By Ben Aveling
NSW Labor is experimenting with a Community Preselection for the Labor Mayoral Candidate for the City of Sydney. The ballot will be part primary, part traditional preselection, with the community and party Members each collectively having an equal say in the final outcome. (By way of example, if there should happened to be twice as many votes from members of the community voted as from party members, then each community vote will be worth half of a party member vote.)
There have been two primaries held in Australia before, one by the Nationals in Tamworth and one by the Victorian ALP in Kilsyth. The Tamworth primary was a success – there were four serious candidates, 4293 people participated, party membership went up, and there was a 14.7% swing to the Nationals, compared to a statewide swing of 2.5%. The Victorian experience was the opposite – the only candidates were staffers, 170 people participated (rumour has it that most were trade-unionists organised by one faction or the other), about 5 new members joined, and the swing against the ALP was 10.0%, compared to a state-wide average of 6.8%.
The Sydney Community Preselection will be a success if it can ‘bring people back to the party’. But for that to happen, it must also ‘bring the party back to the people’. It must deliver a locally focused candidate with strong recognition and support in and outside the party – someone with a reputation of working for and with residents, not a would-be careerist who has ‘earned a turn’ from their faction.
The problem is that there will only be less than 3 months to run the community preselection as it must occur 3 months before election day according to the rules. While an improvement on the previous process, it is not enough time to allow people to build and organise a campaign.
There are other issues, for example, how do we know if a community preselection is a success and also the need for a distinct Labor vision for the City of Sydney that will resonate with its diverse demographics, which I do not feel have been adequately addressed as yet. I address some of these issues here: