This was originally published on the 11th April, 2017, and was republished here on the 4th January, 2022.
The final results are in for the Western Australian state election, an election in which conventional political wisdom was meant to return in full. Instea, it decided that it quite enjoyed being on holiday and just kind of phoned it in, delivering Labor not just a win — like it was supposed to — but one of the most crushing victories in Australian electoral history.
After his failed half-bid to become the leader of WA Labor all of a year ago, former Foreign Minister Stephen Smith stated that ‘no-one believed the 42% Newspoll figure’. No-one except, as it turns out, the voters of Western Australia, who 42% of whom did indeed vote for Labor. Not only did this mean Labor won the outer suburban seats they were expected to win; not only did this mean they won the inner suburban seats that were a possibility, but not guaranteed; it also meant that they won regional seats by orders of magnitude, and the translation of their primary vote to the upper house meant that they can now expect to have a much easier time getting legislation through than previously thought.
The Liberals have been left with thirteen of the thirty-one seats they previously held, mostly in the western and central southern suburbs of Perth. The scale of the Liberal destruction in the regions was out of the blue, for in seat after seat they have lost around 20% of their primary vote compared to the last election, when it was expected that their vote would actually hold up more outside of Perth than it would inside of it. They’ve even come fourth behind One Nation in Central Wheatbelt.
Their Legislative Council results are even worse. Historically, votes for major parties in WA elections have been similar across both houses, though often slightly higher in the LC. This time around, Labor’s vote continued this trend — though it was slightly lower, rather than higher — but the Liberal vote was a mere 27%, a 4% drop on their already dismal Assembly result and a 20.5% swing from their 2013 Council result. This has meant that they have lost their otherwise guaranteed influence of the Upper House, losing half their seats there.
With a loss this size, the WA Liberals have come into contention for the most prestigious electoral award on Earth, the WAELOAT (henceforth known as the Whale Oat)! That’s the Worst Australian Electoral Loss of All Time.
To be a contender for the Whale Oat, you must do as follows: you must be the major governing party and lose an Australian election — state or federal — with a loss in your primary vote of 10% or greater. The more votes you lose, the better. Seats lost are taken into account, but isn’t as important as vote share lost.
Here is a list of all the incumbent governments to lose an Australian election with a primary vote reduction of 10% or more:
Victoria, 1908: Reform -33.94* (Liberal) Queensland, 1957: Australian Labor Party -22.32* (Queensland Labor) Australia, 1931: ALP -21.74* (Lang Labor) South Australia, 1933: ALP -20.86* (Lang and South Australian Labor) New South Wales, 1901: Progressive -20.04* (Retained government) Northern Territory, 2016: Country Liberal Party -18.80 Qld, 1904: Ministerial -18.14* (Retained government) NSW, 1920: Nationalist -17.82* (Went to new party) Tasmania, 1982: ALP -17.46 Western Australia, 1911: Min -16.82* (Stood multiple candidates in many seats, did not contest in ten others) Vic, 1955: ALP -16.50* (Anti-Communist Labor) WA, 2017: Liberal -15.88 Vic, 1952: Liberal and Country Party -15.84* (Holloway Liberals) Qld, 2012: ALP -15.59 NSW, 1941: United Australia Party -15.57 Qld, 1989: Nationals -15.55 NSW, 1932: Australian Labor Party (NSW) -14.89* (Lang Labor) WA, 1971: Lib -14.34 NSW, 2011: ALP -13.43 Tas, 1996: Lib -12.91* (Entered as minority government) SA, 1997: Lib -12.41* (Retained government) Tas, 1925: Ntl -11.93* (Wasn’t governing party by time of election) SA, 1979: ALP -10.78 WA, 1914: ALP -10.42* (Twelve uncontested seats, retained government) NSW, 1988: ALP -10.27 SA, 1993: -9.72* (Entered as minority government)
Most of this list is asterisked, because there were circumstances around the vote that explain away the large vote loss. Many of these are due to the creation of new parties that are rebelling in some way against the government, and end up splitting that vote. Some are due to being a minority government at the time that voters clearly wanted to get rid of, and the remainder are parties that remained in government despite the declining vote, and therefore clearly do not count. Removing those results leaves us with this list:
NT, 2016: CLP -18.80 Tas, 1982: ALP -17.46 WA, 2017: Lib -15.88 Qld, 2012: ALP -15.59 NSW, 1941: UAP -15.57 Qld, 1989: Nat -15.55 WA, 1971: Lib -14.34 NSW, 2011: ALP -13.43 SA, 1979: ALP -10.78 NSW, 1988: ALP -10.27
We can continue to cut results here, but notice first that this decade has more elections in this list than any other, which is particularly noteworthy given the 1990s and 2000s do not appear on the list at all.
The 1988 NSW election can be the first to go: the ALP government lost 15 seats in a 109 seat chamber, which isn’t really that many, and also held more seats than the Coalition in the upper house, despite a 9% swing against them. The 1979 South Australian election can go for much the same reason, as can the 1971 Western Australian election, though that was affected by malapportionment which meant the Liberal government lost barely any seats despite an enormous vote drop. The 1982 Tasmanian and 2016 Northern Territory election can be eliminated as well, because the size of their electorate is too small, though the NT election gets kudos for the total destruction of the government, who were left with two seats. This leaves us with five elections to choose from, for there can be only one Whale Oat.
WA, 2017: Lib -15.88 (& -20.56) Qld, 2012: ALP -15.59 (-) NSW, 1941: UAP -15.57 (-) Qld, 1989: Nat -15.55 (-) NSW, 2011: ALP -13.43 (& -15.54)
New South Wales, 1941: United Australia Party -15.57%
The United Australia Party had become a ironic name by the early 1940s. The latest in a long line of anti-ALP parties to emerge in pre-WW2 Australia, the UAP found electoral success easy-going in the early post-depression years, coming into federal government in 1932 and staying there for the next decade. In that same year they also entered government in New South Wales and Victoria, and while the latter was undone by infighting between their coalition partners and became the minor party in their coalition government, in NSW they, like at federal level, lasted until 1941.
That year, Robert Menzies was forced to resign as federal leader and, therefore, as Prime Minister, due to internal unhappiness about his leadership. This had also been true at state level, and there was a general feeling of directionlessness in the party. The NSW Premier, Bertram Stevens, had been replaced as leader in 1939 by Alexander Mair, only one year after Stevens had led the party to victory. Labor, meanwhile, was riding high, with Jack Lang finally being deposed as leader after splitting the party in the early 1930s. His replacement was well-regarded barrister William McKell, who decided to put forward a comprehensive ‘master plan’ for the state during his campaign.
Mair, by contrast, argued that no major changes could be implemented until after the end of the war, contributing to the image of directionlessness, and his coalition partners in the Country Party stuck with him on this. The result was an absolute drubbing for the UAP/Country coalition, losing a combined 33 seats out of the 59 they held. These seats didn’t all go to Labor though, as the aforementioned splits meant that some candidates ran as Independents and won, but bear in mind also that of the fourteen seats the UAP did win, four of them were uncontested. The swing against the Country Party was quite small — only just over 2% — yet they lost ten seats, which they gained back at later elections. The UAP, on the other hand, folded after losing 23 seats on a 15.5% swing, getting a mere 20.29% votes and having no elected upper house to potentially lessen the blow of their loss.
This was, by any measure, a terrible loss. 20% of votes, even in a state that had a large regional population, is dismal for a supposed major party. At the previous election, Labor and the UAP both had about 35% of the vote, which is actually a better performance by the UAP, because they were only running in metropolitan seats, whereas Labor ran across the state. This time around, the UAP lost significant ground in Sydney, leaving them with nothing to fall back on.
Queensland, 1989: National Party -15.55%
If you ever need an example of a political fall from grace, look no further than the end of the National government of Queensland in 1989. After 32 years in government, and the last six of those without the Liberals as a coalition partner, the Nationals fell apart, deposing their long-serving leader Joh Bjelke-Petersen in 1987 after his failed bid to move into federal politics.
Queensland’s political history probably deserves its own article, but it is littered with governments reworking the electoral system to suit their own interests. The early parliament was dominated by businessmen, and thus Legislative Council was unelected, serving as an impediment to any party forming government without supporting their pro-industry policies. The Labor Party was such a party, and once it entered government in the early 1900s it set about finding ways to get rid of the Legislative Council altogether, thereby giving it free rein to do whatever it wanted. It succeeded in this task by filling the appointments to the Council with Labor members, who then voted the chamber out of existence in 1921.
From then until 1949, the Legislative Assembly was formed out of equally distributed districts. Labor was in power for all but three years from 1915 to 1949, but this is largely due to the timing of the one, single term Country and National Progressive government, which was elected in early 1929. The Great Depression hit not long after, and Labor won government back in 1932 as a result, though the CPNP had tried to minimise the possibility of this happening by reducing the number of seats in the chamber by ten, with Labor-held seats being the main target. The troubles within the CPNP and then the UAP after them meant that the ALP had little to worry about until the retirement of their leader, William Forgan Smith, in 1942. Following this, the Labor government decided to introduce first-past-the-post voting, replacing contingency voting and meaning that they would benefit from a split in the votes of the opposition coalition.
In 1949, Labor chose to increase the size of the chamber by thirteen seats. However, these districts were malapportioned to favour Labor, as the greater distance a seat was from the capital, the less voters it would need to have to constitute a seat. It just so happened the Labor was the predominant party in the regions, and so they gained the majority of these new seats. First-past-the-post and malapportionment were so advantageous to Labor that it took a split in the party to take them out of government in 1957, with their leader Vince Gair being expelled from the ALP and forming his own Queensland Labor Party. FPTP now worked against Labor, splitting their vote between the ALP and the QLP and allowing the Liberal/Country coalition to form government, though the malapportionment towards the regions meant that the Country Party had more seats, and therefore their leader became the Premier.
The Coalition immediately readjusted the zoning and added another three seats to the chamber for the 1960 election, and then reintroduced preferential voting for the 1963 election, ensuring that three-cornered contests would no longer work against them. Queensland Labor dissipated in time, but Labor had lost its electoral advantage in regional areas and now had its support concentrated in the major cities. As a result, it continued to increase its proportion of the vote during the late 1960s and 1970s, but was barely able to make a dent in the Coalition majority. After a new redistribution for the 1972 election, Labor achieved a vote share of 46.75%, yet won only seven more seats than the Country Party, who had a flat 20% of votes. Another Electoral Act change occurred in 1977, which benefitted the Nationals (as they were now named) at the expense of the Liberal Party, with the former hoping to turn the latter into a South-East Queensland-only party.
They achieved this goal in 1983, which took place after the Liberals tore up their coalition agreement with the Nationals, who had been increasing their grip on government at successive election under Bjelke-Petersen. The Nationals won the election, governing for three years with the support of two Liberal MLAs due to not quite having a majority in the chamber, and in the process adding another seven seats to be contested at the next election which, naturally, were predominantly regional seats that they could win. They did exactly that in 1986, winning nineteen more seats than Labor despite getting 2% less votes than them, and could now govern in their own right.
However, success had got into Bjelke-Petersen’s head, and he decided to immediately launch a ‘Joh for Canberra’ campaign, doomed to fail from the start and resulting in his being replaced as party leader. The Nationals were not able to recover from this, looking very much like the three-decade government that they were, and received an absolute drubbing at the election. Once again, though, redistribution had worked against a government trying to retain its hold on power, as the Nationals had lost their votes to Labor, and their old coalition partners were now so enfeebled that they could not hold back the tide: the Nationals won 27 seats with 24% of the vote, while the Liberals won 8 seats with 21%.
The severe malapportionment of the Queensland Parliament at the time means that we can’t really get a sense for how badly the Nationals lost, because their earlier victories were based on the same fallacy. However, the fact that Labor was able to overcome this mountain at all — though it was partly due to the way the Bjelke-Petersen government had gone about trying to achieve an outright Nationals majority — gives some sense of the scale of the victory, particularly as the Nationals were only able to overcome a similar mountain because the Labor Party itself split in two. Compared to this, the split of the Coalition — especially given the return of preferential voting — was not so dramatic, and was even to the Nationals’ advantage for a time.
New South Wales, 2011: Australian Labor Party -13.43%
Having been in government for ten years without a change in leadership, the NSW Labor Party decided to change its leader — and, therefore, the Premier of New South Wales — three times in a four year period in the lead-up to the 2011 NSW state election. Each time, the leader was deposed due to poor polling, and the change in leadership stopped the rot for a small amount of time, before it would continue heading south for Labor. Did people dislike Labor, the changing of leadership, or both? One could argue that this election is unique amongst the five on offer, because it is not as bad as it could’ve been for the governing party. After all, it’s the only one of the five to have a swing against the party of less than 15%.
But this hides the fact that, in terms of seats lost, this one was shocking for Labor, and came only four years after a relatively strong performance, though not quite as strong as the UAP in 1939 before their enormous loss in 1941. Labor won only 20 seats out of 93, losing 32. Most of these losses occurred in in the south and west of Greater Sydney, with only a thin strip remaining in Labor hands, from the city centre to the inner-western suburbs. In the Upper House, too, Labor did poorly, winning only five of the 21 seats on offer in the proportionally elected Legislative Council. Their proportion of the vote in both houses is a much depleted 25.55% in the Assembly, and 23.7% in the Council, though one wonders if in earlier times the Greens’ 10% vote share would’ve mostly belonged to Labor.
Queensland, 2012: Australian Labor Party -15.59%
The newly elected Labor government after the 1989 election, with the support of the Liberals, decided to end the generations of malapportionment and re-introduce ‘one vote one value’ seat distribution. Thanks to the scale of their victory, the introduction of this legislation actually had very little effect on which seats each party won in 1992. Labor spent all but two of the years between 1989 and 2012 in government, and did not lose an election at all in this time. Meanwhile, the old Coalition had decided to go one step futher than being in coalition again, and actually merged in 2008 to form the Liberal National Party. At the 2009 election, there were warnings for the Labor Party, for the LNP had very nearly matched them in terms of votes, but was unable to get a large enough swing in enough seats to force a change of government.
Three years later, the Labor government was facing certain electoral defeat. Polling had heavily advantaged the LNP for over a year, and 21 years is a long time to be in government. Furthermore, Bob Katter had launched his Australian Party, which seemed likely to eat into the Labor vote in the regions moreso than the LNP vote. Combined with the slim margins with which they were holding on in South-East Queensland, everything was pointing to a heavy defeat for Labor. Somehow, it managed to be even worse than that for them, as they suffered a 15.6% swing against them, and lost all but seven seats in the chamber. The new LNP government, by contrast, had 78 seats.
The mind-boggling seat difference is in some ways the opposite of the other Queensland result, as it makes the swing look more impressive than it actually was. The Bligh government had held onto a swathe of seats by relatively small margins at the last election, so even a 5% swing would’ve looked like a landslide in seat tallies. The emergence of Katter’s Australian Party — polling 11.5% of the vote at their first election — turned the result from a bad one to a terrible one for Labor, as they lost quite a number of votes to the party and, due to the existence of optional preference voting in Queensland at the time, many of those votes didn’t come back at all. The overcorrection at this election was rectified three years later, where there was a swing against the LNP large enough to create a hung parliament.
Western Australia, 2017: Liberal Party -15.88%
So, how does the Liberal Party’s ignominious defeat compare with its rivals?
Of its four opponents, it seems to have the most in common with the Queensland 1989 election. Like then, it involves a former party of the Coalition governing on their own and being decimated, while their former partners only suffer minor losses. However, the Nationals governed Queensland for a far longer period of time and were therefore more vulnerable to a big loss. On the other hand, the distribution of seats was greatly in their favour, even moreso than Western Australia was before ‘one vote, one value’ legislation was introduced, which meant that Labor had a significant hurdle to overcome in regional Queensland that they did not have to in Western Australia, where they could rely on simply winning Perth.
It has something in common with NSW 1941, in that it was the defeat of a governing coalition by the ALP, and the size of the loss of seats for the UAP and the Liberals (at least in the Legislative Assembly) is comparable, as well as the loss of primary votes. It also shares with the two most recent results (NSW 2011 and Qld 2012) the fact that it is a recent Australian election in which there has been an enormous swing against a sitting government, for which there is now such a pattern that it seems as though we quite confidently state that there appears to be something going wrong within the Australian political system, and voters, being forced by law to vote, are responding by veering in one direction and then the other more heavily than ever before.
Primary vote share actually works against WA 2017 being the worst loss, in that it is the only one of these five in which the governing party had a vote share greater than 30%. However, I believe this is due to the proportion of Western Australians who live in Perth being much greater than their equivalents in New South Wales and Queensland, and thereby meaning the Liberals have a larger base (and, conversely, the Nationals a smaller one) than in the other states.
However, there are two important factors that I believe makes the 2017 election the Whale Oat: the Legislative Council result, and the length of government.
Proportionally, the lower house seats that the Liberals have been left with is not the worst of the lot. Queensland Labor had a miserable 7 out of 89 in 2012, a mere 15% of what they had prior to the election. By comparison, no other party lost more than 70% of their seats. On balance, the Liberals’ Legislative Council result matches this, having lost just under half of their seats in that chamber. The thing is, that is still much worse than in any of the other elections.
This is because three of those elections (both of Queensland’s along with NSW 1949) did not have elections for both chambers. This leaves a comparison only with NSW 2011, in which the Labor government had a swing against them of 15.4%, 2% larger than in the lower house. From this, they won 5 of the 21 seats on offer (or just under 25%). The WA Liberals, on the other hand, have had a swing in the upper house of nearly 21%, winning 9 of the 36 seats on offer (or bang on 25%). Why is the latter worse than the former? Because the structure of the chamber is different.
In New South Wales, the entire state makes up one electorate from which the upper house is elected using proportional voting, so a party has to make sure its vote holds up across the state in order to win the same number of seats at each election. In Western Australia, there are six regions which each elect six MLCs at each election, so even if your vote collapses in, say, South Metropolitan, if it holds up everywhere else, you will only lose one or two seats.
The Liberal vote did not hold up anywhere. Interestingly, the results are almost the inverse of in the lower house, where it was in the regions that they were losing 20%+ of the vote. In the upper house, that’s happened in the three Perth regions, along with South West, in which they were equally dismal in both houses. The fact that voters were so willing to break away from their voting pattern — as mentioned earlier — of voting for the same party in both houses makes this a uniquely awful result. A 20% swing against a sitting government is unheard of in Australia, even moreso for it to happen in a metropolitan area.
Furthermore, that is a 20% swing against a sitting government that has only been in for two terms. The two other recent election landslides involved a four term, sixteen year government in New South Wales and a five term, fourteen year government in Queensland. The WA Liberal government didn’t even make it to nine years before getting dumped. This is actually the kind of result you would expect a popular first-term government to get in their bid for a second term, like the Liberals in 2013, rather than at the election of a new government. It’s actually hard to see how Labor can better the result, such is the scale of their victory.
To sum, the 2017 Western Australian election saw a major governing party lose over half their seats, record an unparalled drop in their primary vote in the Legislative Council to go with their unusually bad Legislative Assembly result, and it is difficult to see how they could’ve had a worse result, even though they had only been in government for two terms. For these reasons, I proclaim that the WA Liberals are the new holders of the Whale Oat — the Worst Australian Electoral Loss of All Time. Congratulations to them, and long may they reign.