Plenty of margin for error in Dutton’s ‘Trump bump’
By David Crowe
The most important number in the latest Resolve Political Monitor is the margin of error. At 2.4 percentage points, it is an essential reminder of the uncertainty in measuring voter intent.
That reminder is even more important than usual when so many Americans are reeling from Donald Trump’s victory after feeling so sure he was on course for defeat.
The core findings in the Resolve Political Monitor show that victory in Australia hangs in the balance: the Coalition has taken a small step to a new high, while Labor is not moving at all.
The result is 50:50 in two-party terms when preferences are allocated in the same way people voted at the last election.
But we make sure to include the margin of error - and the sample size - because the survey is never as simple as those two final numbers. Given the full range of the margin of error – plus or minus 2.4 per cent – the result for the Coalition might be 52.4 per cent in two-party terms. Equally, it might be 47.6 per cent instead.
The range is the same for Labor: as good as 52.4 per cent, as weak as 47.6 per cent.
But the trend is certainly running Peter Dutton’s way. The opposition leader has made personal gains when voters are asked about his personal performance, so he has a positive net rating. The Coalition primary vote is 3.3 percentage points above the result at the last election.
Some will see this as a “Trump bump” that delivers a gain for Dutton after the US election. But correlation is not causation: there is nothing in the latest survey to show that Australian voters are shifting to Dutton because of the American outcome.
Whether the Trump victory is a factor or not, the numbers are grim for Anthony Albanese. The core support for Labor is now 2.6 percentage points below the party’s result at the election and the prime minister has lost his edge over Dutton as preferred leader. It is only a small consolation for Labor that its primary vote was lower, at 28 per cent, in September.
Many Australians are aware of the trend against the government. When voters were asked at the end of last year who would win the next election, 41 per cent said Labor. Now 42 per cent say victory will go to the Coalition. This is not a forecast, of course, just a guide to voter sentiment.
Jim Reed, who runs these surveys for us at Resolve Strategic, notes that the American polls underestimated Trump for the third election in a row.
“The same features that make their electoral system inferior to Australia’s – non-compulsory voting and the electoral college – also make it more difficult to measure,” he says.
“But it’s also a product of hundreds of cheap polls being used as media fodder.”
The polling aggregates in the US tried to make sense of the blizzard of polls, but in simplifying the results to two simple averages, they created a false sense of precision.
The final Resolve survey at the Queensland election had Labor on a primary vote of 32 per cent and the Liberal National Party on 40 per cent. The election result was 32.5 per cent for Labor and 41.5 per cent for the LNP. The outcome was within the margin of error for the survey.
The key point is that we are dealing with a range of outcomes, not a single number. Carpenters can check the length of a wall with a laser measure; pollsters have to allow for uncertainty.
Cut through the noise of federal politics with news, views and expert analysis. Subscribers can sign up to our weekly Inside Politics newsletter.