I’ve been fascinated by how the two halves of Tasmania have moved apart at this election. The Liberal Party lost support overall, but managed to gain swings in the three northern electorates. This has been the culmination of a trend which has seen the gap between the north and south larger than it has been at any point in the last three decades.
For our purposes I’m defining “north” as Bass, Braddon and Lyons. Admittedly the seat of Lyons stretches to the Hobart suburbs, but it’s mostly a large rural electorate with most of its population outside of the Hobart area.
On the other hand you have Clark, contained entirely in Hobart, and Franklin, which covers the remainder of southern Tasmania, but is about 50% Hobart.
The gap between these two areas has seen the Liberal Party much stronger in the north than the south, and the Greens vote much higher in the south than the north, with a gap not previously seen.
Unlike a mainland state using single-member electorates, the number of electorates has been consistent for the last century, with only relatively minor changes to boundaries. This makes it much easier to compare the relative position of electorates over decades.
Now I should also acknowledge that there were candidate effects that were relevant in the anti-Liberal swing in these areas, namely the retirement of former Franklin MP Will Hodgman, and the independent candidacy of Sue Hickey, a former Hobart mayor elected as a Clark MP in 2018.
The trend is most dramatic when looking at the Liberal and Green votes, and I think it’s a particular problem for the Greens in the future.
This first graph is simple. It just shows the Greens vote in each electorate over the last ten elections, all the way back to the first time the Green Independents contested every electorate in 1989:
Electorates mostly have moved in line with each other, and the order has mostly remained the same. Braddon has consistently been the worst electorate, with Bass and Lyons a bit better. Franklin has usually come second to Denison/Clark, with an exception in 2010 when Greens leader Nick McKim was defending his seat in Franklin and the Greens were running a new MP in Cassy O’Connor in Denison.
But while the order has stayed the same, the gap between the best and worst electorates has grown, and has never been worse than in 2021.
Another way to look at this same information is to subtract the statewide vote from each seat’s vote.
The gap shrunk to its smallest point in 1996 and 1998. 1998 was the worst ever result for the party in terms of votes and seats. The gap started to widen in 2002 when the Greens bounced back and regained three seats lost in 1998.
Denison/Clark has been well out in front of the northern seats consistently since 2002, but Franklin is now as far ahead of the statewide total as it’s ever been. Meanwhile the vote in Bass has now fallen further away from the statewide average. The gap between second and third is now 9.7%.
Meanwhile Braddon has bounced along the bottom, consistently polling 6-8% below the statewide total. This means the Greens have only been able to win in Braddon at a record high statewide vote in 2010.
This gap matters because the Greens have two safe seats in the south, but are a long way from winning a second seat in either southern electorate. Their best prospects for growth remain in Bass and Lyons, but the party had a relatively small swing in Lyons and lost ground in Bass.
We have seen a trend in lower house results in the bigger mainland states where the Greens have stayed consistent across the state but have built up support in their inner city heartland. That’s helpful under a single-member electorate system which rewards concentrated support, but the Hare-Clark system punishes that concentrated support.
But this story isn’t just about the Greens. It’s also very obvious for the Liberal Party.
This chart is the same as the last, but for the Liberal Party.
The Liberal vote in Bass is now about 11% higher than the statewide total, which is more than we’ve ever seen before. The gap is 8.3% in Braddon, which is about the same as in 1996. Lyons remains relatively close to the statewide average but has ticked above it in 2021, leaning more towards the Liberal Party for any time since the early 1990s.
Denison has always been a relatively weak electorate for the Liberals but this weakness has been growing, even before the Hickey candidacy widened the gap further than any time in the last three decades. The Liberal vote in Clark is now 17% less than the statewide vote.
You can also simplify this widening gap by looking at the average Liberal vote in the three northern electorates compared to the statewide vote, and the same for the two southern seats:
The two lines parallel each other, but the southern line is more exaggerated since they don’t cover the same number of electorates.
I’ll note that the gap between north and south was smallest from 1998 to 2010, elections which the Liberal Party lost. I won’t crowd this post with a similar chart for Labor, but the Labor vote was actually higher in the north of the state during that period of government, but Labor was more popular in the south during the previous 1989-1998 era, and from 2014 to 2018. The crashing Labor vote in Clark has balanced out the two regions.
So Labor does better in the north when they are on an upward trend, and the Liberal Party’s north-south gap widens when they are winning majorities in the House. That gap is the thing putting the Liberal Party in majority territory, and if Labor is to win back Tasmania they’ll need to win back some more seats in the north of the state.