Noel Turnbull’s blistering critique of commentary on the recent by-election leaves us asking (like Julius Sumner Miller) ‘Why is it so ?’
Perhaps we should start by recognising (with all respect) that political journalists are primarily there to generate entertainment: stories which their readers will find interesting, persuasive, and perhaps helpful. The trouble is that by-elections are not very interesting, and generally not very significant. Some participant has departed the scene or is about to, and a replacement must be selected; usually, it is fairly clear who the replacement will be.
So journalists have to find a perspective – an ‘angle’ – from which the by-election will become interesting. It helps if it is one of the few seats where the outcome cannot be confidently predicted, and in Eden-Monaro, there is the extra spice of the belief that the party which wins the seat would win government. (That this had not been the case in the last two elections does not make the idea less interesting.) Some see a historical inevitability at work. Some want the by-election to be seen as a barometer, testing how ‘the voters’ feel about the Prime Minister, or about the issues that the candidates talked about in their campaigns. Other commentators are more interested in the impact that the result might have on the standing of the party leaders.
Underlying these perspectives is the perception that the election is an aggregation of individual choices, which reflect judgments about the process of government: the election outcome tells us what ‘the voters’ think about it. The evidence for this is not very strong: voters do not seem to be very interested in what candidates are taking about, and are unlikely to explain their vote by reference to this. Voters may have a high opinion of the PM but not vote for his party. And most voters, rather than choosing which candidate to support at each election, consistently vote for the same party. Their voting intention is less a choice about programs than a form of self-identification, largely reflecting socio-economic position, which is (usually) renewed at each election, or (less commonly) renegotiated. Changes in the party vote in any seat are more likely to reflect changes to the boundaries, which have moved particular sorts of voter into, or out of, the electorate, than changes in the political orientation of individuals.
But journalists will continue to talk in terms of this narrative of the vote as a choice on the basis of voters’ concerns because it is ‘a good account’ of the election: it makes it interesting, and significant, and is part of the democratic narrative of ‘government by the people’. And it invites all sorts of speculation about how ‘the voters’ (meaning, the minority of the electorate whose party affiliation is so insecure that they may vote differently from last time) will react to issues and leaders, about which political operatives are usually happy to be quoted. So while elections are, for the most part, occasions for established allegiances to be recorded, there is enough uncertainty around the edges around the edges to give space for chatter.
Once the election is over, the focus is on the outcome; there is little interest in the process, and even less in how well any of this chatter helped us makes sense of the behaviour of the voters. If we were interested in explaining the voting in Eden-Monaro, one factor that was clearly important was was that the way that those with unstable allegiances voted depends on the choices on offer. As Tim Colebatch (yes, we are) pointed out (Labour’s preferential treatment, Inside Story, 7 July 2020), at the 2019 election, there were eight candidates: as well as the Usual Suspects (Liberal, National, ALP, Greens), two micro-parties and two independents contested the seat and got 8.1% of the vote. The major parties kept to a tight exchange of preferences – 87% of Green preferences going to the ALP, and the same proportion of Nationals preferences to the Liberals. The preferences of small party and independent voters tended to favour the conservatives, but less tightly: for instance, only 53% of the Palmer United preferences went to the Liberals or Nationals.
In the by-election, it was a different story: there was now a new party, the Shooters, Farmers and Fishers, which had won three seats in the state lower house, six micro-parties and three independents. So there were more choices on offer, and Palmer’s United Australia Party (which in 2019 had drawn votes from the ALP and fed many of them to the mainstream conservatives as preferences) was not standing.
As was expected on the retirement of a popular incumbent MP, the ALP vote fell, by 3.30%, and the UAP candidate, who had won 2.77% of the vote in 2019, was not standing, so even if we assume, for the sake of the analysis, that in general, people would vote the way they did last year, except those who had only voted Labor because its candidate was Mike Kelly, we could expect 6% of the vote to have been reallocated; where did it go ?
Well, some of it might just have stayed home. Despite compulsory voting, turnout tends to be a bit lower at by-elections, and there were five thousand fewer votes cast at the by-election than a year earlier. All four mainstream parties got fewer votes than in 2019, but the Shooters (who did not stand in 2019) got just over five thousand votes. The Liberal vote fell, but rose as a proportion of a lower turnout, though only by 1.35%, while the Nationals vote went down 12.5%%, so support for the Coalition fell by 1200 votes, and its share only rose 0.79%.
This did not stop the Australian seeing the result as ‘a savage by-election swing against Labor’ and ‘a powerful endorsement of the Government’s handling of COVID-19 with Scott Morrison’s personal popularity driving the swing’. But the figures are clear: there was no swing to the government at all. The fall in the Labor vote is equal to Peter Brent’s estimate of Mike Kelly’s personal vote. (Five things we learned in Eden-Monaro, Inside Story, 5 July 2020) Labor and the Greens both lost a significant number of votes, but they did not go to the government, but either to another opposition party (the Shooters) or to non-voting.
What can be seen is the significance of the small players’ preferences. With the major parties running neck and neck, these preferences can determine the outcome (like the crossbench in the Senate). And in a rural seat, voters are less likely to follow the ticket. They may not know what it is – it is difficult even for the main parties to find enough volunteers to cover all of the 70-odd polling booths for eight hours on polling day – and rural voters are more likely to have heard of the people standing as independents. Even when there is a ticket, they may exercise their own judgment: in 2019, only 38% of Christian Democrat voters followed the ticket, and all seven of the candidates got some CD preferences.
In the by-election, four of the small players expressed no preferences, four of them, and the Shooters, preferenced Labor, and only the Christian Democrats preferenced the Liberals. In 2019, the Liberals did slightly better than Labor on preferences (though not enough to overcome Mike Kelly’s first preference lead). This year, Labor had a distinct advantage (14.5% to 11.3%), which was just enough to creep past the Liberals’ first-preference lead.
So the Eden-Monaro by-election did exactly what it was meant to do: to allow people to decide who should represent them in Canberra. By compelling them to rank all the candidates, the system ensured that their voice would be heard right to the end of the count, and it determined the result: if that was the choice, more voters preferred to have McBain as their member than Kotvojs. And that was what they were asked: they were not asked about Scomo or Albanese or bushfires or Covid-19, and no matter what the commentariat might like us to believe, any attempt to distil from the electoral outcome a judgment by a mythical actor called ‘the voters’ on any of these issues is at best wilful self-deception.
One last aspect of the outcome: if Informal was a party, its candidate ‘None of the Above’ would have run third, behind the two majors but out-polling the Greens, Nationals and Shooters. The question is whether this should be seen as an indication of a need for electoral education, in many languages, as a sign of alienation from the whole electoral process, or as a useful ‘safety valve’ for compulsory voting: you have to be present, but you don’t have to say anything.