Group voting tickets – here we go again

I feel like I’ve been banging on about group voting tickets for a long time, but with the upcoming Victorian election again featuring this anti-democratic feature for its upper house elections, I thought I needed to summarise some of the worst problems with this system.

Under the group voting ticket (GVT) system, voters can vote above or below the line. If they vote above the line, they can only cast a single ‘1’ for a party group, and the party group controls where that preference then flows.

GVTs were first invented in the 1980s for use in the Senate as a solution to a problem caused by requiring voters to mark every preference on ever growing ballot papers, leading to higher informal rates. It quickly spread to upper houses in New South Wales, South Australia and Western Australia. By the time it was being implemented in Victoria in the mid-2000s, the problems in the system had already become clear and New South Wales had already adopted a different system, which has now been adopted in all of those other jurisdictions, leaving Victoria as the only jurisdiction using GVTs in 2022.

So in this post I’m going to run through some of the worst problems with GVTs, all of which have been solved in other states.

Firstly, the GVT system produces some bizarre disproportionate outcomes. In 2018 I compared the actual results to a proportionate outcome. The Greens polled 9.3% but won just one seat, while Derryn Hinch’s Justice Party polled 3.8% and won three.

Part of the reason this happens is because only certain parties have an incentive to do deals. The Greens are large enough that they are usually there at the end of the count and it’s hard to overtake them, so most of the other minor parties have no incentive to deal with them. But they are small enough that they don’t win seats on full quotas, unlike the major parties.

Even amongst the smaller minor parties there is no particular relationship between number of votes and number of seats. Transport Matters (TMP) won a seat on 0.6% and Sustainable Australia did so off 0.8%, but another six parties polled more than TMP and didn’t win any seats. The Democratic Labour Party polled 2.1%, more than three parties which won seats, but missed out.

Secondly, the process of giving parties automatic control of their party’s preferences creates some bizarre outcomes. In the absence of tickets, parties can and do recommend preferences, but many voters choose to mark their own preferences, and it requires some effort from the party to communicate to their voter (by having a volunteer hand out a piece of paper, or by paying for advertising). With tickets, you don’t need to do any of that, you just need to nominate and attract some people to vote for you, and you get to decide where all of their votes go.

This creates an incentive for small parties to run in all regions, even if they don’t have a local presence. There’s no need for volunteers to hand out how-to-votes, but by running in multiple places you can swap preferences with other parties, hopefully getting the best preference flows in the region where you are strongest. Without such complete control of preferences, this incentive goes away, and has thus resulted in smaller Senate ballot papers since GVTs were abolished.

(Incidentally this allows us to pinpoint which of the small parties has the best chance of winning in a particular region, because usually all of the other parties in the preference-swapping alliance have ranked them highly. Here are my predictions for the 2018 Victorian election and the 2021 WA election).

Ticket votes all flow to the same location, which creates a lopsided effect. It becomes very easy for candidates to overtake other candidates, and thus candidates with tiny votes can gradually overtake their rivals until they have an election-winning share of the vote. See whose preferences ended up with the Daylight Saving party when they won a seat in Western Australia last year.

When voters control their own preferences, this doesn’t happen. Since preferences scatter amongst numerous other opponents, you can usually only overtake a rival if you are relatively close to them in the first place, and candidates with a tiny vote don’t snowball. Antony Green assessed how this phenomenon was completely eliminated at the 2022 federal election.

And while minor parties have strategic incentives to do deals with other small parties, including those that no-one has ever heard of, leaving Labor, Coalition and the Greens to the end of their preference, real voters’ preferences do not follow this pattern. Rather, a voter for a particular minor party will usually then preference the larger parties which are ideologically closer (so Animal Justice voters tend to favour Labor and the Greens, and voters for right-wing minor parties tend to give preferences to the Coalition), and parties with minimal public platform don’t receive many preferences. There is no “minor party vote” – there are a variety of ideological voting patterns, some of whom choose to vote for a small party or a big party.

The same is also true of major party voters. Labor (particularly Victorian Labor) has a habit of preferencing small parties, including some quite right-wing ones, ahead of the Greens, but usually their voters immediately preference the Greens. Steve Fielding wouldn’t have been elected in 2004 if Labor voters had directly decided whether they would preference him or the Greens.

As an example of how voter behaviour doesn’t follow the party line, this blog post from Antony Green shows the low rates of voters following the how-to-vote card in Western Australia for the Senate in 2019.

GVTs also add a lot of complexity to the system and make it much harder to understand, even for professional election analysts.

Under voter-controlled preferences, usually the candidates closest to the quota are the most likely to win, and it requires a remarkably strong preference flow for someone to overtake and win. But when whole parties’ blocs of voters can be transferred en masse, it allows for a lot more potential outcomes, and it is much harder to predict where your vote might end up. When pretty much every party makes preferencing decisions that don’t fit with their political principles, it’s impossible to be sure your vote might not end up electing someone you oppose.

When a party on a tiny vote can parlay a slim lead over another party on a tiny vote into an election-winning position, slim margins become very important. When voters control their own preferences, it’s rare that these orders of elimination matter. This became most obvious in Western Australia in 2013, when a change of about a dozen votes was enough to change the result for not one but two Senate seats. This immense pressure to get the result exactly right was too much for the AEC to handle, and they ended up losing ballots and forcing a re-election for the entire state, but would be far less likely under the new Senate system.

Sadly it’s too late for Victoria to fix its electoral system for this election, but thankfully for individual voters they do have a choice. It’s quite easy to vote below the line. I’ll be back tomorrow to explain how to do it, and why everyone (yes, everyone) should do it.

I will also be returning to this topic with some other analysis over the next four weeks, but this post is here as a summary of why this system is bad.

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