) Bill 2026 was introduced in the Assembly by Anthony Carbines on 28 July 2026, deemed urgent, and passed the lower house the next day without amendment. It went to the Council on 30 July, where Ingrid Stitt led for the government, and passed 33 votes to four on the night of 11 August. Royal assent followed on 18 August and the change took effect on 19 August — the day after assent.¹

Both major parties and the Greens voted for the change. The four votes against were the four members whose parties had been built on the system being abolished.³

• Georgie Purcell, Animal Justice — said the change would “significantly decrease” her chances of re-election, and that “it's not over until it's over”.³
• Rachel Payne, Legalise Cannabis.³
• David Limbrick, Libertarian.³
• Jeff Bourman, Shooters, Fishers and Farmers.³

From this election, an above-the-line voter must number at least five groups in order of preference, and a below-the-line voter at least five candidates. The only between-party preferences that count are the ones the voter writes. Within a group the order is fixed by the party, which submits it at registration, and a ‘1’ for a group flows to that group's lead candidate and then down its list. Saving provisions adopted from the Senate keep a ballot formal where a voter numbers fewer than five but marks a valid first preference; a cross or tick in a square counts as a ‘1’; and a ballot marked on both sides is taken as below-the-line only.¹

What did not change matters as much. The eight regions, the five members per region and the 16.7% Droop quota — the share that guarantees a seat: one divided by one more than the seats in a district — are all untouched. Altering the region structure would require a referendum, which a parliamentary inquiry recommended for 2028 or 2030 but which was not part of this Act.³

The quota, and the seat that decides it

In a five-member region the quota is 16.7% of the formal vote. One quota is 16.7%, two seats is 33.3%, three is 50%, four is 66.7% and five is 83.3%. Above 16.7% a party wins a seat; below it, under the new system, it almost certainly does not.²

The quota decides who is in the room

In most regions Labor and the Coalition each win about two seats. That leaves the fifth — the hinge — to decide who holds the balance of power. Under the old system these hinge seats were where preference harvesting paid, and Green's evidence is that they were used to keep the Greens out. Applying the new rules retrospectively to 2014, 2018 and 2022, he finds there would have been 5, 7 and 6 fewer minor-party members respectively. Across the three elections, non-Green minor-party seats would have fallen from 22 to 4, with the Greens rising from 10 to 19, Labor from 47 to 50 and the Coalition from 41 to 47.²

Legislative Council seats, actual against no-group-voting-tickets, 2014–2022
Election | Count | ALP | L/NP | GRN | Others
2014 | Actual | 14 | 16 | 5 | 5
2014 | No GVTs | 15 | 18 | 7 | —
2018 | Actual | 18 | 11 | 1 | 10
2018 | No GVTs | 19 | 14 | 4 | 3
2022 | Actual | 15 | 14 | 4 | 7
2022 | No GVTs | 16 | 15 | 8 | 1²

What it is likely to produce

Antony Green's expectation under the new system is a Coalition on about two seats per region, so about 16; Labor reduced to one or two per region, so roughly 10 to 12; the Greens up to six, one in every metropolitan region and possibly one regional; and One Nation on eight or nine. If the Liberal Party forms government after the election, even with a lower-house majority, it will not have a majority in the upper house and will have to rely on One Nation there.³

That reverses the usual crossbench dynamic. A Coalition government at roughly 16 seats plus One Nation at roughly 8 reaches or passes the 21 votes needed to work the chamber. Labor plus the Greens at roughly 19 or 20 falls short of it. One Nation, polling in the 22–25% band, is above one quota in every region and is the one party that gains under the reform rather than losing to it: it was rising on first-preference votes, not on preference deals, and Green's view is that it would elect a substantial bloc with or without group voting tickets.

The microparties are the other half of the story. Animal Justice, Legalise Cannabis, the Libertarian Party, Shooters Fishers and Farmers and the Democratic Labour Party all held seats built on preference harvesting at 16.7% quotas. Those members have said they supported abolition, but only alongside a structural reform that scrapped the eight regions for a single statewide electorate — which would cut the effective quota to about 2.4% and would need a referendum. That reform was not in the Act.³

Unsupported block: pub.terrain.quote

Green has estimated the reform would have cost the two major parties seats in all three previous elections as well as the microparties, and the reason is arithmetic that works through the whole region rather than any one contest. The abolition removes the mechanism that let a candidate win from below 1%, but it also removes the mechanism that manufactured a majority in the chamber for whoever could assemble the most tickets. Nobody has run the 2026 Council on a published model: the forecasts here are expert expectations from a party-level read of the polling, not seat-by-seat simulations, and the range across them — One Nation at 7 or 8 on Green's read, as many as 16 on Glenn Druery's — is wide.

Three things hold across every credible read of the new system. No party wins a Council majority on its own. One Nation becomes a substantial bloc and is the most likely holder or co-holder of the balance of power. The microparty crossbench as it existed after 2022 is not recoverable at a 16.7% quota. The per-region projections going into the campaign are estimates, not counts: no votes have been cast under the new system before 28 November 2026.