Effective Vote Spreading: Labor's Hidden Hero At The 2022 Federal Election

This article is about two incorrect narratives about the 2022 election.  I commonly see false claims on social media that Labor only won the 2022 election - not just in majority but at all -because of Green and teal preferences.  These claims are made by certain right-wing posters, mostly of the silly and Trumpy variety, who seek to delegitimise the result because the primary vote winner didn't win the election and the winner's primary vote was unusually low.  But there is another narrative that is more mainstream, which is that while Labor's win was fair enough, the Coalition was hard done by in seat terms because of its seat losses to teal independents.  I show here that that narrative is not really true either, and that the real reason the Coalition's seat share was so bad compared to Labor's was that its vote was poorly distributed in the classic two-party seats.  Most of this article is very numbery so it's been graded Wonk Factor 3/5.

Of course, Greens voter preferences do greatly benefit Labor, and had Green voter preferences split 50-50 Labor would not have won ten seats that it did, and who knows who would have governed in that mess.  But Green preferences favouring Labor is simply part of the scenery, and some other parties' preferences assist the Coalition.  The Coalition only "leads" on primary votes because it is a coalition of two parties that, after decades of fighting each other in some states, choose to mostly work together instead of wasting resources competing everywhere.  Labor and the Greens could sort out their differences and make a similar arrangement if there was any strategic point in doing so, but in their case there currently isn't.  

As concerns teals (whether they won or not), while their 2PP preferences heavily favoured Labor, in most seats where they ran that did not help Labor since Labor failed to make the final two.  This included seven seats that teal independents won, and six where independents who were generally teal-adjacent made the final two but lost.  Yes there were some seats where such candidates were cut out and the contests finished as classic Labor vs Coalition contests, and yes teal preferences helped Labor in those.  But Labor mostly didn't win those anyway (Boothby is one they did win), and there is not a single one where Labor won but would have lost had the teal voters' preferences split 50-50.  In strategic terms the teals were a nuisance to the Coalition, forcing them to fight a second front and making criticisms that may have driven votes to Labor in other seats.  In terms of votes actually polled, however, all they did was take six seats from the Coalition in an election it had already lost outright.  Labor won 72 classic seats where it did not need an edge on their preferences, plus five seats where the Coalition was excluded in lopsided Labor vs Greens contests.

In terms of converting 2PP leads in non-classic seats (where someone who is not a major party candidate makes the final two) it's true that the Coalition lost nine of their 2PP wins to the crossbench (eight independents and Bob Katter) to Labor's seven (four Greens, Andrew Wilkie, Dai Le and Rebekha Sharkie).  But that's the same gap as 2019, when the Coalition lost four to Labor's two.  Also in those nine seats, the Coalition's average 2PP margin was only 5.2%, compared to 9.9% to Labor in Labor's seven.  Furthermore, in non-classic seats that the majors won, the Coalition's average 2PP margin in its won seats was 12%, but Labor's in its was 25.6%.

Ultimately the seat split in the 27 non-classic seats was actually unfavourable to Labor.  Off primaries in these seats of 33.0% Coalition, 22.3% Labor and 44.7% for others, the 2PP preference flow (67.6% to Labor) was so high that Labor won the 2PP in these seats 52.54-47.46.  This is not surprising as nine of the non-classic seats had very high Green votes and there were also a lot of inner-city teal seats where there were strong flows to Labor (partly off the back of apparent strategic voting) and not a lot of primary votes for One Nation or UAP.  Yet despite winning the 2PP in these seats overall, Labor only won five of them to the Coalition's six, while the crossbench won 16.  This left 124 classic seats and Labor needed to win 71 of them for a majority.

Given that Labor won the 2PP in the non-classic seats by slightly more than its overall 52.13%, Labor hence needed to win 57.3% of the classic seats off 52.04% of the 2PP vote.  That's far from impossible (because single-seat systems tend to magnify even small leads in a seat totals sense) but it's no sure thing either.  

Labor overachieved in classic seats

In fact, Labor went into the election in already decent shape on paper in what turned out to be the 124 classic seats of 2022.  Since the Coalition still held the six 2022-non-classic seats it would lose to teals in 2022, the two it would lose to Greens and six it would hold against independents, it hadn't needed that many of the rest to win a narrow majority in 2019.  As a result of this, in these 124 seats the majors went into the 2022 election with a 62-62 notional split, despite the Coalition having an average 51.3% 2PP in them.  

The average 2PP swing in these seats in 2022 was 3.42%, just shy of the overall 2PP swing for the election.  By that swing as a uniform swing in these seats, Labor would have gained Bass, Chisholm, Boothby, Higgins, Braddon, Reid, Swan and Longman and missed a majority by one.  But in practice swings are not uniform, and as the last four were on margins only a whisker below the swing, Labor would have most likely not done that well.  In fact, Labor did better - it missed Bass, Braddon and Longman but won five seats higher up the tree: Robertson, Pearce, Hasluck, Bennelong and Tangney.  

Preferences helped in that Labor won seven classic-2PP seats from behind, but that was actually a low number by recent standards, and a reason for this is that the preference flow in the classic seats (59.6% to Labor) was actually quite modest.  Despite trailing on primaries 34.8-36.3 in the 124 classic seats, Labor actually led on primaries in 65 of these seats to the Coalition's 59.  Throw in the five seats where Labor easily beat the Greens and Fowler where they led on primaries but lost and Labor led on primaries in 71 seats overall compared to the Coalition's 73 (the other primary vote leaders were Wilkie, Haines, Steggall, Katter, Sharkie and two Greens).  That the Coalition would only barely if at all have governed in a parliament of primary vote winners shows how absurd it is to imagine the Coalition would have won under first past the post, where many Greens and Labor voters would have voted strategically and several seats would not have been won by the 2022 primary vote winner.  

There's a narrative that the Coalition got the harsh end of the stick in the non-classic seats by losing a bunch of safe seats to teals but as these figures show that's not really true.  In fact they won one more of the non-classic seats than Labor did despite losing the 2PP in them by more than the overall result.  The key here is that the seats that the Coalition lost to teals used to be safe 2PP seats but teals or no teals they're mostly not safe anymore, because of demographic realignment away from the right in affluent seats.  Five of the seats won by teals in 2022 became 2PP marginals in the process.  The two seats the Greens took from the LNP became Labor 2PP seats.  Warringah and Indi became Coalition 2PP marginals and Mayo became a Labor 2PP seat.  North Sydney very nearly flipped on 2PP (there appears to have been an uncorrected error such that the real 2PP should have been 50.4-49.6) and it may well have even been won by Labor had Kylea Tink not contested it.  At the end of the 2022 election there were only two seats the Coalition didn't occupy but where they won the 2PP by more than the marginal threshhold: Mackellar and Kennedy.  

Of the 124 classic seats, the median Labor 2PP was 1.63% above the mean, and this resulted partly from a pattern in which the 2PP swing to Labor was larger in Coalition seats and ALP marginals than in Labor's safe seats.  In part, this represents the general realignment where once very safe Labor seats are moving away from the party demographically, but it also raises the question of why the Coalition did even worse than the national 2PP predicts in the middle-ground seats.  

This graph shows what happened with the distribution in the classic seats.  It's simply an ordering of their ALP 2PPs from highest to lowest.

I've put a thin grey line over the top of most of the graph to show that the steps downwards are mostly pretty flat except that Labor is dropping off quite slowly in its 0-10% range (which means winning more seats).  At the far right of the graph there is a sharp dropoff where the Coalition wastes 2PP margin in its nine safest 2PP seats (Maranoa, Gippsland, Mallee, Parkes, Barker, New England, Farrer, Riverina and Lyne).  

In comparison, this is what the graph looked like after Labor's last majority in 2007:

In this case, there is a similar but less pronounced tail of lopsided seats on the Coalition side but there is also one on the Labor side, and Labor slightly underachieves in its own 0-10% range.  This would hurt it in 2010 where the 2PP in classic seats was 50.02 to Labor but the Coalition won two more classic seats than Labor did, although Labor should have had an advantage based on personal votes.  Four of the lopsided Labor seats shown here have since become Labor vs Greens seats, but most of the rest are the Watson/Chifley/Calwell/Scullin type outer suburban seats that are now transitioning away from Labor (and in some of which Labor faces potential Muslim voter problems at the upcoming elections).  Since 2007, Labor's spreading of its two-party vote in the seats it wins has become more efficient and the Coalition's has become less.  A more extreme version of this pattern was seen in the 2022 Victorian state election.  

The story of the election then is that Labor was clearly preferred by voters to the Coalition both in the non-classic and the classic seats.  It was actually in the classic seats, not the non-classics, where Labor did well and the Coalition badly in turning 2PP vote share into seats.  

No, Labor was not fished out of the slop of a probable loss, nor even helped as much as in the past, by the Greens and teals.  No, they were not lucky to get a majority (beyond that we don't have proportional representation which would have made them have to work with other parties).  No, the Coalition wasn't ripped off in seat tally terms by the teals.  No, the Coalition would not have won under the discriminatory crud that is first past the post.  No, lopsided seat shares for modestly performing inches-left-of-centre parties would not go away if we got rid of preferences (as we have just seen in the UK where Labour won 63% of seats off 33.7% of the primary vote compared to Australian Labor's 51% off 32.6%!)  

The fact is that the Coalition was soundly beaten everywhere in a way that suggests that it failed to distribute what 2PP vote it got efficiently in the classic bread and butter 2PP seats.  Padded by a handful of lopsided rural seats, the close-ish 2PP from the 2022 election flatters a Coalition performance that was otherwise about as bad in classic seats as in 2007.   This failure looks like it will be baked into the startline for the next election: by my numbers if the draft redistribution goes through largely unamended then the Coalition will need about 51.3% 2PP just to equalise the seat tally.    Perhaps the next election pattern will be radically different, perhaps not, but there is no ground for righties whinging about the 2022 election.  Their side lost comprehensively, and they should get over it and focus on why it happened.