Which state will be the tipping point?

Australian constitutional referendums aren’t fought at the seat level. They aren’t really fought at the national level. It’s actually a contest in six states.

A referendum requires a majority in four out of six states to pass. It’s also necessary to achieve a national majority, but there has never been a case where a referendum has achieved a majority in four states but fell short on the national majority. Indeed, out of 44 constitutional referendums, there has only been two where the national Yes vote was lower than the Yes vote in the fourth-best state (the tipping point state), and those were both held in 1910.

I’ve previously analysed margins of victory in past referendums in the tipping point state, and which states have tended to vote together in favour or against referendums.

For this post, I’m looking at some recent polls that have included state breakdowns, to get a sense of how (whatever the overall Yes or No vote might be) we might expect the different states to shake out.

I’ve identified eight polls conducted between the end of August and early October. Each poll reported state figures for at least the three largest states, with three of the eight polls publishing figures for all six states.

First it’s worth noting the caveats around using state polling. We don’t get state breakdowns of national polls for every poll, because the sample sizes tend to be smaller – so either you publish polls with greater margins of error, or combine multiple polls, or sample more participants to get a sufficient sample.

And usually when we do get state breakdowns, Tasmania is usually left out.

This referendum has been unusual in that we have seen so much state polling, but that makes sense considering the structure of the constitutional change process.

I’m also not going to get into the reliability of these polls for actually predicting the outcome – I’m mostly interested in what it tells us about the relative position of each state. And these polls are pretty much entirely on the same page when it comes to that question. In the table above I have shown the Yes vote

Only three of these eight polls include Tasmania, but it is always showing a majority for Yes. It looks like Tasmania is currently the best state for Yes, and quite possibly on track to vote Yes while the other five states vote No.

This does make sense when you consider the position of the Liberal Party specifically in Tasmania – the only Liberal government in Australia is supporting the referendum, as is prominent Liberal maverick Bridget Archer.

There are four polls that cover at least five states. If we assume that Tasmania is also the most pro-Yes state in the early September Essential poll (which covered the other five states), then we have rankings from four polls.

New South Wales and Victoria are generally ranking second and third, but the polls are split on who is ranked second or third. New South Wales has a higher Yes vote than Victoria in four out of eight polls, and vice versa in the other four.

Western Australia and Queensland are clearly the strongest No states. Apart from the most recent Essential poll, the Yes vote has not exceeded 40% in either state in any of these polls.

Finally, this leaves South Australia. In the four states that include South Australia, it stands alone as the tipping point state. It always has a higher Yes vote than WA and Queensland, and lower than NSW, Victoria and (where polled) Tasmania.

Historically Yes votes have generally found it easier to win a national majority than to win a majority in a fourth state. This doesn’t seem to be such a big issue this year. While some polls have had SA having a lower Yes vote than the national total, it’s only 2% in two polls and 4% in another. In the most recent poll, SA was slightly more pro-Yes than the country.

Generally the most populous states NSW and Victoria have tended to be more pro-Yes than less populous states. Tasmania has usually been ranked as one of the least supportive states in referendums. Moving Tasmania to the other end of the spectrum makes the task of winning four states a bit easier, relative to the national win.

At the moment the polls seem to be pointing to a clear win for No, likely with four or five states voting that way. The difficulty of polling Tasmania means I wouldn’t even rule out a No win there. But if things turn out closer, I think we’ll be watching South Australia while not ruling out the possibility that the deciding factor, for the first time in a referendum, would be the national vote.

Liked it? Take a second to support the Tally Room on Patreon!