The Victorian lower house results showed a remarkable disconnect between voting trends and seat results. There was a clear swing against Labor – 5.8% on primary votes, and 2.4% on two-party-preferred votes – yet there was a gain of one seat for Labor.
While Labor suffered a large primary vote swing, there was no increase in the Coalition vote – instead, the minor party vote increased significantly. But thanks to the single-member electorate system, that increasing minor party vote didn’t flow through into a larger crossbench, and instead the major parties largely held their own despite the lower vote.
This has prompted me to revisit a chart I made prior to the election, showing the combined primary vote in the lower house for Labor, Liberal and the Nationals at state elections since 1979.
The 2022 result was not just a continuation of the trend, but an acceleration. No other election saw such a large drop in the combined major party primary vote. It was the worst result during this time period for the Coalition parties, and the second worst for Labor (who did slightly worse when they lost power in 2010).
Prior to the federal election, I pointed out that there is not much of a relationship between the non-major party primary vote and the number of crossbenchers elected. At the federal election, there was both a low major party vote and a large number of crossbenchers, such that this point was mostly forgotten about. But it has played out in Victoria.
Another way to see this trend is to look at the Gallagher index, which is a method of measuring disproportionality at elections. The index starts at zero when the proportion of seats and votes for each party match perfectly. Parties with highly proportional systems tend to have scores around 1-2, while non-proportional systems tend to be the ones that pass 10. The recent federal election produced a score of 16.5, the least proportional result since at least the 1940s.
The Victorian state election reached a score of 21.8.
This is not the highest I've found, but it is the highest for over half a century.
Disproportionality was at its highest levels through the 1950s, 1960s and early 1970s when the Democratic Labor Party was at its strongest in Victoria. The party regularly polled well into the teens statewide but never won any seats. Their preferences also strongly favoured the Coalition, meaning that the Coalition tended to win more seats off a particular primary vote than Labor.
For example at the 1967 election, which has the worst score, the Liberal Party won 37.5% of the primary vote and won 44 out of 73 seats (60%), while Labor with 37.9% of the primary vote managed only 16 seats, barely more than the Country Party's 12 seats achieved off 8.65%. Meanwhile the DLP polled 14.3% yet was unrepresented.
Disproportionality was then at its lowest levels in the 1980s and 1990s when over 90% of votes were for the major parties. It shot up in 2002, and while it dropped slightly over the last two decades, 2022 is the worst in 50 years.
I guess the ultimate point of this blog post is to point out that, despite winning a comfortable parliamentary majority, Victorian Labor's primary vote is bouncing along at near-record lows. A larger minor party vote manifests itself less as a large crossbench and more with greater disproportionality. I will return to the lower house in a few days to look at the non-classic races that did occur in 2022 (even if not many resulted in a crossbencher winning), but I will also be looking at the upper house outcome. Despite the distortions of group voting tickets, the overall upper house results bear far more of a resemblance to voting statistics.
This is of course a result of one-member electorates, but the elephant in the room is that as the ‘major’ party vote continues to drop, sooner or later a tipping point will be reached and the crossbench will suddenly pick up large numbers of seats and the ‘major’ parties will find themselves as what they will truly have become, large minor parties. And at that point, the one member system will start to work against them … unless they fiddle with the electoral system to prevent that happening.
I think David that you are more likely to see either the coalitions that support the parties change or a new set or parties emerge, that will see them both get back over 40%. I don’t think this fragmentation can last – and I see PR as actually being a negative not positive in that case.
I think that your analysis misses a little because you do not consider that the Greens are a “major party”. If you do then the disproportionality readjusts a little. The core issue here is the falling level of home owner/occupier which rose from 45% in 1946 to 70% in 1960 (due to Labor’s ideological failure in agreeing to selling public housing to renters) and held that level through to 2006. Since then it has declined to 64% and will be 62% by 2025 due to Howard’s capital gain discount/negative gearing policy which most LNP Members are heavily personally invested in (ie they are negative gearers if you read their Parliamentary disclosures). Each decline of 1% in home ownership carves 1% of the Liberal vote. So the rising renter voter is being shared by Labor and Greens and solidifying the Green vote at 12% trending to 15% as a major party. So the rise and fall in Labor vote will be mostly between it and Green as the Liberal vote trends to high 20 percent?
I don’t think there has been so much of a decrease in support for the major parties, as an increase in the number of candidates.
If we had proportional representation I daresay we would have an increase in support for major parties as people wouldn’t want instability.
Well the Greens are obviously not a major party, just like the DLP were not a major party when they were regularly polling 14% in the 1960s.
In theory a new major party could emerge, but this story would still be interesting to see how parties that once had close to a monopoly on the vote (over 99% in 1985) are now down to under three quarters.
I do think falling rates of home ownership explain a lot of the shifts in political trends but don’t understand the rest of your comment – what rise in the Labor vote?
The other solution would be to draw the boundaries like South Australia does, to reduce the number of “false majorities” or simply increasing the number of seats in the legislature might increase the chance of the seat count more likely to match the vote totals.
Why does Victoria only have 88/89 seats whereas Queensland has 93 with a smaller population? It is time for an increase, both chambers of parliament in Victoria can pass it and then the electoral commission can draw new boundaries in time for the next election.
South Australia stopped doing that because it didn’t work. And it didn’t try and fairly represent all parties, simply ensure that the “right” major party won a majority – which doesn’t really make sense when almost 30% is voting for someone else.
Ben, I share your feelings, but I think it’s important to note that the redistribution, on the VEC’s calculations, left Labor with 58 seats on 2018 voting, not the 55 it won on the boundaries at the time. On its figures, Labor won Bass, Bayswater, and Glen Waverley from the Liberals – but lost Caulfield (pretty borderline, true), Hawthorn & Nepean to the Liberals, Morwell to the Nationals and Richmond to the Greens for a net loss of two seats.
Luck also played a part. Labor won Northcote by 184 votes, Bass by 202 (both 0.2 per cent) and Pakenham by 0.4 per cent – the three closest ones all went Labor’s way. The next three closest – Mornington, Benambra and Mildura – all had independents as runners-up. The numbers post-Narracan will end up 56-28-4, but had the losers had won those six, they might have been 53-27-8.
I guess my serious point is that no one should underestimate how good the Victorian ALP is at protecting its own territory. We saw it in 1988 when Kennett’s Liberals won 50.5% of the 2PP, but Labor held on to its majority. The Greens saw it in the inner suburbs time & again before they finally broke through in 2014. I’ll be interested to read your thoughts, but I think Labor’s marginal seat campaigning, and its ruthless use of slush funds to win those seats were clearly factors in that one-sided win, and all other parties & candidates have to reckon with it next time.
(Ignore my pending comment above I accidentally got the email wrong)
I don’t buy that it was 58, I don’t believe Caulfield was notionally Labor. I actually believe the numbers that had it marginally Liberal at around 0.4%.
Estimates are never accurate and it is extremely hard to calculate how preferences would have flown considering neighbouring seats had different candidates.
I agree Hastings, Bayswater, and Bass were notionally switched, but the author of this website also had them switched, but did NOT have Caulfield listed as notionally Labor.
Instead of making psephologists like Antony Green and Ben calculate the margins. Why can’t the AEC publish the estimates themselves unless they already do? Considering they draw the boundaries you would think they would know because they have to gerrmander the maps so they are fair to both parties.
Mildura is exceptionally hard to say where that was on the new boundaries. But because it took some territory from Ripon and you had no National candidate in that seat in 2018, and it just combined the heavily liberal booths with the Nat booths. There was no independent or at least a high profile one run in Ripon, so the estimates for that seat should have been taken with a huge pinch of salt.
I would also say the NSW boundaries could produce a result like in Victoria. The coalition could hypothetically get a 2-3% swing against it but gain a seat or 2 in March, allot of the seats Labor needs to win are on a 6% margin or more.
Dan in nsw the situation for the
Liberal party.coalition forces is not good. I suspect the liberals and nationals are only certain of maybe a dozen seats. The rest are competive for some reason or other of course they will probably win more than that number but it is not certain. Seats like Parramatta East Hills and Penrith will probably Fall to
Labor because they are so close
Bega will remain Labor due to the candidates. I expect the greens to hold their seats although Balmain is not certain.
The Scandal plagued seats Kiama and Drummoyne are difficult for the libs and I expect Labor to win 1 of those seats.
The ex sff independents I expect to retain their seats. I expect Wagga Sydney and Lake Macquarie to retain their seats. I don’t expect Labor lose any of their current seats. Overall I expect an uneven 6% swing to Labor this gives the opv bonus to Labor not the liberals. There are more seats that are competive than the ones I predicted but these in some cases involve wins by parties other than Labor or have relatively high 2pp nominal Margins
What I would like to see how the federal figures translate in terms of state seats and visa versa after the state election. What has happened at either level of government dies not confirm a result but shows what is possible
Daniel, I believe the vec did publish their own estimates. They had caulfield and hastings as notional labor by a handful of votes.
Agree with your assessment that the coalition in nsw still has a fighting chance to retain most of their seats. I only see labor favoured to gain Penrith and possibly Parramatta at this stage, Liberals still in with a chance to hold rest of their seats including east hills.
Yoh An is correct, the margins I’ve been using are VEC margins. I explained my switch to using their margins and what I knew about how they calculate theirs, and the extra data they have that I don’t, in this blog post.
Tim, Labor has ended up with more seats than before. Yes on a redistribution basis it is less but a number of those seats are so close as to basically be a tie. The 55 seats they won in 2018 is real. My point is that if there was any fairness in the system there would be feedback between votes and seats and we haven’t seen that.
According to Adrian Beaumont curtesy of TheConversation. The Greens would have won 7 seats in the upper house if they used the senate voting system we use federally. 0 for legalise weed, 16 for Labor, and 15 for L/NP. Others would have dropped to 2 from 5.
I think it is a must that the upper house is fixed before lower house reform, GVT’s make it broken, and is not very represented. I would love to know what Labor’s position is on it, or do they support the system?
I mean it is their loss if they want less seats and a bigger crossbench, But why on earth would they continue to support a system that opens the door for extremists to get into parliament with very few votes? Western Australia has a broken system until it was changed recently, So why not Victoria?
The Liberals should campaign on changing it if they are elected in 2026, If Labor does not, It may cost the Liberals preferences from minors however many swing voter’s who hate GVT’s may consider voting for the party who want’s to see the end to it, And that time should be sooner rather than later.
Here is the link to the article from Adrian Beaumont
https://theconversation.com/final-victorian-election-results-how-would-upper-house-look-using-the-senate-system-196291
Daniel – Because the GVT system gives the party itself greater control over who gets into parliament. They don’t want to help get Greens elected, they want to keep the crossbench destabilized and dependent on the ALP’s goodwill. The Coalition have similar concerns
Labor would much rather have to deal with extremists and other single-seat players, because they’re not any sort of threat to their dominance in future, whereas the Greens certainly are.
Wilson, agree with you that Labor attempting to block the Greens gaining strength is a bad look for them. However, I feel that the ALP are managing the relationship better than the Coalition (Liberal-national intraparty relations). By distancing themselves from a more extreme partner, Labor can have better success in the ‘centre space’ and retain marginal, outer suburban seats. At the same time, the ALP still retains an ‘informal’ relationship with the Greens but not a full coalition partnership.
In contrast, the Liberals have moved too far to the populist right in trying to align with National Party views, which has hurt them in the affluent, inner suburban areas where voters prefer a more centrist brand.
I partially agree with you Yoh An, but I’d also say the Liberal Party itself has many populist right figures, and the blame cannot be laid solely, or perhaps even mainly, at the feet of the Nationals. The figures in the National Party rebelling against Littleproud’s decision not to support the First Nations Voice to Parliament shows that they’re perhaps not as populist right as has been previously thought.
This “informal relationship” between Labor and the Greens seems marked by a lot of vitriol wherever the Greens are a threat to seats held by Labor. The last Queensland state election got rather nasty in the battle for South Brisbane, as did the contest for Northcote in the recent Victorian election. And Federal Labor campaigned in inner Brisbane in 2022 with the slogan “the only way to change the government is to vote 1 Labor”, clearly aimed at scaring voters who were considering voting Greens. It may be a better look for Labor in the eyes of the public, but they’re also making future collaboration between the parties in parliament more difficult in the process.
Agree Wilson that whilst Labor is successfully distancing themselves from the Greens, they are not doing it in a ‘polite’ manner given the large number of personal attacks used against incumbent MPs and/or candidates. Maybe a strategy similar to the Coalition where Labor and the Greens agree not to contest seats against each other would work better to create a more harmonious environment.
Comments are closed.