Polling conducted in the run-up to the New South Wales state election accurately reflected the result.

It was a great performance for Australia’s polling industry. Most agencies, including Freshwater Strategy, correctly indicated final vote shares, within an acceptable margin of error.

As the final votes are counted, the results show that Freshwater Strategy’s (FWS) final poll was closest to the result overall.

FWS has repeated its successful public polling debut, ahead of the Victorian state election in November 2022, where it correctly called the final outcome and final vote shares.

In New South Wales, polling conducted by FWS, exclusively for The Australian Financial Review, was on the mark, not once, not twice, but four times, with two-party/candidate preferred results all within margin of error for the final results in the final state-wide poll, and all three seat polls conducted and published during the campaign.

FWS polling correctly indicated that the most likely result would be a minority Labor government, that the Liberals’ Rory Amon was on course to hold Pittwater, Labor’s Warren Kirby would win Riverstone, and that Goulburn was on a knife-edge but most likely to be held by Liberal incumbent Wendy Tuckerman.

Comparing final published polls

In the final 10 days of the campaign; Freshwater Strategy, YouGov/Newspoll, Resolve and Roy Morgan all published vote intention polling.

Please note; although we go on to compare each poll, readers should be aware that polling is only ever a snapshot of vote intentions in time, never a forecast.

For transparency, all the final polls are compared below.

When we compare the final published polls against the election result, we can calculate the 'error' for each published figure (i.e. the difference between the poll figure and the election result). The average Primary vote error is then calculated.

The two-party preferred error is then calculated, for each polling agency, and then included in the final 'Total Error' figure.

The results show that Freshwater Strategy's final poll, published in The Australian Financial Review, had the lowest average error, and was closest to the final result. However, all polling agencies were fairly close to the final result, with average error easily within an acceptable margin. We congratulate Newpoll for being closest to the final TPP result, albeit incorrectly suggesting in their coverage (before and after election day) that there would be a majority Labor government.

It should also be noted that final TPP results from all pollsters were within an acceptable margin of the election result. Well within the margin of error.

Reporting the results

The research team at Freshwater Strategy were careful to brief the AFR reporting team of the implications from our final poll.

While our poll showed that, days out from the election, the most likely outcome would be a Labor minority government, polls are always snapshots in time, not forecasts. There remained a significant possibility of Labor being able to form majority government in their own right, but only if voters began to change their minds in the final days of the campaign.

We also flagged the very small and diminishing possibility, of the Coalition being able to form a minority government, should most of the crossbench choose to work with them. Again, only if voters switched decisively and in strong numbers, to the Coalition in key seats during the final days.

While the reporting of other polls was more 'conclusive' of a Labor majority, even though that outcome never arose, our insight reflected the range of possibilities fairly. Proportionate to the probability of their occurrence based on the statistical fundamentals of conducting survey research.

Nevertheless, the most likely outcome occurred.

We are proud of the way our polls continue to be reported by The Australian Financial Review, and although Freshwater Strategy was the most accurate pollster on this occasion, we are also proud to be part of an industry that did well to accurately convey vote intentions and public opinion collectively.