Liberal Party paid $1.5m for polling that failed to predict election result

News.com.au

08-05-2025

EXCLUSIVE
Furious Liberals have revealed the party paid about $1.5 million to Freshwater Strategy for polling that was still telling them on the day of the election they would pick up seats.
The amount paid for the polling project can be revealed as some strategists push back on criticism insisting Peter Dutton made captain's calls.
As angry donors prepare to close their cheque books and complain about the millions of dollars handed over to make Mr Dutton prime minister, party strategists are counting the cost of the disastrous campaign.
High on the list of regrets is the money spent on the political research project designed to get Mr Dutton elected.
Liberal director Andrew Hirst and the pollster Freshwater have flatly refused to comment on the confidential contract but senior Liberals have told news.com.au that the final figure will be around $1.5 million.
'I understand the final would be north of $1 million. I don't know the exact amount but $1.5m would be around the mark at least, depending on any outgoings he has incurred,'' a Liberal source said.
The contract is known to be particularly expensive as it included focus group research with voters as well as quantitative poll research.
Peter Dutton's 'big surprise'
In the final week of the campaign, Mr Dutton said he was hopeful of a 2019 result when Liberal prime minister Scott Morrison triumphed over predictions of a Labor victory.
'I think we're seeing a 2019 situation, where you've got a lot of interesting contests playing on the ground, where we've had a very significant effort by great candidates.
'And I think there'll be some big surprises on election night, because people have had enough.'
He was right on the big surprises prediction but not on where their votes would land.
Liberal MPs insist polling was not accurate
'We had bad numbers, way off the mark, totally out of line with all of the published polling,' Liberal Senator Johnathon Duniam said.
'Our own polling here in Tasmania pointed to the wipe-out we ultimately got and so there are some people in our campaign headquarters who are going to have to answer some questions for us around what went wrong here.'
Hitting back at attempts to blame polling for the Liberal Party's woes however, some strategists insist that's simply 'not fair'.
'At no stage was anyone saying to Peter Dutton, 'you're going to win,'' a campaign insider insisted.
'I'll be absolutely honest with you, the polling definitely did not show us going backwards, either.'
The Freshwater poll never predicted a seat count for the Liberals in the 70s. At the highest point of the campaign, it predicted a count in the high 60s.
By the end of the campaign, Liberal Party director Mr Hirst expected, based on the polling, a seat count in the low 60s and the Labor Party with a slim majority.
Mr Dutton's constant suggestions that he could win the election and secret Liberal polling was at odds with published polls was nothing more than political spin.
Liberals want to dump pollster
A blame game has now erupted with the Liberals preparing to dump the pollster when his contract runs out in June and blaming the research for convincing Mr Dutton the situation wasn't as bad as it turned out to be on election night.
News.com.au understands there was a greater reliance on focus groups than there was in previous campaigns.
The focus groups are a room of around 10 people held in key electorates multiple times a week.
'The jury is still out on the efficacy of the polling. There are some MPs who weren't that close to the polling complaining about it,'' a Liberal strategist said.
'Did the track tell Peter Dutton to go out there and campaign for nuclear policy? No.
'Was it telling him to campaign to cut 41,000 public servants? Far from it. Was he advised not to do that? Of course. So, there's a little bit of corporate warfare here.'
The polling was overseen by Freshwater Strategy's Dr Mike Turner and involved a tracking poll that was conducted every second night in 15 key electorates across Australia in key marginal seats with 1200 votes.
'The starting point was really that our internal data was wrong,'' a Liberal strategist said.
'If we knew that was the situation we would have spent much more time and resources defending seats rather than trying to win seats.
'Clearly, then there's just a broad rejection of what we were offering Australian people.'
Freshwater Strategy
A pollster and strategist, Dr Turner has worked as a senior adviser on several pivotal election campaigns in the UK, Australia and New Zealand.
The polling company also provided a two-party preferred national poll for The Australian Financial Review that predicted before the election that the Albanese government faced a hung Parliament or around 76 seats.
With 80 per cent of the vote counted the Labor Party is currently on 89 seats and the Liberal Party 40 seats.
But as insiders point out, the tracking poll that political parties rely on is never a hard count of seats.
That formal work on modelling the seat estimation was actually contracted to a different company.
Nevertheless, Liberals who worked in the inner sanctum thought that some of Mr Dutton's predictions were based on Freshwater data.
'That would be guesstimates from (Freshwater), absolutely,'' a source said.
What is the tracking poll and what does it do?
What voters and the media sometimes fail to understand is that the political research for elections known as a tracking poll do not provide a hard count of seats or a two-party preferred prediction in the same way published polls do in newspapers.
The polling is also tightly controlled and only provided to the leader, the leader's office and Liberal Party directors which means many of the people complaining about the polling have never seen it making it difficult to analyse in a fair way.
A post-mortem will now be held into the result, with an experienced Liberal Party loyalist such as Brian Loughnane tasked with analysing where it all went wrong.
What did Freshwater's polling say?
Just days before the election, a Freshwater Strategy poll predicted in the Australian Financial Review that Labor was likely to form a minority government at Saturday's federal election.
The polling conducted predicted a 1.5-point swing towards Labor in a fortnight on a two-party preferred basis, with the government leading the Coalition by 51.5 per cent to 48.5 per cent.
The current estimated two-party preferred vote is 54.8 per cent for Labor and 45.2 per cent for the Liberals.
The figures in the poll just before the election represented a swing of 0.6 per cent since the 2022 election, and if applied across all electorates, the number of seats Labor holds would fall to 76 – the figure required to form majority government – down from the 78 it held at the start of the campaign.
The company, which includes a number of staff from the Liberal Party's previous pollster Crosby-Textor or CT, was established in 2022.
In a post-election opinion piece for the Fin, the pollster blamed a few factors.
'Like most other polls, Financial Review-Freshwater Strategy polling picked the trend, accurately estimated Labor's primary vote, but we underestimated its final TPP. After some initial analysis, we put this down to three main factors,'' the pollster Dr Turner said.
'First, polling appears to have over-estimated Labor 'defectors' to the Coalition. Particularly, those who voted No at the Voice referendum. Early indications suggest that 'Labor-No' voters just didn't switch over to the Coalition in the big numbers estimated.
'Second, for all the noise about the preference flows being different in a way that would substantially benefit Coalition performance, it appears that the outcome simply did not materialise. The primary vote collapse for the Coalition was too much for any benefit from additional preference flows.
'Third, the late swing. Given that all pollsters seem to have underestimated the swing to Labor, and everyone's fieldwork would have been over the earlier days in that week, it strongly suggests that there was a late swing among 'soft' or undecided voters in the final days that was very hard for pollsters to pick up.
'The Financial Review/Freshwater Strategy poll has a strong track record of calling election outcomes. We were the closest pollster for the Voice to Parliament referendum, we accurately called recent Victorian, NSW, Queensland and West Australian state elections. But this time some others were closer to the mark.'
Why the election swung so much
Dr Turner argued that Labor's strategy to move quickly to frame Mr Dutton, was highly effective.
'Every pollster showed the same trend of Dutton's net approval collapsing throughout the campaign, and Albanese's lead as preferred PM increasing,'' Dr Turner said.
'Labor timed its move well, characterising Dutton as someone who would cut essential public services if elected. The consequence of voting for your local Coalition candidate was getting a Dutton-led administration that was a 'risk' to services that people already had, and valued.
'In times of uncertainty people tend to stick with what they know. No doubt Trump's global tariff interventions played some part in creating greater hesitation about the future, particularly as the sharemarkets went flying.
'But when most of the 'soft' or undecided voters chose Albanese's Labor in 2022, as our polling suggested, it's easy to see how the final stages of a campaign could be so impactful to the final result.
'With former Labor voters considering switching to the Coalition but breaking back home to Labor in strong numbers, it's likely that some of the seats that Labor was hoping to pick up only materialised in the final days of the campaign.'
The Voice referendum and the 2025 election
As the post-mortem begins into how it all went wrong, Liberal strategists have told news.com.au that they believe they know three reasons why the polling did not predict the result.
The first reason is that the Voice referendum results were factored into the Liberal polling methodology over-inflating support for Mr Dutton among Labor's 'No' voters.
The second reason is that Gen Z and millennial voters are notoriously difficult to poll because many don't answer unknown numbers on mobile phones.
The third reason was that the Liberal Party didn't poll in Mr Dutton's seat in the final weeks of the campaign as it was thought it wasn't necessary as they were 'throwing everything' at the seat anyway.
After 24 years in parliament, voters in Dickson then hit the eject button on Mr Dutton's political career on Saturday night.