The election of Wilson Tucker from the Daylight Saving Party at March’s Western Australian election has become the catalyst for abolishing group voting tickets in Western Australia.
Mr Tucker polled 98 votes or just 0.2% of the vote in the vast Mining and Pastoral Region. His low vote is not surprising as four referendums over five decades have shown little support for daylight saving in this vast region covering the state’s most remote areas.
Anyone familiar with how to engineer results using group voting tickets knows the system can elect parties with little support. But even I, with two decades of covering the perversity of elections using group voting tickets, find myself startled that such an egregious distortion of the electorate’s will could be constructed.
It is the most magnificent example of preference harvesting yet achieved by well-known preference ‘whisperer’ Glenn Druery. It is the crowning glory of his art, but will also be the death knell of the group voting ticket system he used to achieve it.
The back-story to Mr Tucker’s election gets even weirder. Tucker left Western Australia three years ago and has been working as a software engineer for Amazon on the other side of the Pacific Ocean in Seattle. It is a better paid job than his new four year position in the WA Legislative Council.
That is assuming, in this period of pandemic, he can get a flight back, is allowed entry to Australia and can cross the Western Australian border. (Update: I’m informed Mr Tucker has arrived back ready to take his seat.) Tucker’s term is due to begin on 22 May. If he is unable to return and vacates the seat, a re-count would create the farcical situation where his his running mate, Janet Wilson, would take his seat despite receiving zero votes at the state election.
What are Group Voting Tickets and what is ‘Preference Harvesting’.
Most people who follow my blog know exactly what I am talking in referring to group voting tickets and the tactic of “preference harvesting”. However, as overseas writers sometimes turn my criticism of group voting tickets into criticism of preferential voting, I need to explain both terms.
Most countries with proportional representation allow parties to put forward candidate lists or voting tickets. Open list systems allow voters choice of candidate within party ticket. Closed list systems do not and parties determine the order of election of their candidates.
Group Voting Tickets are similar, except because Australia uses proportional representation by single transferable vote (PR-STV), these voting tickets list candidates of the party in order, and then the candidates of every other party in order. So it is like a closed list system where a party determines the order of election for its own candidates, but also gets involved in the election of candidates from every other party.
Group voting tickets were introduced for Australian Senate elections in 1984 as a solution to the then scandalous rate of informal voting. The previous requirement that voters number every square on the ballot paper, sometimes more than 50 numbers, regularly produced informal voting rates above 10%.
The solution adopted was to divide the Senate ballot paper with a thick horizontal line. Voters were given the option of voting ‘above the line’ (ATL) for a single party box and adopting the party’s group voting ticket, that is a full list of preferences for all candidates on the ballot paper. Or the voter could vote for candidates ‘below the line’, but had to number every square as before.
Before group voting tickets, parties and candidates had never had the power to control preferences. Voters had complete control, though parties tried to influence how voters completed their ballot papers by distributing how-to-vote material. Parties that distributed lots of how-to-votes used to produce impressive rates of preference compliance, but well short of the rates we see produced by group voting tickets.
Minor parties generally distribute fewer how-to-votes and consequently had less ability to influence preferences. Before group voting tickets, preference flows from minor parties were essentially random.
Group voting tickets increased major party control over preferences and for the first time gave the same power to minor parties and independents. It turned preferences into a tradable commodity, a good that could bartered with other parties. Within 15 years, the allocation of preferences began to shift from being about ideology to being about tactics.
Which is where ‘preference harvesting’ comes in. It is about trading preferences to advantage. Over time minor and so-called ‘micro’ parties began to understand how to game the system, to swap preferences amongst each other, to add together a collection of unrelated and relatively unknown parties in the hope that one of the participants would be elected.
It required group voting tickets with complex chains of preferences that only a tiny fraction of voters would ever complete if required to number their own ballot paper. Group voting tickets could take a first preference vote cast by a voter and send it on a magical mystery tour across the ballot paper with the voter having little hope of knowing who their ballot paper might end up electing.
The tactic encouraged more parties to nominate, and also to adopt attractive party names that could be confused with better known parties or attractive causes. It was all with the aim of making ballot papers more complex and confusing, making it harder for voters to find the parties and candidates they did know amongst the vast array of unknowns. By 2013, there were so many columns on Senate ballot papers that the Electoral Commission began issuing magnifying sheets to help voters read their ballot papers.
All this herded more voters into using the above the line voting option, putting more votes into the preference harvesting pool.
It was a triumph of tactics over ideology. It turned one seat at most elections into a lottery with the prize of a well-paid and potentially powerful job for between four and six years. The price of a ticket in this lottery was an agreement to lodge a group voting ticket keeping preferences within the alliance of parties.
And few preference harvesting arrangements have come off as spectacularly as the one that has elected Wilson Tucker.
And so who elected Wilson Tucker?
The Mining and Pastoral Region ballot paper had 53 candidates distributed across 21 columns. A voter had two options for voting. They could mark one of the boxes above the line which meant their vote defaulted to the party’s full list of 53 preferences. Or they could number every box below the line from 1 to 53. Any missing or duplicated number invalidated the vote.
Given a choice between completing one preference or 53, only 2.1% of ballot papers counted as below line line votes, with 97.9% of voters opting for one number above the line. Very few parties recorded BTL rates above 5%. This was an election where more than 95% of preferences had been pre-arranged between the small number of people who negotiated the preference deals.
So high was Labor’s vote at the 2021 election that some of the preference deals came unstuck. In Mining and Pastoral Region, the Labor Party polled a second decimal point under four full quotas, the Liberal Party 0.75 quotas and the Nationals 0.72.
Which left one and a half quotas split across the other 18 groups on the ballot paper. Under every other electoral system in the world, the vote for the other 18 parties could never be aggregated together to elect a member. Under PR-STV systems used elsewhere in Australia where voters rather than parties control preferences, these 18 groups could never be aggregated together to elect a member.
But this is the world of group voting tickets. The preference deals were all pre-arranged, and once the they were applied to the votes, Wilson Tucker emerged as a Member of the Legislative Council to the surprise of both himself and the voters.
The table below has disaggregated the distribution of preferences in Mining and Pastoral Region to isolate the source of all votes used to elect Mr Tucker.
Wilson Tucker received just 98 votes, 95 as ATL votes and 3 first preferences for him below the line. He also received another 76 BTL votes during the distribution of preferences. Janet Wilson, number two on his ticket, received no votes.
Adding these votes together, it means just 174 votes, 0.35% of the formal vote or 2.3% of the votes that elected Mr Tucker, were votes cast by voters for him as the lead candidate of the Daylight Saving Party.
The other 97.7% of votes that elected Wilson Tucker were group voting tickets from other parties sent to him by preferences and the almost random order that GVTs cause candidates to be excluded. Thirteen other parties contributed to the votes that elected Mr Tucker, including 11 parties that polled more votes than the Daylight Saving Party.
Two-thirds of the votes that elected Wilson Tucker were cast for either Legalise Cannabis WA, The Greens or the Shooters, Fishers and Farmers Party. I have no knowledge of these party’s policy on daylight saving.
And look at who Wilson Tucker defeated. The National Party polled 50 times as many votes but was defeated by the Daylight Saving Party. Converting the quotas for both parties into equivalent percentages for a single-member electoral contest, it is like a party with 36% of the vote being passed on preferences by a party that started with 0.7% of the vote.
Towards the end of the distribution of preferences, there were four parties competing for the final two vacancies. These were the lead candidates of the Liberal Party, National Party, Shooters Fishers and Farmers Party (SFF), and Wilson Tucker of the Daylight Saving Party.
The graph below plots the progressive vote through the count of the lead candidates for the final four parties. The horizontal axis represents the number of new first preferences released to the preference count on the exclusion of candidates and groups.
On party totals, Labor led the count and was guaranteed four seats. The final two vacancies required the distribution of preferences to fill.
In order on first preferences, the Liberal Party sat second, Nationals third, SFF fifth, and the Daylight Saving Party way back in 18th place in a field of 21. Two of the three groups below the Daylight Saving Party directed preferences its way, and from that point on, Wilson Tucker leapfrogged candidate after candidate to reach a quota.
As the graph shows, there were two preference harvesting operations underway, one trying to elect the Shooters, Fishers and Farmers candidate, the second which ended up coalescing around Wilson Tucker.
There were five groups on the ballot paper that didn’t contribute to the Daylight Savings Party’s quota. The tiny ticket vote for Independent Anthony Fels went by a circuitous route to reach the National Party. The WAxit Party, No Mandatory Vaccination, Australian Christians and One Nation were all excluded and flowed to the Shooters, Fishers and Farmers.
On the exclusion of One Nation, SFF candidate Matt Priest led Wilson Tucker 4,387 votes to 3,896. Both trailed Liberal Neil Thomson with 5,285 and National Nicholas Fardell with 5,170.
The exclusion and distribution of Green candidate Kimberley Smith sent 2,279 Green ticket votes to Tucker, putting him ahead of the SFF.
The SFF was the next excluded, and its preferences put Tucker over the quota to fill the region’s fifth vacancy.
After Tucker’s election, three further tickets that had reached the SFF were released. 492 ticket votes from No Mandatory Vaccination flowed to the National Party, while 530 from the Australian Christians and 1,428 from One Nation flowed to the Liberal Party, electing its candidate Neil Thomson to fill the final vacancy.
The above graph and table show just how perverse the group voting ticket system is. The luck of finishing fourth last allowed Tucker to pick up two smaller group’s ticket votes, starting the sequence where he leap-frogged group after group before eventually receiving Legalise Cannabis and Green preferences to leap over the SFF, capture their preferences, and beat both the Liberals and the Nationals to a quota.
This is all great fun, but is it what voters intended? Advocates of group voting tickets argue it makes it easier for small players to get elected and harder for major parties to control the upper house.
Perhaps voters like this, even if the same voters then complain when governments can’t get legislation they want through the upper house.
But group voting tickets were introduced to cut informal voting, not to produce particular outcomes in the upper house.
Yes, on occasions group ticket voting can be engineered to act like a ‘random dude’ provision, putting some unknown candidate into the Legislative Council.
But neither the voters nor the candidate themselves can have any knowledge that this may be an outcome of the election. Though Glenn Druery always tells me these outcomes are never random nor unintended – they are carefully planned.
It’s a sign of electoral system failure that, instead of it electing members representing the will of the electorate, it occasionally tosses in some random dude engineered by a small number of participants swapping preferences without the knowledge of the public.
That’s not how elections should work.