Depending on your point of view, the election of One Nation Senators to represent New South Wales and Western Australia is either vindication of Malcolm Turnbull’s 2016 Senate electoral reforms, or an argument why the system should not have been changed.
The argument against is that under the group voting tickets abolished by the reforms, parties would conspire to prevent One Nation winning Senate seats.
Famously in 1998, Labor and the Coalition swapped preferences to stop One Nation electing Senators on preferences in all states except Queensland where the new party polled a quota in its own right.
Elsewhere the major party deal resulted in the Australian Democrats electing several Senators despite polling fewer votes than One Nation.
The problem with supporting party control of preferences is that elections are about electing representatives to represent the will of the electorate, not about engineering the exclusion of certain parties.
Majority view at the time might have been that keeping One Nation out of the Senate was good, but it is a view at odds with the idea that the Senate is elected by proportional representation, and a party with significant support should have a reasonable chance of gaining representation.
Preference Harvesting
But what some viewed as the advantage of group voting tickets in keeping pariah parties out of the Senate was later undermined as preference deals became ever more complex.
A system that started as an institutionalised version of a how-to-vote card, slowly morphed into a monster with opaque preference deals that voters had no hope of understanding.
Instead of final seats being decided by voters, they were increasingly determined through preference deals arranged by preference ‘whisperers’.
The tactic of preference ‘harvesting’ developed where multiple minor and so-called ‘micro’ parties would swap preferences with each other ahead of all larger parties.
And thanks to group voting tickets, these tiny parties with no ability to campaign or distribute how-to-votes, could still deliver 95% of their votes as preferences into what became known as the micro-party alliance.
More and more parties nominated. Ballot papers hit the maximum printable width of one metre, requiring the Australian Electoral Commission to shrink the font size and issue magnifying sheets. In NSW at the 2013 Senate election there were 46 party columns on the ballot paper and 110 candidates.
Voters had two voting options, the arduous task of numbering preferences for every candidate below the line, or the easy option to select a single box above the line and adopt a party’s unknown ticket of preferences.
Understandably, many voters took the easier option. With so many parties on the ballot paper, it became hard for voters to spot the parties they knew amongst flood of new parties with attractive and sometimes confusing names.
The nadir of the system was the 2013 Victorian Senate result when Ricky Muir of the little-known Australian Motoring Enthusiast Party was elected to the final Victorian seat defeating a Liberal candidate who started the count with more than 20 times as many votes.
Such a victory would have been impossible under a system where voters completed their own preferences. Ahead of the 2016 Senate, the Turnbull government abolished group voting tickets with a reform bill backed by the Coalition and the Greens but opportunistically opposed by Labor.
Why the Senate Reforms have worked
The new system abolished party control over preferences and introduced an element of optional preferential voting. The ballot paper instructions were to number a minimum six party squares above the line on the ballot paper, or a minimum 12 preferences in candidate boxes below the line.
Elections since the 2016 reforms have seen the number of parties and groups contesting election halve, making it easier for voters to find candidates and parties they know.
But it also weakened the flow of preferences between parties. At the two half-Senate elections since the reforms, all Senate seats have been won by parties with the highest vote, or partial quota, at the start of the count.
The two victories by One Nation are the first half-Senate election since the reforms were introduced where a party has come from behind to win after preferences.
But neither victory was from a ridiculously low vote. Both the elected One Nation Senators polled more than two-thirds of the vote for their Labor opponent at the start of the count.
The flows of preferences received by One Nation may have been assisted by the party appearing in the preference recommendations on Coalition how-to-votes. The smaller ballot paper probably helped voters notice One Nation while completing preferences.
Since the reforms were introduced, preference flows between left-wing parties have been stronger than between right-wing parties.
The 2025 Senate result could be a sign that supporters of small right-wing parties are learning that preferences can be their friend rather than their enemy.
One Nation’s victory is a vindication of the 2016 Senate reforms in ensuring that Senate elections were decided by the votes and preferences of voters, not the opaque deals of preference traders.
One Nation’s victories have denied Labor two extra Senate seats, Labor finishes with 28 seats rather than 30, but 28 plus 11 Greens still gives the two parties a Senate majority before having to deal with the other ten members of the cross bench.
But it may not be good news for the Coalition. One Nation now holds four seats that were previously Coalition seats.
The Coalition may have to work more closely with One Nation than some of its members would prefer, a situation that may exacerbate internal tensions within the Coalition parties.
Thank you X 1000 . Well by the amount that got Ricky Muir in! I was so curious about this and now I understand.
This is a vindication that preferential voting should be abolished altogether in the upper house and lower house. I have not met a person Labor or liberal that agrees with preferential voting.
COMMENT: Really? You must be asking the wrong people because there’s little pressure from elected representatives.
Both the Whitlam and Hawke governments proposed the introduction of optional preferential voting. The Coalition was rabid in its opposition.
Since Labor started to win elections on preferences around 1990, it has lost its previous principled interest in optional preferences. But parts of the Coalition once so firm in supporting for full preferential voting are slowly undergoing a damascene conversion to supporting optional preferential voting.
Rather than simple majority voting, it is more likely there could be a shift by the Coalition to supporting optional preferences, as already applies in the Senate. It is already agreed Coalition policy for state elections in NSW and Queensland.
government ptoposed to
Conversely I don’t know anyone who thinks preferential voting is a bad idea.
Our electoral system is the envy of the rest of the world, Raymond.
I want to offer my thoughts, be gentle…!;) ALP would have won this current election as the Libs stunk it up… But why should someone win a majority in primary vote by 5% or more, yet lose by a handful of votes? Is that fair?
One thing I do not like is I am forced to vote for some parties that do not align with me. It works for either side of the political spectrum. In my area, we had 9 candidates, with 4 being from parties I would rather not have to give them a place. I feel it should be a minimum of 4 or 5, if you want to add more, have at it.
Doesn’t seem democratic to me. Also, somehow if the senate sheet was easier to deal with, we might get better results… Too many idiot parties that do not even get 1% of the vote! And I think ballots should be numbered.
COMMENT: The number of columns on Senate ballot papers has halved since the 2016 reforms were introduced.
The loudest voices for changing the voting mechanisms usually comes from the parties that are currently losing. Mainly because they feel that they are getting less than they deserve some how. In almost every country there is always rumblings after every election, even in the US after the 2020 election that was “rigged…” It would not surprise me if the Coalition Came To Jesus on electoral reform as they are getting squeezed mightily from the left (teals + losing votes to Labor/Greens) and the right (National agitation, minor right wing parties). With the Coalition primary vote now down under 32% and not gaining much on preferences one could easily see them wanting a change. The problem, though, is that I am not completely sure that they would do much better with a Senate system in the House or with a “savings plan.” With the exception of One Nation, the Coalition has done poorly on preferences from the larger minor parties and if preferencing became options after say the 4th position those votes would more than likely end up wasted. The problem the Liberals have now is not an issue of process but of where they stand in the electorate. Without rummaging through the ashes of the campaign I am not sure any electoral system used in the Western world would have produced a substantially better result for the Coalition at this election.
Going left is not helping the coalition gain voters. They are losing them. What’s the point in having two main parties on the left, and fringe radical group teals and greens to back them up? That doesn’t sound democratic to me – more like statism.
Hi Antony
Speaking of the Senate, do you know the historical basis for section 39 of the Commonwealth Electoral Act, which seems to be a slapdown of Queensland specifically? Is it related to the Night of the long prawns, or some other Joh era scheme?
I’ve been curious about it for a while but haven’t found too much unfortunately.
COMMENT: It is a provision that overrides paragraph 2 of Section 7 of the Commonwealth Constitution.
To give more detail, that section of the Electoral Act was passed in the late 70s or early 80s in response to Joh publicly considering using the section of the constitution that Antony mentions in order to gerrymander Senate elections in Queensland. This would have heavily favoured the Nationals over the Liberals and Labor, and Labor was able to convince the Fraser Government to pass new legislation to prevent that. I don’t know the history behind why that section of the constitution only gave the Queensland Parliament this power in relation to Senate elections though.
I agree wholeheartedly with your conclusions in this article. Preference harvesting and group voting tickets were an anti-democratic abomination. One former Senator openly boasted of establishing additional micro-parties with arresting names simply in order to channel preferences. Now the Senate results in WA and NSW have shown that the switch to optional Senate preferences for the 2016 election was never a device for eliminating minor Parties: minor Parties with a substantial voter base can prevail and electors will allocate preferences to Parties that they actually recognise and favour.
Antony Green is a genius. He makes complicated issues easy to absorb with his brilliant pen.
Optional preferential voting in the House is a mistake.
Too many votes end up exhausted on the minor parties, which means that the actual “function” to select the candidate that is the combination of votes for them as the “best” (first preferences) and “least worst” (second and subsequent preferences) would be lost on exhausted votes.
Of course, some candidates and parties would prefer that votes exhaust rather than go to another candidate.
“supporters of small right-wing parties are learning”, you say? 🤔
COMMENT: And the rest of the sentence reads “that preferences can be their friend rather than their enemy”. Extracting parts of sentences devoid of context to make dumb jokes is boring.
Hi Antony,
Let me say first that you’ve been a delight to watch for all these years, and your insight will be missed.
That said, I’m not sure that I can wholly agree with your opening contention. Looking at these results, I wonder if we just have a new future problem. In both WA and NSW, the One Nation candidate was “elected on exhausted votes”. This is not inherently problematic, if the exhausted votes represent a true non-preference between the remaining candidates. However, I find myself questioning how many truly do, as versus those which represent a non-preference between the major parties, without consideration of the subsequent effects. If the exhausted voters had been asked directly, One Nation, or ALP, how many would have genuinely expressed no preference?
Obviously, this is impossible to know. However, considering the margins, with the majority of exhausted votes flowing via Legalise Cannabis, I think it is conceivable that this “effect” produced results that weren’t necessarily considered by those who caused it. I concede, my concerns may not ever amount to a problem that needs solving. However, if they do, as far as I can discern, the addition of an optional negative, or “veto” vote would provide a systemically elegant fix, albeit one which might grossly complicate voting instructions.
Anyway, these were just some thoughts I wanted to express. Thanks for providing the forum, and for everything else over the years.
COMMENT: The only way around it is to bring back ticket voting with full preferences which we abandoned because of the way it was rorted. Or we make full preferences compulsory and put up the informal vote. There are fewer exhausted preferences in 2025 because we are finally getting smaller ballot papers from which voters can make more sensible preference decisions. 95% of voters mark six preferences. In some of these states there was a final race between Labor and One Nation and some voters had exhausted their preferences without making a choice between Labor and One Nation.
Do you think it would be possible to change the senate to be more similar to systems in other countries, were if you get x% of the votes above a threshold you get x% of the seats and you wouldn’t really have to worry about preferences. Or would it be too difficult because it’s elected by state and and too get something of that level you would need to dramatically increase the number of seats.
COMMENT: There are six seats per half-Senate election. The sort of system you describe isn’t practical.
I think the big problem with this counting system is that candidates elected 1 to 5 need a full quota while candidate 6, as in WA and NSW, gets in short of a quota. The system needs to be changed so that all candidates are elected with a full quota. A simple fix would be to press the button once, if all six reach quota, count over, if not, press the button again but set the quota at the number of votes the final candidate was elected with in the first button press. This way, all candidates will require the same number of votes to be elected.
COMMENT: Any attempt to do that will advantage the candidate from a ticket that has already elected members. In the case of NSW if you reset the quota, that would mean the two already elected Labor candidates would have a surplus that would add to the third Labor candidate and advantage them in the race against One Nation candidate who is still trying are short of a first quota. If you lower the quota it is always the parties with multiple quotas that are advantaged.
There is no question that these reforms have worked. It has reduced the size of the ballot papers, whilst preserving the preferential voting system.
What started out as an attempt by the major parties as means to isolate minor ‘extreme’ parties, backfired on them as the minors worked out how to game the system to their advantage. What was good enough for the major party “goose” turned out to be even better for the minor party “gander”.
Repealing GVTs has spoiled the game and has certainly reduced the size of the ballot paper. But do you have any reason to believe that FPV ATL, as opposed to the current OPV ATL, would cause the number of groups to expand again?
Have you done an analysis on the vote exhaustion, in part or full, of these senate contest? To my mind vote exhaustion is the biggest argument against OPV. FPV is required for House voting. As noted by Ron (above) FPV allows you to discriminate between bad and worse with at least the remnant of your vote, which many would like their vote to do, but probably do not understand this nuance.
You said that repealing GVTs was “opportunistically opposed by Labor.” What is the rationale for this statement and is this why Victoria is the only jurisdiction in which it still operates?
Speaking of which, I too had written some VB software within Excel to graphically simulate the count in Victoria in 2004, which you stated as your “proudest moment” on ABC TV. I reverted to your published calculator, for later elections to save having to load in the registered GVTs each election, but when I needed it most for the last Vic state election it threw an error, albeit that I tried it on a number of diferent devices. Do you plan to make it available again for the next Vic election? I’ve long lost my software :(.
COMMENT: Labor supported the recommendation to abolish GTVs that came out of the Joint Standing Committee on Electoral Matters after the 2013 election. Two years later, after a change of opposition spokesman on electoral matters, switched and opposed the change. Abolishing tickets was supported by all previous Labor shadow ministers on electoral matters as one of them, Gary Gray, made clear during the debate. I would note that Labor in both SA and WA has since adopted the original JSCEM proposal without bothering about the minimum 6 preferences that the government was forced to adopt when Labor backed out of the fully optional model.
On the question of exhausted votes, I note that you have done some analysis on your post on the Senate contest in SA.
“Overall 30, 730 votes have exhausted preferences representing about 0.18 quotas worth of votes.”
Whilst this may not of closed the gap of 0.09 quota between ONP and ALP, the point is that it could be significant in closer contests.
COMMENT: Until the current system there were no exhausted preferences. This was achieved prior to 1984 by accepting a double digit rate of informal voting. Between 1984 and 2013 we did it with group voting tickets but that distorted the proportionality of the electoral system in a way that could not be produced by voters expressing their own preferences. The proportionality of the Senate system has been restored by the 2016 changes and I think that putting up with some exhausted votes is preferable to the two previous systems that forced full preferential voting.
I was a scrutineer the first federal election after the reforms were enacted, for a minor part that did okay however I did have to speak to the polling manager on more than one occasion to remind his staff that they had to say at least one to 6 boxes not just one to 6 boxes. At my polling place this year they did seem pretty fastidious about it which was pleasing to see.
As always thank you Antony. Undoubtably the 2016 reform was good for reducing informal votes & encouraging voters to engage in the system. It reduced the ballots paper from a table cloth to a hand towel.
Interestingly a bit of discussion occurred amongst HTV booth workers (yes LNP, Labor & even green HTV workers do talk a lot amongst themselves constructively on booths during the day), that six preferences might be too many in the senate and maybe 4 would be good enough. It was a view expressed about a distaste for preferencing ON by LNP HTV booth workers, but, who felt inclined to follow the HTV card recommendation because they did not know anybody else who they would put safely in their six.
I think that it might also be time to reform the House of Representatives as to preferential voting in the Lower house. Although I personally prefer optional preferential, including because it maximises formal votes, that is not what this post is about.
In many lower socio-economic electorates informal votes doubled or worse this year. Look at Werriwa & Whitlam for example. There seems to be a co-relationship to the number of candidates. The more candidates on the ballot paper, the more informal votes especially in lower socio-economic area. (a sort of mini table cloth effect?) Whitlam is a good electorate to observe this because there is a geographic socio/economic divide which demonstrates the truth of this proposition – eg Barrack Heights
It will not be long (ie by say 5 May) that well healed cynics, & even others like Clive Palmer, will have noticed the potential chaos that can be caused to particularly the formal labor vote in low social-ecomonic strongholds by running many, many candidates. Much cheaper than spending $80 million on HTVs & advertising.
The answer is not to limit the number of candidates who can contest or to strike ridiculous deposits. But the answer may be to review compulsory preferential voting in the Lower House to introduce part optional preference voting like the Senate of say 6 votes to be a formal vote.
(Overseas the number of candidate/size of the ballot paper is already being deployed as I have suggested is likely above. This was perhaps one reason why the conservative leader lost his seat in the last Canadian election being opposed by 91 other candidates, albeit in a first post the post vote, in an attempt to highlight the need for electoral reform apparently).
Isn’t it ironic that the Senate is now more represenative of the voters who elect them than is the House of Representatives?