If there is a substantial expansion of the size of the federal Parliament, it will be a rare event. It has only happened twice before since federation, and hasn’t happened at all in my lifetime. About two thirds of Australian voters hadn’t cast their first vote before the last expansion in 1984.
At that year’s election, the House of Representatives was expanded from 125 to 148, adding 23 new electorates across the five mainland states. Almost every seat was redrawn, and the map looked quite different aftwards.
So with something so rare, how do we know how it would likely play out?
This is the first of two blog posts where I look at the electoral map of the early 1980s. For today’s post, I look just at how the map changed, before the 1984 election. I compare the results of the 1983 election to Malcolm Mackerras’ estimates of margins for the new electorates to identify how the distribution of marginal seats changed, and how the pendulum changed.
For the next blog post, I am going to look at how the new map affected the election – where MPs decided to run and which seats were left without a sitting MP, and how incumbency influenced swings against the Labor government.
I should clarify that in this blog post I will not be looking at all at the results of the 1984 election. Any comparison of 1983 and 1984 is simply comparing the different boundaries, both based on 1983 voting data. Everything in this blog post could have been known before election day in 1984.
First, what were the total number of seats per party? There were no non-classic seats back in 1983, so this is really just a story about the major parties.
Labor’s share of seats was almost unchanged. If they held 88 notional seats, that would have been as close as possible to the 1983 result. They held just one more.
The Liberal Party hold a slightly higher number of seats relative to the National Party, but it is more difficult to judge which of the two parties would win seats where both had previously competed. In the end the Liberal Party lost two seats to the National Party and the National Party lost one to the Liberal Party, which closes half the gap between the 1983 result and 1984 boundaries.
But it’s not just about how many seats each party holds, but what margin they are held by. If one party can replace marginal seats with safer seats, that can change the landscape, particularly in the 1983 context when the Hawke government’s first victory was one of Labor’s biggest. Proportionally it was just a bit less than the Albanese government’s landslide victory in 2025.
This chart shows the distribution of Coalition and Labor seats by margin. The shapes look fairly similar, although there are some notable differences amongst the most marginal seats.
Labor won a much higher proportion of the very marginal seats (under 4%) in 1983, but on the 1984 boundaries these seats were split roughly equally. This means other Labor seats must have become safer.
Taking the same data and presenting it in a different way, this graph shows two lines showing the cumulative number of seats with a given Coalition 2PP (or less).
The two lines are almost identical. There are some small deviations amongst safe Labor seats, but the main point of interest is on a Labor 2PP of 52-54%. In general, these seats became better for Labor than they had been in 1983.
This is important because this is the range where Labor would lose its majority. Prior to the redistribution, a 2.15% uniform swing was needed for Labor to lose their majority. After the redistribution, this grew to a 3.1% uniform swing. This could have made a difference if the Coalition had done slightly better. Instead they had a swing of 1.5%.
So the redistribution mostly left the distribution of seats the same, but made Labor’s position a bit better in the case of a very close election. What about geographically? Did the location of marginal seats change?
Not much changed on a state basis. Defining marginal seats as those with a margin under 6%, there wasn’t a lot of change. New South Wales was steady. Queensland had more marginals, as did Victoria and Western Australia, while South Australia had fewer marginals. Across the whole of Australia, the proportion of marginal seats increased slightly.
To get more specific about locations of marginal seats, I’ve produced a map. Seats that had a margin of under 6% (based on 1983 results) are marked in red, while all others are in white. You can toggle between 1983 and 1984 boundaries.
The marginals tend to be in similar areas, but not exclusively so. To take Sydney as an example, the seat of Dundas ceased to be marginal, but only just. Philip Ruddock’s margin grew from 2.7% to 6.9%. There was a band of marginal seats along the NSW coast and wrapping around Sydney, and this band is mostly intact.
Likewise looking at Melbourne, a lot of the marginals are in similar places, but there are differences. Burke becomes more marginal, and McEwen is a new marginal seat.
I was inspired to look into this topic because I often do analysis of marginal seats for clients and this raised the question – will the marginal seats be the same? The answer is, mostly, yes. But not entirely so. Areas that are particularly of interest in federal elections, like the northern suburbs of Sydney, or the eastern suburbs of Melbourne, will undoubtedly continue to be so, but the precise make-up of the seats and the relative marginality of seats within an area will surely change if the parliament is expanded.
Great analysis. Thank you for the detail. BTW, I think you meant PM Albanese, in 2022, not 1983.;)
I hold the view that an expansion of parliament is less beneficial (in the long run) for Labor, but it could affect the Liberal Party’s internal balance. The more competitive seats there are in urban areas, the higher the threshold to form government is, forcing the Liberals to carry more of those seats. Thus, potentially inflating the Moderate and Centre-Right factions in the party to the detriment of the National Right and the Nats in the Coalition. That is probably bad news for Hastie (and others) who wish to run a culture- or social-focused campaign. The most significant difference between 83/84 and a potential expansion in the next 3-4 years is the continued concentration of density in urban areas and the dramatic shift of voters <40 to the left, including the Greens. The Liberals tried various campaign styles and policy offers during the Hawke/Keating years, but to no avail. It took a recession and longevity to finally break through. I suspect that will likely be the case here until at least 2031. If I had time, I would examine the results of the new seats from 1983 through the 87-90-93-96 elections to see if any trends emerged. That is a lot of redistributions; however, the data might get lost.
Agree Craig – although Albanese is from the left faction of Labor that appeared to be ‘socialist’ in nature, he seems to be more in the mould of Bob Hawke rather than Whitlam or Rudd being considered as centrist and is also quite cautious when considering major reforms.
Considering the 1984 new seats, it appears they were created in both outer suburban and regional areas just outside the capital cities. These areas were generally more favourable for the Coalition/Liberal Party at the time, but if repeated today it won’t provide as much benefit for the Liberals because some regional areas that have seen tree/sea changers move in such as Central Coast and Blue Mountains have trended towards Labor.
For the Coalition to win federally in the future – it will require them to move towards the centre, develop a strong economic policy and decouple themselves from the fringe, culture war type issues. This will allow them to recover support amongst the groups that have swung away from them (women, CALD minorities and young voters, many of which are tertiary educated professionals).
Do the Labor factions have much to do with ideology anymore? Chalmers seems to be to the left of Albanese!
@ Craig
Great points i think the longer term trend is that there is a growing Urban/Rural divide. Once exception to this is the growth in alternative lifestyle/tree change communities.
Seats like Calare, Monash/McMillan were seen as swing seats but no longer are. Once upon a time Riverina/Hume were swing seats as was Wannon. Calare is one seat Keating lost in 1996 they have never won back.
There are few seats like Lilley, Moreton and Perth which used to be key swing seats but maybe now safeish Labor seats (Moreton until 1990 was a Coalition leaning seat). Kingston maybe another one that is now safe even if Amanda Rishworth retires Labor can retain it.
I think Labor will support an expansion to secure Whitlam and Werriwa that are talked about as future Labor gains. If Werriwa gains Liverpool/Warwick Farm from Liverpool and sheds McMansion suburbs like Cecil Hills etc it will be more secure for Labor. Labor would want Whitlam to be a purely Illawara based seat an expansion of parliament could enable this.
I think the right faction increasingly likes rural-urban hybrids like McEwen, Canning, La Trobe, Hume, Hawke etc. Labor would like some of these electorates Split into more pure Urban and rural ones. It would make fewer competative seats where appeals to realignments will work.
The Nationals may gain a seat or two in both NSW and QLD.
Nimalan, I think the reason why seats like Lilley and Moreton have trended towards Labor leaning seats is due to changing demographics. I believe during the 1980’s/1990’s era, Brisbane as a city was much smaller and the suburbs contained within Lilley and Moreton were seen as mortgage belt areas.
Now with the growth of the city, the mortgage belt area has now moved outwards away from the core BCC area. As a result, both Lilley and Moreton have many suburbs which have gentrified, and this has led to their leftward trend.
@ Yoh an
Great points
Interestingly, adjusted to 2022 boundaries Lilley and Moreton actually have a better result for the Voice than the Republic in 1999 only 28 seats out 150 did this. So they are likely become more socially progressive with time
Perth is another seats which i think is a bit like Moreton/Lilley as Perth has grown much of the mortgage belt is now in other seats like Hasluck. Perth has also densified. Interestingly, Perth is another seat with a better result for the Voice than the Republic.