The time has come for me to stick my neck out and make predictions for the NSW state elections to be held on Saturday March 25. First, I give the predicted numbers in the Legislative Assembly: Labor 43, Liberals 29, Nationals 12, Greens three and Independents six, total 93. Dominic Perrottet resigns, and Chris Minns becomes Premier in a minority Labor government after he receives a guarantee of supply from the three Greens and the independent members for Barwon, Lake Macquarie, Sydney and Wagga Wagga. Therefore, he has a guarantee of supply from 50 of the 93 members, a majority.

My reasoning is the same in principle as was the reasoning for my successful predictions for the May 2022 federal elections and the November 2022 Victorian state elections. See articles posted last year on May 19 “My final predictions for Saturday’s election” and on November 9 “Seats to fall but Victorian Labor safe”.

Essentially, I make opinion poll induced predictions combined with Mackerras Pendulum induced forecasts – plus the occasional personal judgment explained below. Therefore, I show now the latest Mackerras Pendulum. It was published in Sydney’s “Daily Telegraph” on page 6 for Monday 6 March 2023.

There was also an article by me explaining the pendulum titled “How state of play has swung round”. The article made no predictions. Among other things it explained that two seats won by the Liberal Party in March 2019 are now shown as Labor. The seats are Bega won by Labor at the February 2022 by-election and Heathcote where boundary changes have been so bad for the Liberal Party as to change that seat from actually Liberal to notionally Labor. Both seats are shown in red on the pendulum.

By averaging the published opinion polls, I predict that the state-wide two-party preferred vote will divide 53 to 47 in Labor’s favour. That would be a 5% swing. Going to the pendulum, it tells me that on a uniform swing Labor would gain five seats, East Hills, Penrith, Goulburn, Upper Hunter and Tweed. So, what about Sydney and Barwon? The independent in Sydney, Alex Greenwich, will retain his seat easily – the buffer statistic is the swing needed by Labor to take Sydney from a notional Liberal member. Labor might win Barwon since it now includes Broken Hill, but I think the sitting member, Roy Butler, will hold on. He was elected for the Shooters, Fishers and Farmers Party but is now an independent.

At last year’s Victorian elections, the Nationals performed strikingly well, while the Liberals performed badly. For that reason, I forecast the Nationals will retain all their 12 current NSW seats. However, in Mackerras-type psephology there is a concept known as “cancellation of deviations”. The Liberals, therefore, will lose two seats not named so far. Which two? I don’t know, of course, but I mention that in these three seats the sitting Liberal is retiring – and I give their buffer statistics, Riverstone 6.4, Parramatta 6.5 and Ryde 8.9.

In modern psephology there is a concept known as “retirement slump”. There will be unusually large swings to Labor in Parramatta, Riverstone and Ryde but I predict one will stay Liberal and two will fall to Labor. Let it be noted, however, that retirement slump may affect other parties, not only the Liberal Party. For example, the very popular Greens member for Balmain, Jamie Parker, is retiring which means Balmain could fall to Labor. Ballina and Newtown will stay with the Greens, and I guess Balmain also – but I would not be surprised if Labor won Balmain. A majority Labor government, therefore, remains a serious possibility.

Finally, readers may wonder why I show Leppington in red along with Bega and Heathcote. The answer is that Leppington, in south-west Sydney, is the new seat in the redistribution. It replaces the abolished Lakemba, which was also in south-west Sydney but much closer to the central business district. Lakemba was a was a very, very safe Labor seat.