 |
| Huge if true, but ... |
Recently my attention was drawn to an article on the Canberra Declaration website by one Dave Pellowe. Pellowe is better known for being on the receiving end of a later dropped 2024 Queensland anti-discrimination complaint over comments he made about Indigenous religious beliefs while explaining why he would not include a Welcome to Country in a Christian religious event. (I'm vaguely curious about what exactly he said, as my home state has a long history of potentially interesting complaints like this being made then almost always dropped - but not curious enough yet to find it.) The Canberra/Daily Declaration site was better known to me through the involvement of one Julie Sladden, an anti-COVID-vaccine retired doctor and right-wing culture warrior who was bizarrely endorsed by the Tasmanian Liberals for not one but two state elections.
I'm not sure I'd come across Pellowe talking nonsense about Senate voting before but this is not the first time he's done it. His Twitter bio reads "Solomon prayed: Give me an understanding heart so that I can (steward democracy) well & know the difference between right and wrong. 1 Kings 3:9". I hope that he will see this article and realise that having an "understanding heart" to "steward democracy" requires understanding the facts and consulting reliable sources rather than just former UAP and One Nation figure Lex Stewart. Stewart and Pellowe have been making very similar complaints about Senate reform and the 2016 election, and they're both wrong.
The fundamental claim Pellowe is making is that the current Senate voting system greatly advantages the Greens over the minor right because minor right votes exhaust and as a result minor right parties fail to catch the Greens who, according to Pellowe usually win the final seat because they have "the biggest fraction of a quota".
But even the most fleeting look at post-reform Senate results shows this is false. Since Senate reform, the Greens have only won five of the 24 final state seats (the others being won by the Coalition six times, Labor four times, One Nation four times, Jacqui Lambie Network twice and Family First, Liberal Democrats and UAP once each). The five cases where the Greens won the last seat were their second seats in WA and Tasmania in the 2016 DD and the final seats in NSW, Qld and WA in 2019. The reason they haven't won any since is that as the major party primary votes have fallen further at recent elections, the Greens now tend to be fifth on primaries and not sixth.
Looking at these actual cases, four were not even remotely close. In WA 2019 the Greens easily crossed quota during the exclusion of the HEMP candidate so they would have won even had every right wing voter whose vote exhausted put them last. In WA 2016 the Greens with .887 quotas led the WA Nationals with .644 quotas when the latter were excluded, and had rapidly increased their lead over said party during the distribution. In NSW 2019 the Greens with .958 quotas at the end of the distribution defeated One Nation with .677 quotas. In Qld 2019 the Greens with .958 quotas defeated Labor (not even a right wing party) with .768 quotas. In these three cases, had there been full voter-directed preferencing the Greens would have easily got quota, since it is not only right-wing preferences that exhaust (and even some right-wing preferences flow to the Greens, especially when the alternative is Labor). The only case of a right-wing party losing to the Greens that can be attributed to right-wing (particularly Liberal) votes exhausting was Tasmania 2016 where the Greens, starting with a 1.451 to 0.334 quota lead over One Nation, ended up getting home by just .005 quotas (141 votes). With the increased flows between right-wing parties in recent years, I don't think that would happen anymore.
Even that singular case of what Pellowe falsely claims to be normal was not because the system was set up to advantage the Greens - rather the previous Group Ticket Voting system was starting to unfairly disadvantage them. What Pellowe calles "the power for voters to delegate decisions about their vote to their preferred political party" was in fact a system that in its later years coerced voters to delegate their decisions, because the only alternative was numbering every candidate box below the line. An arduous process for even a highly engaged voter, with a high risk of an unintended informal vote. And that system wasn't rewarding ideological parties that assigned their preferences in a principled manner, since there was little reason for most other parties to deal with them.
Rather for minor parties to succeed in such a system they usually had to do pragmatic deals with parties that were problematic for them and hope those deals came off. Parties running on principles, especially from the left, tended to suffer more of a backlash for doing so and in some cases had their campaigns wrecked by it. Parties like the Greens that tended to poll high primary votes were perversely not worth doing deals with since they wouldn't get excluded and therefore wouldn't have any preferences to pass on to make them work trading with The result tended to be networks of unrelated random single-issue parties doing deals to try to gang up on the Greens and major parties and win seats off risible primary votes with preference flows nothing like what voters would have chosen for themselves. How anyone claiming to be a Christian thinks that this was in any way better than what we have is way beyond me, and Pellowe's nostalgia for this nonsense sits poorly compared to other fellow travellers who have seen it for the venal system that it was and opposed it.
The reason we have this semi-optional preferences system is that once Group Ticket Voting was clearly broken beyond repair, the alternative for Senate voting was either to allow some exhaust or to go back to the other nightmare that existed pre-1984, of all voters having to number every box with national informal rates often into double digits. Allowing some exhaust is the least of the three evils.
It is true that the Greens do well out of the current arrangement - but that's not because of the differences between semi-optional preferencing and an unworkable full preferential system. Partly it's because the granular nature of contests for six or twelve seats per state eliminates a lot of minor parties with no support, which tends to cause all of the Coalition, Labor and the Greens to outperform their vote share slightly compared to what would happen with a single national PR contest with an extremely low threshhold. Also it's because half-Senate elections with six Senators per seat are a sweet spot in the Greens' ability to get enough vote to win one out of six seats everywhere off about a tenth of the vote. If the Senate is increased in size, they aren't likely to win extra seats, so this advantage will decrease.
Pellowe also engages in apologism for, believe it or not, the Labor Party, who he claims "opposed the changes, arguing the changes were designed to “entrench” the power of the major parties and disadvantage minor parties — and it did exactly that." In fact Labor filibustered the Senate for a whole day with nonsense against the new system, making embarrassingly false claims and predictions about the level of exhaust and the impact on minor parties. Many ALP Senators argued that Senate reform would produce a crossbench confined to the Greens and Nick Xenophon candidates. In reality the 2016 DD elected four One Nation Senators and Senators from each of Family First, Liberal Democrats, Derryn Hinch Justice Party and Jacqui Lambie Network. Subsequent half-Senate elections have continued to elect One Nation (this year with three seats), the UAP in Victoria in 2022 and JLN. While the Coalition did seem to think the system would be better for them and worse for the crossbench than it was, the crossbenchers elected from 2016 on were elected by voter intention and accountable to the voters who elected them. At least this is true of those who stayed with their parties and didn't get disqualified under Section 44, but that's another story.
The level of exhaust
Pellowe claims that "7.52% of Senate ballots were exhausted" in 2016. Throughout his complaints about the system he seems unaware that Senate votes do not all carry the same value through the count, and thus a vote may contribute most of its original value to electing one or more Senators, with a partial vote value assigned to it later exhausting. Thus the number of ballots that ended up contributing some value (often very low) to exhaust in 2016 is a lot higher than 7.52%, but the number of ballots that exhausted at full value, contributing nothing to electing anyone, is lower.
There is also an issue with the 2016 election that some of the officially recorded exhaust happened during irrelevant preference throws at the end of counts. In the case of WA 2016 mentioned above, exhaust rose from 4.67% to 8.16% during the exclusion of Kado Muir (WA Nationals). But Muir finished 13th, and at his exclusion the result was already decided with Rod Culleton (ON) and Rachel Siewert (Greens) winning the final two seats, so the Muir throw was completely irrelevant to the result. Taking out these cases, the rate of meaningful national exhaust rate in 2016 was 5.08%. Similar rates of meaningful exhaust have been recorded since; 4.8% in 2019, 5.7% in 2022, and a remarkably low 4.0% in 2025 - even despite the rise in the minor party vote over successive cycles!
Pellowe makes more claims such as that 88% of all exhausted votes gave their first preference to a non-major party candidate and that 25% of non-major votes "simply vanished". No source is given for these claims and I greatly doubt Pellowe, Stewart or anybody else involved has the ability to perform the very complex work involved correctly, but the second probably rests on misunderstandings of the point about partial exhaust above, and/or other misunderstandings. In NSW 2016 for example, 14.5% of non-ALP/Coalition/Greens votes did not number any winning candidate and 12.2% did not number any winning candidate nor the final loser (those votes would have materially exhausted at full value). In Victoria the figures were 18.1% and 11.5%. The total rate of full material exhaust would be higher than the material exhaust figures I give there because of cases where the voter included an elected candidate in their vote but the elected candidate had crossed the line before the vote could reach them, but I doubt it would be that close to 25%.
Pellowe also talks about the supposed 25% having their vote "die before it could help elect a non-major party Senator", endorsing the fallacy that voters for minor parties would, if they made a choice, overwhelmingly prefer other minor parties to the majors. Many of those who do make a choice do send their preferences to major parties. All this adds up to an absolute whopper in the subheader that "millions of conservative votes quietly died — benefiting the Greens more than anyone realised." In fact the total vote value of all 2016 exhaust including meaningless exhaust was only 1,040,865 votes, many of those votes weren't conservative, some of that exhaust was partial value of votes that had helped elect somebody, and nearly a third of the exhaust was after the results had been decided.
Amid all Pellowe's nonsense against a fine and necessary improvement to an electoral system that was being gamed to death by preference harvesting, there is one thing that I agree on. This is that the most effective way to vote in the new Senate system is to number every box (for most voters above the line is fine). I've written about this at length. It is not, however, desirable or necessary to scare voters with conspiracy theory nonsense and junk psephology in order to convince them to do this. The system isn't broken - the old system was, and Pellowe's objections to the current system are.
Stewart's and Pellowe's previous articles
Pellowe's Spectator article, which I missed as it came out on election eve, claims that Lex Stewart is a "psephologist and advanced mathematics nerd". Stewart also calls himself that, minus the "nerd". Those of us who are psephologists with advanced mathematics skills (nerd status optional) don't know the guy in this capacity. Some of us do know him as someone who makes pitiably weak fraud accusations against the AEC via videos that contradict his own submissions. Besides the Spectator, the most prominent source I can find calling him a psephologist is the so-called "Cairns News". Even his own profile on the Daily Declaration doesn't call him a psephologist; it calls him an engineer.
Pellowe's article mostly piggybacks off Stewart's, which makes this bizarre claim:
"This happened in 2016, when the Greens, with only 8.7% of the vote, secured 12% of the Senate seats. Meanwhile, the combined votes of all the minor parties (right-wing, centrist, and others), which were 20% of the electorate, well over the required 14% quota, missed out due to the ‘exhaustion’ of votes not numbered beyond six."
Missed out? How precisely is four seats for One Nation, one for LDP, one for Family First, one for Hinch, one for Lambie and three for Nick Xenophon Team "missing out"? For sure it's a lower proportion of the seats than the combined primary vote for minor parties, but that's because a person who votes for one minor party won't in general prefer a minor party from a different flank to a major party closely aligned to them, so preferences between excluded minors don't flow anywhere near 100%, and nor should they. Indeed, preferences from minor parties often flow more strongly to the majors, Greens and One Nation than to other minor parties (even similar ones). All that, rather than votes exhausting, is the main reason minor parties combined have a lower seat share than vote share in the Senate.
He goes on to claim that in 2019 the results ignored the non-Green minor parties (actually they won a seat in each of Queensland and Tasmania, for what that's worth) and that " If many voters had numbered beyond six, a non-Green party would have been elected to the final Senate seat instead." As I have noted above this is not true - the Greens hit quota in one of the states where they finished sixth and would easily have done so under full voter-directed preferencing in the other two.
Stewart then also goes on airbrushing the ALP's disgraceful 2016 defence of preference harvesting by claiming that Hansard shows "ALP MPs did not name the Greens but stated that the new method gives a “medium” party an unfair advantage." Perhaps Stewart was only looking at the Reps debate but if so he should have known the place to look was the Senate where even in the lowlights package I linked to above, most of the Labor Senators spouting utter bunkum against Senate reform had a go at said minor party.
Stewart's final paragraph entitled "Justice and Fairness in Voting Systems" is ridiculous. He argues (without the slightest evidence of intention, and none of effect either beyond examples that I've debunked above) that the system was deliberate rigging in favour of the Greens and against voter intention. Voter intention had far less to do with the preference harvesting era in which vast ballots effectively coerced most voters into voting 1 above the line and assigning their preferences to their party which would deal them to parties near and far in an attempt to win a seat. This could result in unknown and unscrutinised parties winning seats off a fraction of a percent of the vote, absolutely nothing to do with what voters were thinking. When voters had to choose their own preferences, the facade of these fake near-100% preference flows between parties collapsed. Stewart even invokes "transparency" in implied defence of the old Senate system in which seats were determined by a combination of luck and operatives doing harvesting deals behind closed doors, and in which Senate preference allocations could be deliberately obfuscated to make it hard for even the most expert observers to tell what a vote would do!
What would be closer to the point on Stewart's part is introspection about why right-wing preferences can exhaust. I suggest that he should know. The main reason exhaust is a problem for the minor right is because Australia has too many unnecessary minor right-wing parties, making it possible for pointless spats between them to turn into a How To Vote card saga whereby Trumpet of Patriots leaves One Nation off. This isn't a sign of a bad system; it's a sign of the minor right being maladapted to it through its own pointless splintering and political self-indulgence. But even taking that into account, his claim that it is causing the Greens to win seats still isn't true.
Did the old system help minor right parties?
The new Senate system is OK for minor right parties if they actually get votes. One Nation won three seats this year, two by overtaking Labor on preferences. UAP won a seat in 2022, and there were all the wins for right minors in the 2016 DD.
But the idea that the old Senate system was good for minor right parties needs a closer look. One Nation were disadvantaged by other parties dealing them out in 1998, meaning that they only won one seat off 9% of the vote, losing to the Democrats in two states where they were ahead on primaries in the race for the final seat.
Not counting Palmer United, whose position in 2013 was populist and not really clearly right-wing (unlike the later UAP), the remaining right-wing wins included David Leyonhjelm (Liberal Democrats NSW 2013) off a perfect storm of a great draw on a monstrous and confusing ballot paper. With or without group ticket voting, changes since have prevented such a case happening again. Then there was John Madigan (DLP Vic 2010) who won off 2.3% of the vote by using the Group Ticket preferences of a pile of minor right parties to just get over the Coalition then win a seat that should have gone to a major party (nothing to do with the Greens who were in with quota). The other two cases were Family First in Victoria 2004 and SA 2013.
But both of these were not because of the choices of right-wing voters but because of bad Labor and Democrat GTV preferencing decisions in 2004 and bad Labor GTV preferencing decisions in 2013. (The Greens also made the same bad decision in the case of SA 2013 but their decision didn't cause the outcome). With voter-chosen preferences, even without exhaust, Family First would not have got near winning either of these seats, which would have been won by the Greens in 2004 and Stirling Griff (NXT) in 2013. It's not clear whether the Greens would have won their seat in SA 2013 under voter-chosen preferences or whether it would have gone to Labor.
The minor right didn't deserve any of these post-1998 seats and wouldn't have got any of them under voter-directed preferences with manageably sized ballots, with or without exhaust. But as the One Nation 1998 case shows, Group Ticket Voting wasn't necessarily good for minor right parties if they actually got votes.
The Daily Declaration claims Senate reform is a threat to freedom. The opposite is true; it was about freedom. The previous system did not give voters the freedom to effectively choose preferences for themselves. It only gave most voters the effective freedom to surrender their preference to their party's whims, and the parties had shown that almost none of them could be trusted with those preferences. It is poor form for these supposed Christians to be bearing false witness against an excellent reform and those who fought for it.
