As luck would have it, Malcolm Turnbull’s reverse at the Super Saturday byelections landed just as opinion polls started pointing out a potential path to his re-election.
Not that the government is there yet. Its famous losing run in Newspoll extended to 37 last week, and no other established pollster has run a poll that failed to credit Labor with a two-party lead throughout this period, which began shortly after the 2016 election.
However, the polls of the past fortnight have found the Coalition three points ahead on the primary vote in Newspoll, for the first time since December 2016; at 41% in Essential Research, its strongest primary vote in all poll published since September 2016; and level pegging in Ipsos’ little-publicised two-party measure that allows respondents to allocate their own preferences, rather than assuming they will flow as they did at the last election.
So far though, the improvement in voting intention has been modest compared to Malcolm Turnbull’s approval and preferred prime minister ratings.
The latter has particularly been a favourite of journalists talking up a threat to Bill Shorten’s leadership — though their doing so has perversely ended up strengthening him, to the extent that it was the cause of exaggerated expectations for the Liberals’ prospects on Super Saturday.
While the press gallery was undoubtedly getting ahead of itself, there has been an historic tendency for prime ministerial approval to act as a leading indicator of voting intention. Trends measures of two-party preferred and net leadership approval (i.e. approval minus disapproval) suggest just such a dynamic is slowly playing out at present.
However, the evidence from polls, as well as byelections, is that the Coalition recovery is inefficiently distributed. It would leave them with little basis for confidence about an election held in the immediate future, even if an optimistic view is taken on the capacity to claw back support over an election campaign.
A picture of the situation at state level is provided by the aggregated poll trends which have been bolstered through the recent publication of detailed quarterly data from both Newspoll and Ipsos, together with unpublished state-level data from Essential Research.
This shows the Liberals looking buoyant in South Australia, where Steven Marshall’s state government is in its honeymoon, and the Liberals seem to be gaining support from the many voters who have deserted Nick Xenophon’s party since its disastrous showing at the state election in March (Rebekha Sharkie’s success in Mayo being an entirely local phenomenon).
But besides shoring up the Liberals in their vaguely loseable seat of Boothby, a favourable swing in South Australia is of no strategic consequence, since none of the five Labor seats are remotely winnable on the post-redistribution boundaries.
A somewhat more productive bright spot for the Coalition is Western Australia — which, in a reversal of the South Australian situation, gave them a diabolical stretch of polling that coincided with the honeymoon period of Mark McGowan’s government last year.
This left pessimistic Liberals fearing a realignment that would cost them a minimum of three seats, including Attorney-General Christian Porter’s seat of Pearce.
Labor’s defeat in the Darling Range state byelection on June 23 clearly signalled the end of the state Labor honeymoon, and federal polling now suggests all the state’s Liberal seats are at least potentially salvageable, given help from judicious marginal seat pork barrelling.
However, that will not avail the government if it suffers any kind of reversal in the marginal seat motherlode of Queensland, and it is there that polls suggest the Coalition faces the biggest swing at all.
Given expectations of a strong showing in Longman, it was possible before Super Saturday to imagine either that the Queensland poll trend was seriously askew, or that Labor had only very soft support that would not survive the heat of an election campaign.
Now it cannot be denied that the Coalition really is bleeding support in Queensland to Labor on the left and One Nation right, and that turning the ship around will not be the work of weeks or months, as early election speculation presupposed — if it can even be done at all.