No truth stands alone.

How Close Was the 2022 Australian Federal Election?

Recently, someone I met told me that the 2022 Australian federal election was unusually close by historical standards. This made me wonder – was it? And are there any trends in which elections tend to be closest? Here, we order each postwar election based on how close it was on five metrics, and assess how close 2022 in a historical context.

We’ll start off with a metric which measures how close the election was among the voters (the two-party-preferred), then a metric which measures the ease with which the winner formed government (how many seats it won out of the total), and finally a series of metrics which attempt to analyse how close the election was to going the other way. (If you’re familiar with these metrics, click here to jump to the summary)

Pre-publish edit (10/Oct/2022): This is very similar to Dr Bonham’s work after the 2019 election, though I hadn’t realised he updated it for the 2022 election before writing most of this!

Government two-party-preferred

To start us off – how close was the election in terms of how the nation voted? In this first metric, we ask what proportion of voters nationwide preferred the government over the opposition? If you need a refresher on the preferential voting system, or two-party-preferred (2pp), here are a couple of explainers on both:

Under the preferential voting system, voters rank the candidates on their ballot in order of which ones they prefer to be elected first. For example, let’s say we had four candidates running in an electorate, from the Labor, Liberal, National and Democrat parties.

A hypothetical voter might prefer that the National candidate is elected first of all, but if the National can’t win, they would prefer that the Liberal is elected, and then prefer the Democrat candidate over the Labor candidate. This voter would fill in their ballot as such:

  1. FLUGGE, Trevor
    NATIONAL
  2. TUCKEY, Wilson
    LIBERAL
  3. PEEBLES, Shyama
    AUSTRALIAN DEMOCRATS
  4. CHANCE, Kim
    AUSTRALIAN LABOR PARTY

In a House of Representatives (the lower house, where government is formed) election, all ballots are first processed and counted, and a primary vote (or first-preference vote) tally produced. This refers to the % of voters who put one party first. For example, if one in five voters put the National candidate first, then the National Party would have a primary vote of 20%.

Once all ballots have been processed and counted, the candidate with the lowest primary vote is sequentially eliminated, and their voters’ ballots will be transferred to their next preference. For example, let’s say that in this election, each party has a primary vote of:

Liberal 49.5% Labor 25.1% National 23.1% Democrat 2.4%

The Democrat candidate will be eliminated first, and their votes transferred to each voter’s second preference. For simplicity’s sake, let’s assume that half of Democrat voters placed Labor 2nd, while a quarter each placed the Liberal and National candidates second.

Liberal 50.1% Labor 26.3% National 23.7%

Note that this is entirely controlled by who the voters place second on their ballot – candidates and parties do not have any control over where preferences go. A party or candidate may recommend preferences using how-to-votes and other material, but where the ballot travels next is entirely up to the voter.

You may occasionally hear of “preference deals” and “(party) directs preferences to (party)” in the news or other media. This only refers to the parties’ ability to recommend that their voters put Party A over Party B. If a voter decides to ignore this recommendation and preference Party B over Party A, their ballot will go to Party B’s candidate at full value.

The proportion of primary votes for a certain party which are then transferred to another party is also known as the preference flow. In this case, the preference flow for Democrat votes would be 50% Labor, 25% Liberal and 25% National.

While preference flows are referred to as percentages, note that in the House of Representatives, there is no partial vote transfer. If you hear that the preference flow from the Greens to Labor is 80%, that doesn’t mean that 80% of each Green vote goes to Labor. It means that four of five (80%) Greens voters put the Labor candidate ahead of the other candidate on their ballot, while one in five (20%) put the other candidate ahead of Labor.

Preference flows are a useful way to calculate the outcome of a preferential-voting contest. For example, if I told you that in an election, Labor won 48%, the Liberals won 32% and the Nationals won 20%, if you know what the National -> Liberal preference flow is, you can calculate the final Labor-versus-Liberal result in that election.

Speaking of which, let’s finish our example preferential-voting election. As the National candidate has the lowest vote share of the remaining candidates, he is eliminated. Since our hypothetical voter from earlier voted 1 National 2 Liberal, their vote is then transferred to the Liberal. Had they instead voted 1 National 2 Democrat 3 Labor 4 Liberal, their vote would instead be transferred to Labor (as the Democrat candidate has already been eliminated).

For simplicity’s sake, let’s assume that 80% of all voters who voted 1 National or 1 Democrat 2 National then places the Liberal candidate over the Labor candidate.

Liberal 69% Labor 31%

The vote shares of the final two candidates is often referred to as the two-candidate-preferred, or 2cp for short. It is sometimes also referred to as the two-party-preferred; however this can be confusing for reasons explained below.


Because Labor and Coalition parties tend to have the largest vote shares at Australian elections, in most electorates the race comes down to a Labor-vs-Coalition contest (as parties/candidates with smaller shares of the vote get sequentially eliminated).

In these electorates, it’s easy to figure out what share of the population prefers Labor over the Coalition or vice versa, using the two-candidate-preferred (2cp) figure produced through the instant-runoff system.

However, in some electorates, the final two candidates may not consist of a Labor and a Coalition candidate. Candidates from one of the smaller parties (e.g. Greens, One Nation, Katter’s Australian, Centre Alliance) may receive more votes (or more preferences) than one or both major parties.

Additionally, there are instances when the final-two pairing ends up being one Coalition candidate against another.

When the final-two pairing is not a Labor versus Coalition contest, this is known as a non-classic contest.

For all non-classic electorates, the Australian Electoral Commission (AEC) will go back through the ballots, and do a count of how many voters placed the Labor candidate over the Coalition candidate(s), or vice versa.

This allows the AEC to determine what share of voters ‘prefer’ Labor over the Coalition in every electorate. This proportion is known as the two-party-preferred or 2pp for short.

The AEC also calculates statistics such as how many voters for each party placed the Coalition over Labor or vice versa, allowing the determination of two-party preference flows. For example, 82.21% of Greens voters placed Labor over the Coalition at the 2019 federal election.

Two-party-preferred can be quite a useful statistic both at the seat level and nationally. For example, if you live in a non-classic seat with an Independent MP, and said Independent MP is retiring at the next election, you might be interested to know if your seat is more likely to fall to Labor or to the Coalition. In the event of a hung parliament, minor-party and independent MPs may be interested to know if their seat’s voters would prefer that they support a Labor or Coalition government.

Two-party-preferred also helps make sense of changes in primary vote share. If I came up to you and told you that Labor’s primary vote has declined by 3 points, with the Coalition gaining one point, the Greens gaining one point and other minor parties collectively gaining one point, it’s not immediately apparent what that means for most seats (which are Labor-vs-Coalition contests).
On the other hand, if I told you that there’s a two-party-preferred swing of 1.7% against Labor, you can look up the two-party-preferred of all electorates at the last election and get some idea of which seats might fall.

Note that two-candidate-preferred and two-party-preferred are not the same thing. While they can be the same if the final two candidates are Labor and a Coalition candidate, two-candidate-preferred specifically refers to the final vote shares between the final two candidates in an electorate, while two-party-preferred refers to the proportion of voters in that electorate who prefer Labor over the Coalition or vice versa.

In most Australian federal elections, the winner of the popular vote (as measured by the two-party-preferred – still the best proxy of the result) also forms government. The 2022 election was no exception in this regard (2022 outlined in black, higher bars represent larger wins on the 2pp):

Two-party-preferred of the major party which formed govt, 1946-2022
Labor’s 2-party-preferred at the 2022 Australian federal election = 52.13%. However there may be a yet-unconfirmed issue with the Labor-vs-Liberal count in North Sydney which may bring Labor’s 2pp up to 52.14%. Additionally note that the figures for elections prior to 1983 were estimates as the Australian Electoral Commission did not examine all ballots for preferences for those elections.

(to skip to the next metric, click here)

Labor’s 2-party vote at the 2022 federal election is almost exactly the average for all governments across the post-war period (52.13% vs average of 52.04%). This isn’t skewed by some of the larger wins – the 2022 win comes in 17th of 30 elections in terms of closeness, making it a pretty middle-of-the-pack federal election. Notably, federal elections tend to be much closer than state elections on this metric; the major party which formed government after state elections won an average 54% of the two-party vote in state elections from 1986 – 2022.

Additionally, note that of the five instances where a party formed government without winning the two-party-preferred, four have been Coalition governments while Labor has only done so once. We will return to this later, but broadly speaking this reflects a combination of:

  1. The historical tendency for governments to do better than oppositions on vote-seat conversions, as they typically have more incumbents and incumbents tend to out-perform the national swing. As the Coalition has formed government more often than Labor, this generally benefits the Coalition.
  2. The historical tendency for the Coalition to do better than Labor in terms of seats for a given 2-party vote, which we explore further in this piece.

Furthermore, neither side has won greater than 54% of the 2-party vote at any federal election since 1977 – a trend which is clearer if you plot out the winning 2pp over time:

Two-party-preferred of the major party which formed govt, 1946-2022

In a non-proportional system, however, winning more votes isn’t enough for you to form government – five of the thirty postwar elections saw one side forming government without winning the two-party-preferred. It’s just as important (if not more so) how your share of the vote converts to a seat haul, which brings us to:

Government share of all seats

For a party to govern, it needs the confidence of a majority of the House of Representatives (50% + 1 seat), ideally with as much buffer room as possible to pass legislation. To capture that, we can look at what % of seats was won by the major party which formed government (2022 outlined in black, higher bars represent a government which won a greater share of all seats):

Govt share of all seats, 1946-2022

(to skip to the next metric, click here)

Despite the 2022 election featuring a pretty average popular vote win, Labor under-performed by quite a bit when it came to winning seats. This actually makes it quite similar to the Morrison government’s re-election in 2019, whose fairly-average popular vote win converted into a fairly narrow majority of all seats. With just 51% of all seats (77 of 151) in comparison to the post-war average of 57.2% (or median of 56.2%), the 2022 election ties with the 2019 election for 4th closest of the 30 post-war elections. This is to some extent due to the increased number of crossbenchers, seven of whom sit on electorates which preferred Labor over the Coalition on a 2-party basis. In ascending order of closeness on an ALP-vs-L/NC basis: Clark, Melbourne, Fowler, Griffith, Brisbane, Ryan, and Mayo.

You can probably treat the first four of those electorates as “would have been held by Labor in a straight Labor-versus-Coalition contest”. I’m a little more skeptical of treating the last three that way; in particular Mayo has historically been a Liberal-leaning electorate and Rebekha Sharkie isn’t clearly left-aligned the way e.g. the Greens are. Hence, it’s difficult to know how many Sharkie voters who preferenced Labor would have done so had it been a close Labor/Liberal contest without Sharkie in the mix.

A trend which you might notice in the above chart is that Labor has generally under-performed the Coalition when it comes to converting 2-party-preferred wins to seat wins. While the magnitude of Labor’s 2-party-preferred victories are roughly clustered around the average (see the government 2pp graph above), Labor’s share of all seats when in government tend towards the low end of the chart. This tendency is more apparent when you plot the two against each other (trendlines calculated excluding 2022):

Winner 2pp vs winner share of all seats, 1946-2022
Trendlines calculated using logistic regression, and excluding the 2022 federal election to provide some context as to how the vote-seat conversion in 2022 fits in with other elections, historically.

The trend of Coalition government over-performance is slightly weaker if you exclude elections from prior to 1973 (when legislation came into effect reducing the allowable variation in electors per electorate from 20% to 10%):

Winner 2pp vs winner share of all seats, 1974-2022
Trendlines calculated using logistic regression, and excluding the 2022 federal election to provide some context as to how the vote-seat conversion in 2022 fits in with other elections, historically. Elections prior to 1973 were excluded from this plot.

However, while a government’s share of all seats is important, an equally important factor in how close an election is how many seats the opposition won. A government which barely scraped into majority can still be reasonably safe if its opposition failed to win many seats – like the Coalition did in 2022 (winning just 38% of seats). To put this in context, this is the worst share of all seats for the Coalition since the 1943 federal election, where Labor won a record 58.2% of the 2-party-preferred, and the Coalition did so badly that the United Australia Party (the Liberals’ ancestor) dissolved to form the Liberal Party we know today.

(opinion ahead!)
I’m not a big fan of “this horrible election result shows that this party can’t take power for at least another term!” narratives – see the 2012 and 2015 Queensland state elections for why. However, I think it’s fair to say that if the Coalition can’t find a way to retake teal-held seats or flip currently-safe Labor seats, the historical advantage it has over Labor in converting votes to seats is probably significantly reduced if not wiped out going forward.
To account for how the opposition performed relative to the government, we can examine how many more seats each government won relative to its opposition:

Government seat margin over opposition

This metric refers to the number of government seats minus the number of opposition seats. So for example, in 2022, the Labor government won a 19-seat margin over the Coalition opposition (77 Labor – 58 Coalition). 2022 is outlined in black below; higher bars represent governments which won a lot more seats than their oppositions:

Govt seat margin over the opposition, 1946-2022

(to skip to the next metric, click here)

When you consider the relative position of the government to its opposition, 2022 comes out looking fairly average, with the Labor government’s 19-seat margin being just a touch under the mean 20.5 seat margin (and being a little higher than the median 17.5). Amongst the post-war elections, the 2022 election would rank 17th closest of 30 – a fairly average performance. Although Labor just scraped into a majority, the Coalition’s loss of several former-safe Liberal seats to independents and an expanded House of Representatives In 1946 – Labor’s largest post-WW2 popular vote win – the House of Representatives was comprised of just 74 members, while in 1983 (Labor’s 2nd largest vote win since WW2), the House of Representatives consisted of 121 members. By contrast, the House of Representatives consisted of 151 members at the 2022 election, making it theoretically easier to win more seats over your opposition for a given popular vote win.

Seat margin is therefore theoretically skewed towards more recent elections, with the larger legislature at each. However, I still think it’s more informative than looking at seat margin as a share of total seats – for example, a government with a 2.7% seat margin over its opposition has very little buffer in a 74-seat House (being a 2-seat margin) but has a bit more buffer in a 151-seat House (as it would have a 4-seat margin).
contributed towards the government winning a pretty typical seat margin over its opposition.

Additionally, when examining the relationship between votes and seat margins – as opposed to raw seat % – Labor’s performance in 2022 comes out much better. While historically Labor has also under-performed on converting its vote into a large seat margin, Labor’s win in 2022 is actually closer to the vote-seat margin relationship for the Coalition than Labor’s historical performance on this front:

Australia,Election,Federal election,Closeness,Electoral history,2022 Australian federal election,2022 election,Two-party-preferred,2pp,Government,Vote share,Seats,Seat margin,Confidence margin,Swing to change
Trendlines calculated using logistic regression, and excluding the 2022 federal election to provide some context as to how the vote-seat margin conversion in 2022 fits in with other elections, historically.

However, with increasing numbers of elected representatives from outside the two major parties, simple seat margin between the government and its opposition may become less applicable as a measure of election closeness. In the event of a very close election, a hung parliament where no party or formal coalition commands a majority becomes reasonably likely. In that situation, members of the House may decide to express confidence in a party which does not have the most seats, as long as said party can command the confidence of a majority of the House. For example, Labor could plausibly form government in a House with 69 Labor, 71 Coalition, 6 Greens and 5 Independents. If all Greens and a single Independent provide support on confidence and supply to Labor (promising to support a Labor-led government in a vote of no confidence and pass bills to acquire the funds needed to keep government running), then Labor would be able to form government. This can render seat margin less useful, especially whenever the members from outside the government/opposition (i.e. the crossbench) are known to lean towards one side or another.

Government’s expected confidence margin

To account for this, we can assume that crossbenchers who are affiliated with one side of politics or who tend to vote with one party will probably vote for that party to form government (vote of confidence), in the event no one gets a majority. Then, we can subtract the expected votes of confidence for the government from the expected votes of confidence for the opposition to get each government’s expected confidence margin in the House, which takes into account the expected leanings of the crossbench.

(description of classification methodology below, and you can read through my reasoning for each classification here)

Crossbenchers were assigned as either left-leaning (defined as being likely to support Labor in a hung parliament), right-leaning (defined as being likely to support the Coalition in a hung parliament), or neither.

The default assumption was to assign a crossbencher as “neither” and re-assign on the basis of (roughly in descending order of importance):
  1. On-the-record statements about which party they’d align with.
  2. Greens were assumed to prefer aligning with Labor, while One Nation (Hanson) and Palmer were assumed to prefer aligning with the Coalition.
  3. Affiliations, past and present. Often an independent will explicitly affiliate with one side of politics despite not having been part of a major party – e.g. socialists, nationalist groups etc.
  4. History of relationship with the major parties. Sometimes, an independent will have won a seat after leaving a party due to ideological differences with said party. If this was documented as being the case, then it nullifies the previous consideration – if someone quits Labor and complains about it being too left-leaning, they should not be automatically considered left-leaning.
  5. The electorate in which the crossbencher was elected. For example, if a crossbencher with a history of affiliation with left-leaning groups was then elected to an electorate which was 65-35 on a Labor/Liberal basis, I would expect them to align with Labor in the event of a hung parliament if possible.

In 2022, a total of four Greens were elected to the House, which would be expected to support Labor in the event of a hung parliament. Unsurprisingly, as a result, the 2022 election is less close on this metric than simple seat margin:

Govt confidence margin over opposition, 1946-2022

(to skip to the next metric, click here)

The expected confidence margin for the Labor government is 23 seats, which compares favourably to the mean 20.9 seats and to the median of 17.5 seats. On these estimates, the 2022 election would rank 19th closest of 30 among post-war elections.

It’s also worth noting how different 2022 and 2019 are on seat margin and confidence margin versus the other metrics. In terms of simple popular vote, they’re quite similar (2022: 52.1%, 2019: 51.5%) and both governments won the exact same seat majority (77/151). However, they could not be more different in terms of how close the opposition was to catching up with the government, with 2019 being one of the closer elections in this regard (9 seat margin, 8 confidence margin) while 2022 being marginally safer than average (19 seat margin, 23 confidence margin).

Estimated swing to change result

However, even confidence margin still doesn’t fully capture how safe a newly-elected/re-elected government was from failing to get elected. A government with a seat/confidence margin of +5 but whose closest five seats were won on large margins is quite a bit safer than a government with the same seat/confidence margin but whose closest five seats were very narrowly won.

Hence, it’s also worth having a look at – what kind of swing would have given the opposition even-odds of winning more seats than the government? There’s a few ways to do it, the most common being to look at the pendulum, and assume that a uniform swing would knock out any seats on a sufficiently small margin – e.g. if there’s 2 seats held by the Coalition on a margin of less than 1%, uniform swing assumes that a 1% swing to Labor would deliver 2 seats to Labor.

Here, I use a slightly more complex model which accounts for the uncertainty in how any swing is distributed (i.e. the conditional probabilistic model). Instead of assuming that a seat “flips” from one party to the other, it estimates a probability that a seat is won by either side given a certain swing – because swing does not affect all seats equally. So for example, if the modeled result in a seat is 50.1% Liberal, uniform swing would assume that that seat is now ‘won’ by the Liberal. On the other hand, the probabilistic model would predict (retrodict?) that the Liberals would win this seat a little over half the time while Labor would win this seat a little under half of the time.

Normally, both methods will produce very similar results. However, if one side ends up with more seats on very close margins than the other, the probabilistic method will estimate a lower swing-to-change than the uniform swing methods, as it would predict that some of those very-close seats would be won by the other side due to no swing being fully uniform (see example in footnote). For example, in 1998, a swing of 0.6% to Labor would have left Labor with 5 seats on margins less than 1%, while the Coalition would have had 11 seats on margins less than 1%.

A simple uniform swing would tell you that the Coalition would win 76 seats and Labor 70, but a probabilistic model would expect the Coalition to win 73.2 and Labor 73.8 on average. Due to the greater numbers of marginals on the Coalition side, the probabilistic model expects some of them to fall if Labor had done better due to the random distribution of any swing, and hence predicts that Labor would not require as big a swing to have even odds of winning more seats than the Coalition.
I have used David Barry’s data to estimate a probabilistic swing to change the result for elections 1946-1990 (code here), and my probabilistic pendulum for elections 1993-.

So how does the 2022 federal election compare to historical federal elections on this metric? Well…it’s complicated:

Probabilistic 2-party swing to change result, 1946-2022

The difficulty with modelling 2022 is that a lot of candidates outside the two major parties won seats, but many of these crossbenchers won fairly narrow victories to which major-party preferences contributed. Hence, if there had been a 2-party swing from one major party to another, it might have been harder for the crossbencher to be elected – e.g. if there was a 4% swing from Labor to the LNP, it probably would have been harder for the Greens to flip Ryan (which they won on a 2.7% margin against the LNP, after preferences).

Bearing this in mind, I’ve presented two estimates for the swing-to-change in the 2022 election. The first assumes that 2-party swing does not impact the outcome of non-classic (major party vs minor party/independent) contests as with other elections; on that methodology, I estimate a swing-to-change of 3.5% for 2022, which puts it above the average swing-to-change estimate (2.8%) and would make it the 20th closest election out of 30.

In my other estimate, I assume that the 2-party swing directly impacts any contests which were major-vs-minor/independent. To continue using the example of Ryan, in this model, a 2-party swing of 1% to the Coalition would reduce the Greens’ margin to 1.7% while a 2-party swing of 1% to Labor would increase the Greens’ margin to 3.7%. Using this methodology, the swing-to-change estimate declines to 2.8% for 2022, due to the number of crossbench seats won from the Coalition on narrow margins (e.g. Curtin 1.3%, Mackellar 2.5%, Ryan 2.7%). This would put the 2022 election exactly on the average swing-to-change for all post-war elections, and make it the 18th closest election of 30.

Swings between Labor and the Coalition usually don’t have a one-to-one relationship with contests between a major party and a minor party/independent, so the best estimate of swing-to-change is probably somewhere between these two estimates.

So overall, how close was the 2022 election?

To sum it all up, the 2022 election ranked:

Or, in graphical terms (2022 highlighted, the average for each metric is plotted as a horizontal line in each graph):

Australia,Election,Federal election,Closeness,Electoral history,2022 Australian federal election,2022 election,Two-party-preferred,2pp,Government,Vote share,Seats,Seat margin,Confidence margin,Swing to change

The 2022 election was therefore fairly average on most metrics of election closeness, as far as Australian federal elections go. The one exception was in terms of the government’s fairly small majority, a metric which will likely become less relevant to government formation unless recent trends of an expanded crossbench are reversed. For example, if the Greens continue gaining seats, we may one day see a situation where Labor forms government with fewer seats than the Coalition, similar to how the Country Party backed in a Nationalist-led government in the 1922 federal election despite the Nationalists winning 3 seats fewer than Labor.

News coverage of the election, however, seems to be lagging in this department, with much of the post-election coverage focused on whether Labor could get a majority when there was no way for the Coalition to form government even if Labor came up just short of a majority. In combination with delays due to how we count postal votes, this probably made the 2022 election seem a lot closer than the 2019 election, despite the Coalition being in a worse position in 2022 than Labor was in 2019.


Add Your Comment

× Home Politics Economics Projects
×

You appear to be using an outdated browser, for which this site is not optimised.

For your security, we strongly recommend you download a newer browser.

Although any of the latest browsers will do, we suggest the latest version of Firefox.

Download Firefox here

(Click on the button in the top-right to close this reminder)