Dating back at least to 1983, the major party leaders have addressed the National Press Club in the final week of the federal election campaign. It's an afternoon event with a tiny live TV viewership, and only twice in that time has one of the leaders declined the invitation.
Both of them were opposition leaders - John Hewson in 1993 and Bill Shorten in 2019 - and they each went on to lose an election they were widely expected to win. In fact, if election-eve betting markets are an indicator of general expectations (and it's hard to think of a better one), 1993 and 2019 are the only "shock" federal results of the past 30 years.
Less easy to quantify is the fact that both were "big target" elections, with the opposition taking ambitious economic policies to the ballot box. And in both cases the government ran a ferocious negative campaign.
As they say in the classics, "correlation doesn't equal causation", and it can't be claimed that the Press Club no-shows decided the election result. In fact, both decisions were probably symptomatic of nervousness within the expected victor's campaign team: we're ahead, but it's softening; the last thing we need is a headlined blunder in the final days.
The lesson? Don't take big, complicated, serious economic reforms to an election, especially from opposition. When John Howard regained the Liberal leadership in 1995, he cast off his 1980s baggage bit by bit: his suspicion of Medicare, his talk of drastic spending cuts and workplace deregulation, his worries about Asian immigration and, of course, his enthusiasm for a GST. The "small target" label was born. And it worked.
The current federal opposition seemed to have learnt a similar lesson from the 2019 result. But then, early this year, it announced a great big industrial relations overhaul. I had a rant about the lack of wisdom of this in February; it's the kind of thing an opposition leader under pressure does to shore up support within the party, but it decreases their chances of electoral success. [The Covid crisis saw this dynamic hit extreme levels across the country as Liberal state and territory oppositions estranged themselves from the middle ground by succumbing to internal pressure to demand an early end to restrictions.]
During the 2019 campaign Labor was burdened not just by its policies but also by a stunningly inarticulate leader, habitually evasive in front of the camera. Still, Bill Shorten had done better than expected at the election three years earlier, and the policy suite back then hadn't been modest: the same housing policy and the same ambitious climate change target, though without the franking credits. But few voters considered a Labor victory a likely prospect that year, and few had focused on the consequences if it happened. [And, of course, there was also "Mediscare". The Coalition got their own back three years later with Labor's "death tax".]
Maybe one lesson from 2016 is that when no one expects you to win, you can get away with more. In 2019, by contrast, Labor was the heavy favourite, and a Shorten government loomed large.
That lesson goes against the "bandwagon effect" that is popular among political scientists and observers but has never made sense to me. [Except for special circumstances, such as a likely hung parliament, which can propel voters to attempt to give the expected winner a working majority. This might have played a part in Victorian Labor's shock 2018 landslide.]
Given all that, what will the next election look like?
Assuming Scott Morrison is still Prime Minister, election-watchers - and indeed voters - will add an electoral-prowess loading to poll-driven perceptions of the likely winner. We saw this at last year's presidential election in the United States, with Donald Trump given a much better chance of winning than the polls suggested. And the American doubters were proved half right, because a roughly nine-point polled national lead for Joe Biden became a 4.5 per cent margin on election day (overall, state-by-state polls also overstated his support). The polls picked the right winner, but were even less accurate than four years earlier.
We don't know if Australia's pollsters have improved since 2019. But those suspicions that Morrison will once again pull off a miracle will assist the opposition.
The government will want COVID-19 to be as prominent as possible, but it's tricky to hold the election early without appearing not just opportunistic but willing to put lives at risk for political gain. The Tasmanian government is taking that punt as we speak, with a half-believable excuse, but it began the campaign a mile ahead in the surveys. The federal Coalition, by contrast, is generally a bit behind at the moment - if you can believe the polls.
Labor will be led by Anthony Albanese, Tanya Plibersek or Jim Chalmers. Albanese is probably not as uninspiring as he seems; anyone who got the opposition leader's job in May-June 2019 would be looking worse for wear by now. But no one forced him to grab it. Plibersek has the recognition, is probably Labor's most accomplished media performer, and would generate the most media pizzazz. Chalmers is a Queenslander (a good thing, electorally) and seems to know his way around a balance sheet. But he's rather inexperienced and seems susceptible to those empirically nonsensical ideas about coal-mining communities being the key to electoral victory.
Labor's Achilles heel remains its perceived economic incompetence, the sense that it spends too much when in office. That image is exacerbated by an institutionally ingrained response when the topic is raised: the party's leaders avoid the question by changing the topic, inevitably if they can, to health and education, as if that will push those two Labor strengths to the front of people's minds as they cast their ballots. But voters, particularly Australian ones, tend to be risk-averse, and those niggling doubts about financial security - repaying that loan, financing those renovations - keep niggling right up until they mark the ballot paper.
When the opposition looks like it doesn't want to talk about economics, it increases those doubts. This terror of defending its record stretches back to the Rudd and Gillard days; only the departed Lindsay Tanner seemed able and willing to engage on this terrain.
And did I mention the opposition has burdened itself with a big IR platform?
- Peter Brent writes regularly on polls and politics for Inside Story.