We hear a lot of discussion in the media about 'marginal seats' and political parties generally use the term to indicate the constituencies where they may win or lose ground. But what makes a seat marginal? And why do they attract the attention of political pundits?
Professor John Wanna from The Australian National University (ANU) explains the difference between marginal and safe seats, why the former can determine who wins government, and which seats are worth keeping an eye on during the 2022 federal election.
What is a marginal seat?
The 'margin' in marginal seats refers to the winning proportion of the formal vote the successful candidate received at the previous election.
The term 'marginal' is not typically a projection of what people might expect, but a statistical historical outcome from the last election or by-election. It is calculated by examining the gap between what the winning candidate received (after adding the other preferences of less-well supported candidates) over and above the 50 per cent needed to win the seat, giving us the so-called 'two-party preferred vote'.
Let's say a seat has a margin of four per cent. This means the winning candidate at the last election won eight per cent more of the vote than the losing candidate - for example 54 per cent to 46 per cent. The margin becomes half the per cent the winning candidate won by. At the next election this means the winner can only afford to lose four per cent of the vote that goes directly to the main challenger. If they lose any more of than half the margin, they'll lose to the main challenger. It's vital to remember that often swings away from sitting members do not initially go to the main challenger, but will then usually come back via the allocation of preferences.
The notion of a 'marginal seat' is a contentious concept and there is no universally agreed definition of what actually constitutes a marginal seat. The identification of seats with small margins, say between two major parties or rival candidates, is only relevant in political systems that have electorates in which a single winner is elected; and not in electoral systems where multiple candidates can be elected from a given electorate. That's why the concept of marginal seats tends to be associated with two-party systems and especially Westminster political systems, for example in the United Kingdom, Canada, Ireland, Australia, and historically New Zealand.
How marginal is a marginal seat?
For some political parties a margin of less than two per cent is enough to label a seat as marginal. However, a loose consensus has developed, especially by electoral commissions, where any seat with a margin of four per cent or less is classified as marginal.
Many commentators accept that this is too narrow a definition, and that if a swing for or against a government at the next election is likely to be high, then a margin of five, six, seven, or even up to 10 per cent can be considered marginal. I personally know politicians who have sat on a margin of seven per cent or over and regarded their seat as marginal.
What's a safe seat?
Marginal seats can be contrasted against 'fairly safe seats' and 'very safe seats'. These are seats where the winning candidate at the previous election received at least 56 to 60 per cent or the formal vote for fairly safe seats, and more than 60 per cent for very safe seats.
Political parties tend to preselect party 'favourites' or politically important identities for these safer seats. When election time comes around, the major parties tend to place less emphasis on holding or winning safer seats because they are considered out of reach of political opponents.
Why are marginal seats important?
Why then do pundits tend to focus on marginal seats, and why are they important? Because the margin between the winner last time and the loser is close, election watchers concentrate their attention on these seats that can easily change hands with a slight swing.
Accordingly, marginal seats are often considered the principal seats where a change of government could occur, although this isn't always the case. Marginal seats may be won by the same party at successive elections, and safe seats may change.
In other political systems with single-member electorates the term 'marginal seat' is not used, and instead there is discussion of 'swing seats' or 'swing states' or even 'seats in play'. But essentially these terms mean the same thing - pollsters have identified seats that have relatively small margins between the major protagonists, and consequentially pay closer attention to them because the outcome here can easily affect the overall outcome of the election.
In short, governments that lose their marginal seats often lose government; whereas oppositions that win more marginal seats can then be in a position to win government.
In any Australian election at any level of government attention will almost exclusively be on the 'marginal seats' with media features and speculation of who might win. Opinion polling also plays a factor here as well. Safer seats are likely to receive very little attention save from the local media outlets and local campaign activities.
How many marginal seats are in play in 2022?
The 2019 federal election was very tight, and the final result not widely expected. The Scott Morrison-led Coalition parties won 77 seats in the House of Representatives from a total of 151; Labor won 68 seats, and six were won by smaller parties or Independents.
The Coalition currently holds nine marginal seats on a margin of less than four per cent, whereas Labor holds 14 seats with a similarly small margin. Two Independents hold relatively marginal seats as well.
Some of the tightest seats based on the results of the 2019 federal election include Macquarie in the NSW Blue Mountains, which Labor holds by 0.2 per cent; Bass in northern Tasmania, which the Liberals hold by 0.4 per cent; the NSW seat of Eden Monaro on 0.4 per cent to Labor; Lilley in Brisbane's northern suburbs, which Labor holds by 0.6 per cent; and Cowan in Perth's northern suburbs, which Labor won by 0.9 per cent. Liberal seats of Chisholm in Melbourne (0.6 per cent), Wentworth in Sydney (1.3 per cent) and Boothby in Adelaide (1.4 per cent) are highly marginal Liberal seats.
What are other seats to watch in 2022?
While some candidates won't be confirmed until close to when the election is called, there are some seats already of intrinsic interest.
The activist group Climate 200, which has raised more than $7 million in donations, is financially supporting the campaigns of candidates challenging predominantly moderate Liberals in up to 25 seats, and one Labor seat in the ACT electorate of Bean. Some Climate 200 candidates are well-known locals or identities, some of whom have stood in previous elections as Independents or Labor candidates.
Climate 200 candidate Allegra Spender will challenge Liberal MP Dave Sharma in the NSW seat of Wentworth. Sharma won the seat in 2019 with a margin of 1.3 per cent over Independent Kerryn Phelps.
Professor Monique Ryan is challenging Treasurer Josh Frydenberg in the seat of Kooyong. The seat was considered a very safe seat for the Liberals, who went into the 2019 election with a margin of 12.8 per cent. That margin was reduced to 5.7 per cent over the Greens in the last election.
Craig Kelly has held the seat of Hughes in Sydney since 2010. It was a safe Liberal seat at a margin of 9.9 per cent, but in 2021 Kelly resigned from the Liberal Party and will contest the 2022 as a candidate for the United Australia Party. The major parties have been scrambling to find candidates to challenge him.
Several MPs have announced they will not contest the 2022 election, in some cases vacating seats that have been held by the same member for decades. Moreover, a redistribution by the Electoral Commission has affected seats in two states: Western Australia, which lost one seat, and Victoria, which gained one.
Labor MP Warren Snowdon, who has held the seat of Lingiari in the Northern Territory since 2001, announced in 2020 that he was retiring from politics at the next election. Lingiari was previously been a safe Labor seat, but the margin at the 2019 election dropped to 5.5 per cent. Marion Scrymgour, the first Indigenous woman elected to NT parliament, will contest the seat for Labor, and former Alice Springs mayor Damien Ryan for the Country Liberal Party.
The retirement of Greg Hunt opens up the race for Flinders in Victoria. The margin at the 2019 election was 5.6 per cent. Liberal candidate Zoe McKenzie and Independents Despi O'Connor and Sarah Russell are contesting the seat in 2022.
Professor John Wanna is an Emeritus Professor in political science and the former Sir John Bunting chair of public administration at the ANU College of Arts and Social Sciences.
Read more federal election coverage from ANU: