The big swinging seats


The pendulum orders every seat in the country by the swing required for that seat to change hands. By following the pendulum you can determine roughly how many seats will fall with a particular national vote. Antony Green has built a calculator which can work this out for you.

Following the pendulum only gives you a rough idea of what might happen. Some seats will swing more strongly, and others won’t, thanks to a whole bunch of factors including the incumbent MP, election issues, and demographics.

So how do we identify which seats could see larger-than-average swings and fall despite a seemingly ‘safe’ position on the pendulum?

One factor worth considering is how much the seat has swung since the last Coalition victory in 2004. While many marginal Labor seats swung back to the Coalition in 2010, a number of seats, mainly in Victoria and South Australia, swung even further to Labor in 2010.

There are fifteen seats where the cumulative swing from the Coalition to Labor at the 2007 and 2010 elections is greater than 9%. These are a good place to start when looking for the big swingers.

Read below the fold to find out where those seats are.

SeatState2013 marginCumulative swing
KingstonSAALP 14.6%13.98
LalorVICALP 22.1%13.36
MakinSAALP 12.0%13.13
WakefieldSAALP 10.5%12.62
McEwenVICALP 9.2%11.74
HoltVICALP 14.0%11.72
CalwellVICALP 20.1%11.53
AstonVICLIB 0.7%11.39
DobellNSWALP 5.1%9.88
PageNSWALP 4.2%9.66
IsaacsVICALP 10.4%9.54
BallaratVICALP 11.7%9.47
FordeQLDLNP 1.6%9.44
BassTASALP 6.7%9.34
LyonsTASALP 12.3%9.05

Thirteen of these seats are held by Labor. Forde is the only seat on this list to swing against the ALP in 2010 – it’s on the list due to a massive 14.4% swing away from the Liberal Party in 2007.

The list is complicated by the existence of redistributions, but luckily most of these seats were unaffected by redistributions. Only Bass, Dobell, Forde and Lyons were affected by the redistribution, and in none of those seats was the margin affected by more than 0.5%.

You’ll notice most of these seats, including all of those at the top of the list, are in South Australia and Victoria, which generally swung towards Labor in 2010.

I’ve bolded six seats where the margin at the 2013 election is less than the cumulative swing over the last two elections. While three of these seven seats are considered to be reasonably marginal, and are likely to fall to the Liberal Party, the other four are held by quite large margins.

Makin, Wakefield and McEwen, however, are all held by margins of more than 9% despite a recent history of being considered “marginal”. The Victorian seats of Holt, Ballarat and Isaacs and the Adelaide seat of Kingston don’t quite fit this criteria, but these seats would be very close if Labor’s cumulative swing flows back to the Liberal Party.

There could be a bunch of reasons why these seats look so different on paper to how they looked in 2004: demographic change, particularly popular local members could have such an affect. But it could also suggest that the Labor margin is pumped up with a lot of soft voters who could easily decide to switch back to the Liberal Party.

If you’re looking for seats which could fall despite sitting quite far up the pendulum, it would be smart to start with the Melbourne seats of Holt and Isaacs, the rural Victorian seats of Ballarat and McEwen, and the Adelaide seats of Kingston, Makin and Wakefield.

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


  1. Who wants to have a go at guessing which seat will swing most at the election? My knowledge only extends as far as QLD so I will say Oxley, but if not, then a seat in Western Sydney.

  2. It would be overly humorous if Lalor were to fall.

    That said, Bass and Braddon are going to be big movers.

  3. This is a very good post. At this point, I think all of those seats in bold will fall to the Coalition, except McEwan and maybe one of the SA seats.

  4. One matter which will be relevant and probably explains the loss of support in the polls in Victoria is the increase in unemployment there,whereas there was a decrease in NSW.So that will be one reason for a big swing,if there is one.

  5. Two in NSW, one in Qld, and all the rest are in SA, Vic and Tas. So much for western Sydney.

    I’d like to see the other end of this list, with all the negative numbers. At a total guess I’d say Cowan would be right up there… it’s gone from fairly safe Labor under Howard to fairly safe Liberal under Gillard. Even compared to WA, it’s been heading toward the Libs quite solidly. The same thing has been happening over the last few state elections… compare to Girrawheen and Mirrabooka (and others – this area always gets chopped and changed at state redistributions), which have gone from ~20% margins a decade ago to marginal now.

  6. Cowan was a Beazley seat & only held in 2004 because of personal vote for Edwards (he ran well ahead of Senate vote). Ballarat is in part a cultural left seat that will insulate it especially against Abbott I think. McEwen Libs are rightly giving it attention & it is outer suburban mortgage belt but incumbency is an advantage.

  7. Bird of paradox, you have to remember regarding Cowan, it had an inflated Labor margin during much of the Howard years, thanks to popular member Graham Edwards.

  8. It’s true that Edwards was a popular man, but that’s not the whole story. Ten years ago, there were patches of rock-steady safe Labor seats in the uglier bits of both southern and northern Perth – both often held by no-name political careerists. (Ted Cunningham and Norm Marlborough… where are they now?) Over the last few elections, the north (the general Balga / Mirrabooka / Koondoola region) has gone a lot more blueish than Rockingham and Kwinana. When Cowan and Brand were created in 1984, they were pretty similar, but those areas of Perth have diverged politically over the years.

  9. I think it’s pretty clear that the biggest swings at this election will be in areas which might have ‘delayed’ swings at the last election. Victoria, SA and Tas will definitely swing the most. WA has swung right at the last 2 elections and so it doesn’t necessarily have the capacity to swing further – You’re starting to get to the rusted-on Labor supporters.

    Queensland and NSW, while they both swung Liberal last time and will have smaller swings this time, will remain the key battleground states, as that’s where the majority of marginal electorates are.

    For example, by the time you reach Oxley on the pendulum in Queensland, at a uniform swing of 5.8%, you’ve taken all bar one Queensland Labor seat, but only three seats in Victoria. At the same time, you’ve taken ten seats in NSW.

    Despite this, I think on election night we’ll see swings in the order of at least 5% in every state, with WA at about 5%, Qld, NSW and NT at about 6-8%, Vic maybe 8% statewide, but varying from 10% in some seats to 5% in others, and Tas at 10%

  10. Working on a proxy for electoral volatility (Beta). Adelaide’s outer suburbs, Melbourne’s sandbelt and the Hunter Valley/Central Coast are home to the high Beta electorates where state wide swings are strongly amplified. Currently Bass would have to be the front runner, but it can be eclipsed by a seat in one of these three areas if there is a big movement in the ALP vote.

  11. Remember – maths tells us the seats that in a uniform swing the biggest swings are likely to be the safest seats. If the TPP is 80% and 1 in 10 labor voters change their vote the swing will be 8%. Conversely if the TPP is 20% and 1 in 10 voters change their vote the swing will be 2%. There might be a surprise in the biggest swing being somewhere like Port Adelaide but the seat won’t change due to the initial margin. Interestingly the biggest (and probably Australian record) swing at the last NSW election was Bathurst at 36% which I suspect nobody predicted at the time.

  12. It depends whether the definition of swing is only between Liberals and ALP

    If Y then Bass, If N then Lynn

  13. Recent polling down this way shows swings in Bass @ 18%, Braddon @ 19%, and Lyons @ 21%.

    Someone ought to run a sweep for biggest swing on the night : )

Comments are closed.