The 2022 Senate election marked a new high point in support for minor parties and Independents. The long term trend of declining support for major parties continued and passed a new milestone. For the first time both major parties were outpolled by the combined vote for minor parties and independents.
Senate ‘Other’ vote reached 35.7% against 34.2% for the Coalition and 30.1% for Labor. Senate non-major party support has been higher than Labor’s vote at every Senate election since 2013, but 2022 was the first where it was also higher than the Coalition.
In the House of Representatives, minor party and Independent support remained in third place though at a record level. The Coalition polled 35.7% in the House, Labor 32.6% and all other candidates 31.7%. While support for ‘Others’ reached record levels in both chambers, the gap between support in the two chambers narrowed.
The one-third splits in Senate support did not translate into one-thirds representation. The Coalition elected 15 Senators, Labor 15, and all other parties 10. Within vote for others, the Greens elected six Senators with 12.7% of the vote while the 23.0% support for the rest elected only four Senators.
This discrepancy is down to the nature of the Senate’s electoral system. Support for the Coalition, Labor and the Greens was confined to a single ticket in each state and territory. (There was a second but very low-polling National ticket in SA.) Support for non-Green ‘Others’ may have been at 23.0%, but it was spread across 126 groups plus numerous ‘ungrouped’ candidates. As smaller parties were excluded, their preferences did not always flow to other smaller parties.
Under the group voting ticket system abolished in 2016, party negotiated deals allowed small parties to aggregate their vote. The abolition of tickets returned control over preferences to voters, and three elections since the change have revealed that voters make different preference choices to those produced by the now abolished tickets. The new system has essentially diminished the influence of preferences and made the system more proportional to the level of first preference vote in each state.
The Senate’s electoral system now effectively operates like list proportional representation with final seats allocated to groups with the highest partial quotas on first preferences. The election of Independent David Pocock in the 2022 ACT Senate race shows that preference can still determine who is elected. But such exceptions don’t undermine the basic nature of the Senate’s reformed electoral system – it advantages parties with primary votes over parties that rely on preferences.
Inside this post, I take a closer look at the national voting patterns, and also assess how the electoral system translated votes into the seats.
Trends in Senate Vote Share
The chart below plots support for Labor (red), the combined vote for the Coalition (blue), and a total vote for ‘Others’ (grey) including the Greeens. I’ve added a separate line for the Greens which I’ll discuss below.
(Note: all tables in this post use the original 2013 WA Senate result rather than the 2014 WA Senate re-election. The totals for 2016 are also based on original results, not the re-calculated totals produced by re-counts following several disqualifications of Senators.)
I've selected 1993 as the starting point for the chart as it was an election where minor party support was squeezed in a polarising campaign over the Coalition's 'Fightback!' package. Support for minor parties declined from a record 19.7% in 1990 to only 13.5% in 1993.
Support for both Labor and the Coalition has declined since, but Labor's decline has been steeper. In part that is due to Labor winning fewer elections than the Coalition in the period covered by the graph. But the separate line for the Greens shows that some of the Labor decline is explained by the growth of the Greens. Some of the early growth of the Greens followed the demise of the Australian Democrats, but since 2010 it is hard to argue against the case that Labor's lower vote than the Coalition is down to Labor having lost a significant portion of support to its left-wing rival.
Labor outpolled the Coalition on the re-election of the Keating government in 1993, but the only election since where Labour outpolled the Coalition was on Kevin Rudd's sweep to victory in 2007. The only other election where the two major party groupings were close on first preferences was in 1998 when One Nation polled 9.0% in the Senate, largely drawing its support from the Coalition. At elections contested by One Nation since, the party's support has been lower and its source less easily identified.
The chart below shows support for 'Others' at both House and Senate elections since 1975. The 1975 election was overwhelmingly a two-party contest in a period without significant third parties. The Democratic Labour Party (DLP) had polled between 6.1% and 11.1% at Senate elections between 1955 and 1970 but largely disappeared after the election of the Whitlam government in 1972. It was not until 1977 that the Australian Democrats emerged and created a new era of third party politics.
As the chart shows, 'Other' support has always been higher in the Senate than the House. This is largely because there are more 'Other' options on Senate ballot papers, with many smaller parties choosing to only contest the Senate. The dashed black line shows the difference between 'Other' support in the Senate and the House.
The large 10.9% gap in 1984 was due to the Nuclear Disarmament Party. It attracted 7.2% of the Senate vote but did not contest the House election creating a larger gap between 'Other' support in the two chambers. The gap has only passed 10% twice since, 11.1% in 2013 with the giant ballot papers at the last election using group voting tickets, and 11.5% in 2016 when there were again giant ballot papers at that year's double dissolution election. Group voting tickets may have been abolished in 2016, but the lower double dissolution quota provided encouragement for minor parties to contest the election.
The gap has since declined. The reason why can deduced from House and Senate lines. Support for Others in the Senate has stabilised since 2013, but support in the House has continued to increase. In large part this has been due to parties like One Nation, United Australia, the Liberal Democrats contesting more House seats. With voters having the option of voting for the same minor party in both chambers, the between-chamber difference in 'Others' has declined.
The chart below shows the decline in the number of parties and groups contesting the Senate since a peak in 2013.
(The number of Territory ballot paper columns since 2013 have been 13, 10, 7 and 11 for the ACT, and 12, 7, 9 and 8 for the Northern Territory.)
While the number of candidates increased at the 2016 double dissolution due to their being more vacancies and a lower quota, the number of groups on the ballot paper declined. The numbers have continued to decline, the largest state NSW down from 44 groups in 2013 to 23 in 2022, the fewest since 1998. Tasmania, the smallest state, also saw a declined from 23 groups in 2013 to 14 in 2022, the average across the six states following the same pattern.
This decrease is due to the abolition of group voting tickets. The ability of parties to control preferences using tickets meant there was no penalty in the old system for small parties running against each other and splitting support, as long as the small parties swapped preferences with each other using group voting tickets. Tickets also encouraged parties to contest every state and to trade preferences based on different deal beneficiaries in each state. In 2013 under ticket preferences, 29 parties contested four or more states, but only 12 did so in 2019 under the new system.
With 90-95% of votes locked into group voting ticket deals, support for 'Other' parties could be aggregated and result in parties electing Senators from a very low vote. These games ended with the 2016 reforms, with voters filling in their own preferences and producing much weaker flows of preferences between parties.
The new system means that unless a party can accrue a significant first preference vote, it will be excluded from the count and have the votes it holds distributed as preferences. Weak flows of preferences then hinder the ability of parties to come from behind and after preferences.
Allocating Seats to Parties
The Senate's electoral system is known in academic literature as Proportional Representation by Single Transferable Vote, or PR-STV for short. Much of the literature on PR-STV is based on the Irish version where results are determined by two types of preference transfers. These are transfers "within party" for candidates of the same party, and "between party" to candidates of other parties. In Ireland, preferences are fully optional so exhaustion of preferences between candidates means parties only stand as many candidates as they can hope to elect. Irish candidates are also printed as a single list and not grouped by party.
PR-STV for the Australian Senate uses a very similar candidate based counting system as Ireland, but the structure of Senate ballot papers means the issue of within-party preferences can be ignored. The existence of 'above the line' voting and the grouping of candidates by party means that "within party" preference flows approach 100%. Votes are still counted and distributed by candidate, but for analysis purposes the votes by candidate can be accumulated to a single party total. Even in the rare cases where a large below-the-line vote results in party candidates being elected out of order (e.g. Lisa Singh from bottom of the Tasmanian Labor ticket in 2016), the number of Senators who will be elected can still be predicted from the party total.
By comparing Senate PR-STV results to possible outcomes under a simple List-PR with final seats filled by highest remainder, you can examine the impact that preferences have on the outcome. Elected members can be placed in one of the following List-PR categories.
- Filled quotas - seats allocated to groups based on filled first preference quotas
- Highest remainder - seats allocated based on having the highest partial quota on first preferences
- Trailing wins - candidates elected from a trailing partial quota. Only possible in preferential systems.
These categories can best be explained with reference to the ACT Senate election. On first preferences, Labor's total vote elected a member with a filled quota, the party polling 1.0012 quotes. Labor's second candidate had to be excluded and preferences distributed before lead candidate Katy Gallagher was elected, but the seat is still treated as a filled quota. Liberal Zed Seselja had the highest remainder with a Liberal partial quota of 0.74, but Independent David Pocock became the only trailing winner at the 2022 election after starting on 0.64 quotas and passing Seselja on preferences.
Table 1 shows the number of elected Senators in each category. The 2013 half-Senate election was the last with group voting tickets while 2019 and 2022 used the new system. The lower quota 2016 double dissolution is not included. Totals for states and territories are shown separately given the different dynamics between elections for six (state) and two (territory) Senators.
The 2013 election with group voting tickets saw a quarter of state Senators elected from trailing positions, 9 of the 36 seats, more than the six Senators elected from a highest remainder. The higher number for trailing wins shows how much preferences were distorting the result compared to first preference proportionality. Some of the trailing wins were extra-ordinary, including the race in Victoria where the Australian Motoring Enthusiasts Party's candidate Ricky Muir was elected with only 0.04 quotas, preference harvesting allowing him to defeat the third Liberal candidate who began the count with an 0.81 partial quota.
Under the new Senate system in 2016, only two of the 72 state Senate seats were won by trailing candidates, and none out of 36 in both 2019 and 2022. The only trailing win in 2022 was David Pocock in the ACT, a race where the early election of a Labor Senator with a filled quota turned the race for the second seat into the equivalent of a single member preferential race.
Table 2 below shows the filled quotas by state and party, the first preference partial quotas for successful parties, and the partial quotas for the final unsuccessful parties. The partial quotas are for first preferences, not the final partial quotas after the distribution of preferences.
To explain what the quotas mean, in NSW the Coalition polled 2.57 quotas on first preferences, electing two Senators with filled quotas and a third from a 0.57 partial quota. The Greens polled 0.80 quotas in total and elected one Senator from this partial quota. One Nation missed out as the next highest partial quota of 0.29. The Greens reached a full quota during the distribution of preferences, while with all preferences distributed, the Coalition's third Senator was reached 0.86 quotas to defeat One Nation with 0.69.
|State||Filled Quotas Elected||Partial Quotas Elected||Partial Quotas Not Elected|
|NSW||2||Coalition||0.80||Greens (+)||0.29||One Nation|
|WA||2||Labor||0.998||Greens (+)||0.24||One Nation|
|SA||2||Liberal||0.84||Greens (+)||0.28||One Nation|
|TAS||2||Liberal||0.89||Labor (+)||0.27||One Nation|
|1||Labor||0.61||Lambie Network (+)|
|ACT||1||Labor||0.64||David Pocock (+)||0.74||Liberal|
Note: (+) indicates parties that reached a quota on preferences. One Nation passed Labor on preferences to win the fifth seat in Queensland. Bold for David Pocock in the ACT indicates the one victory for a trailing candidate.
As the table shows, the highest quota of any unsuccessful candidate was 0.74 for the Liberal Party in the ACT, passed by David Pocock on 0.64 quotas for a trailing win.
The next highest party to miss out was the LNP's third candidate with 0.47 quotas in Queensland, a case where four parties were in the running for three seats and the LNP had the lowest vote. Something similar happened in Queensland in 2019 when four candidates were also in the race for three seats and Labor's second candidate missed out despite the party having a partial quota of 0.58.
The lowest partial quotas to elect a Senator were Labor's third candidate in WA 0.42 quotas, the third Liberal in SA with 0.37 quotas, and the UAP with only 0.28 in Victoria. The last seat in Victoria was an unusual race as the count began with five parties with between 0.2 and 0.3 partial quotas on first preferences and the UAP maintained its narrow lead to the end. Slightly wider initial leads of 0.09 in SA and 0.18 in WA were too wide to bridge on preferences.
The 2022 result confirms 2019 conclusions that the reformed Senate electoral system tends to allocate seats to parties proportional to first preference support. The system behaves like List PR with a highest remainder method of allocating the final seats. While greatly diminished in their importance, there remains a limited possibility that preferences can determine who wins the final seat in each state and territory. This is especially the case if first preference leave parties in close competition for a seat, and also if any of the excluded parties are capable of delivering a strong flow of preferences.