elections

Why NSW Labor’s Election Night Majority Disappeared

On the night of the 2023 NSW election, I along with most other observers had expected that Labor would achieve majority government.

After midnight I turned off the ABC’s predictive tools and assessed every seat solely on the votes counted. The tally I came to had Labor on a certain 45 seats, ahead in another four seats, and based on past trends had a reasonable chance of winning three seats where the Coalition was ahead on votes.

The next morning I rang the Sydney news room suggesting that news reports refer to Labor winning but step back from stating Labor had achieved a majority. On the seats remaining in doubt, you would have expected Labor would win at least two of the doubtful seats to achieve majority government. That was the position I reported on ABC news on the Sunday night.

This did not happen. Labor won only the 45 seats I had marked down as definite Labor wins on my Sunday morning check. In the four seats where Labor was ahead, and every other seat that was close, Labor went backwards with each day’s counting.

On Monday 27 March, the uncounted pre-poll centres were added to the count. A trend against Labor emerged and on the Monday evening I tweeted that Labor would probably miss out on a majority and win 45 or 46 seats. In seat after seat the addition of pre-poll votes on Monday had revealed a consistent decline in Labor’s position.

That trend went even further on the Saturday after polling day when the largest batches of postal votes were added in key seats. For the first time the Liberal Party pulled ahead in Ryde and moved even further ahead in Terrigal. Both seats had looked like Labor gains on election night.

As an election analyst, such post election night shifts set you up for criticism. You are accused of not taking into account that pre-polls and postals would favour the Coalition.

Actually, the model I use builds in a correction for postal and pre-poll voting trends. The model factors in the postal and pre-poll vote trends from the last election.

What happened in 2023 is that the 2023 pre-poll and postal vote results turned out to be very different to the election day results. The trend to the Coalition in post-election counting was much larger than in 2019.

There were also many more pre-poll and postal votes in 2023. The early vote broke more strongly for the Coalition than previously, and the impact of these votes was amplified in the final result by their greater weight of numbers.
Read More »Why NSW Labor’s Election Night Majority Disappeared

Increase in Voters Completing Preferences at the 2023 NSW Legislative Council Election

In 2000 New South Wales became the first state to abolish Group Voting Tickets (GVTs), the system then generally used to elect state Legislative Councils and the Commonwealth Senate.

The NSW decision followed the 1999 Legislative Council election and its infamous “tablecloth” ballot paper. Confusion combined with labyrinthine preference deals made a mockery of any claim that the filling of the final vacancies reflected the will of the electorate.

The system adopted abolished GVTs and introduced a new form of voting above-the-line (ATL) where voters could direct preferences to other parties on the ballot paper by numbering ATL boxes. A single ‘1’ ATL vote was still formal, but a voter could direct preferences to others groups with an ATL voting square by indicating ‘2’, ‘3’ etc to other groups.

With Senate elections continuing to use GVTs where only a single ATL preference counted, few voters made use of the new ATL voting option at Legislative Council elections. At four elections from 2003 to 2015, more than 80% of Legislative Council ballot papers continued to be completed with a 1-only ATL vote and only around 15% of voters indicated further ATL preferences.

When the Commonwealth followed NSW in abolishing GVTs ahead of the 2016 election, it adopted different instructions on how to complete an ATL vote. Senate ballot paper instructions, and advice from ballot paper issuing officers, suggested a minimum of 6-ATL preferences be completed.

The Senate reforms included generous savings provisions permitting ballot papers with fewer than six ATL preferences to remain formal. At three Senate elections since 2016, more than 95% of ballot papers have had six or more preferences, around 80% having exactly six.

Experience with the new Senate ballot paper has clearly encouraged more voters to indicate preferences on NSW Legislative Council ballot papers.

As the chart below shows, only around 15% of voters completed ATL preferences before the Senate changes. At two NSW elections post the Senate changes, 27.6% of NSW voters completed ATL preferences at the 2019 Legislative Council election, and after experience at two further Senate elections in 2019 and 2022, the percentage of ballot papers at the 2023 Legislative Council election completed with ATL preferences rose to 39.2% .Read More »Increase in Voters Completing Preferences at the 2023 NSW Legislative Council Election

Background Paper on the 2023 NSW Election

I’ve prepared a preview publication on the NSW Legislative Assembly election for the NSW Parliamentary Library.

I won’t claim its an exciting publication. It’s a reference work that tabulates, for each electorate, results at the four elections from 2007 to 2019, plus references to the 2015 and 2021 redistributions.

The period covers four elections, beginning with a comfortable Labor victory in 2007, then tracking results through the landslide Coalition victory in 2011, then the 2015 and 2019 elections where Coalition support ebbed away.

Leaving the finely balanced Legislative Assembly that faces the electorate on 25 March.

You can find the publication at this link.

The analysis uses the four elections to categorise 2023 electorates as being safe for the Coalition or Labor, or the sort of electorates that change sides.

Unsurprisingly, the categories of electorates line up with the 2023 electoral pendulum.Read More »Background Paper on the 2023 NSW Election

VIC22 – Werribee – Analysis of Preferences

Werribee was one of the seven districts where the Victorian Electoral Commission (VEC) conducted its re-check of election night votes by data entering ballot paper preferences into a computer system. Werribee had 15 candidates, and as explained in my Melton post, this presented difficulties for the VEC in conducting manual re-checks so data entry was used.

Werribee was of interest in 2022 after the second place finish of Independent Joe Garra in 2018. Werribee had a very safe 13.4% margin versus the Liberal Party, but slightly less safe 9.2% margin versus Garra. Labor MP Tim Pallas was saved from another challenge when Garra contested Point Cook after his home suburb of Werribee South was transferred to Point Cook in the redistribution. But a new Independent emerged in local car dealership owner Paul Hopper. In the end Hopper’s challenge fizzled and he finished fourth with only 5.9%.

With 15 candidates, Werribee saw a very high informal vote at 9.7% on a turnout of 85.6%. A very low 30.1% of votes were cast on polling day, 57.3% were Pre-polls, 8.2% Postal votes and 3.8% Absent votes.

The two major parties attracted 70.7% of the vote between them, Labor 45.4%% and the Liberal Party 25.3%. The other 29.3% was split across 13 candidates, none of whom passed 7%.

As with my previous posts on Northcote, Preston, Hawthorn, Brighton, Melton and Point Cook, this post will use of the electronic ballot papers to analyse preference flow statistics and also to look at the influence of candidate how-to-votes.

The key findings for Werribee are –

  • The Liberal Party how-to-vote, which switched in 2022 to recommend preference for the Greens ahead of Labor, resulted in 71.5% of Liberal voters following the recommendation and putting the Greens ahead of Labor. Note that Liberal preferences were not distributed.
  • For the 12 Point Cook candidates that registered how-to-votes indicating preferences, 25.6% of ballot papers exactly matched the how-to-vote of the chosen first preference party. Labor at 26.9% and the Liberal Party 34.6% were the highest concordance rates, with Independent Hopper third at 27.4%. Fewer voters for the plethora of minor parties and independents followed a how-to-vote.
  • In the two-party preferred count, a relatively low 53.1% of preferences favoured Labor over Liberal.
  • The full two-candidate preferred count finished as Labor 23,517 (60.9%) to Liberal 15,086 (39.1%), 0.4% higher for Labor than the VEC’s count derived from polling place results.

More detail with tables inside the post.Read More »VIC22 – Werribee – Analysis of Preferences

Electoral Pendulum for the 2023 NSW Election

With just 11 weeks to go until the NSW election on 25 March, I thought it was time to publish the Electoral Pendulum that I am advising the ABC to use for the election.

Since the 2019 election there has been a redistribution of electoral boundaries, five by-elections and a number of members who have left the party for which they were elected.

The pendulum inside this posts tries to account for the numerous changes. I’ve also summarised the 2020/21 redistribution, and provided notes on alternate margins for seats.

Depending on nominations, there may be one or two seats where I change the margin between now and March.

The Coalition won the 2019 election with 48 seats to Labor 36 with nine members on the crossbench, three Greens, three Independents and three Shooters, Fishers and Farmers. (Three SFF in total, not one of each.)

As I explain in the post, accounting for redistributions, by-elections and defections, the starting point for the 2023 election is Coalition 46 seats, Labor 38 and nine on the crossbench, three Greens and six Independents. Majority government requires 47 seats.

With the increasing number of NSW seats that are no longer major party contests, talking about the uniform swing each side needs to win is becoming less meaningful. The Coalition holds five seats on margins under 4% and Labor six. The nine crossbenchers will be trying to retain their seats, and there will be new Independents nominating, especially in safe Liberal seats.

Labor needs nine seats on a daunting swing of 6.2% for majority government, but can hope to form government with support from the crossbench if it can gain five seats to finish with more seats than the Coalition, though the presence of at least three Greens may open the possibility that Labor could form government with fewer seats than the Coalition.

Premier Perrottet and his predecessor Glady Berejiklian have managed to govern for more than two years without a clear majority in the Legislative Assembly. It is one of the rare occasions in recent years where a Coalition government has successfully managed a hung parliament. The size of the crossbench, and the chance it will increase in size on 25 March, mean the result of the election may only be the starting point for the formation of the next NSW government. Read More »Electoral Pendulum for the 2023 NSW Election

VIC22 – Point Cook – Analysis of Preferences

Point Cook was one of the seven districts where the Victorian Electoral Commission (VEC) conducted its re-check of election night votes by data entering ballot paper preferences into a computer system. Point Cook had 15 candidates, and as explained in my Melton post, this presented difficulties for the VEC in conducting manual re-checks so data entry was used.

Point Cook, previously known as Altona, was a closely watched seat in 2022 despite its very safe 12.3% Labor margin. Sitting MP for Altona and former senior minister Jill Hennessy was retiring, and new Labor candidate Mathew Hilakari had only recently moved into the area. Point Cook is a rapidly growing seat of new housing estates classed as outer suburban and one of the few western Melbourne seats thought to be at risk for Labor.

Dr Joe Garra was one of the 15 candidates. Garra is a local GP who had been campaigning for years on inadequate hospital facilities on Melbourne’s south-west. He had finished second to Labor contesting Werribee in 2018. He chose to contest Point Cook after his home suburb, Werribee South, was transferred from Werribee to Point Cook by the redistribution. He had a disappointing result, finishing third with only 6.9% of the first preference vote. He was passed by the Greens during the distribution of preferences.

With 15 candidates, Werribee saw a very high informal vote at 10.2%. Only 33.0% of votes were cast on polling day, 51.5% were Pre-polls, 10.2% Postal votes and 4.5% Absent votes.

The two major parties attracted only 64.7% of the vote between them, Labor 40.0%% and the Liberal Party 24.7%. But the other 35.3% was split across 13 candidates with no other candidate passing 7%.

As with my previous posts on Northcote, Preston, Hawthorn, Brighton and Melton, this post will use of the electronic ballot papers to analyse preference flow statistics and also to look at the influence of candidate how-to-votes.

The key findings for Point Cook are –

  • The Liberal Party how-to-vote, which switched in 2022 to recommend preference for the Greens ahead of Labor, resulted in 70.2% of Liberal voters following the recommendation and putting the Greens ahead of Labor. Note that Liberal preferences were not distributed.
  • For the 12 Point Cook candidates that registered how-to-votes indicating preferences, 22.9% of ballot papers exactly matched the how-to-vote of the chosen first preference party. Labor at 30.5% and Liberal 27.2% had the highest rate of ballot paper concordance with how-to-vote recommendations. Few voters for the plethora of minor parties and independents followed a how-to-vote.
  • In the two-party preferred count, a relatively low 51.9% of preferences favoured Labor over Liberal.

More detail with tables inside the post.Read More »VIC22 – Point Cook – Analysis of Preferences

VIC22 – Melton – Analysis of Preferences

Melton was one of the seven districts where the Victorian Electoral Commission (VEC) conducted its re-check of election night votes by data entering ballot paper preferences into a computer system.

Melton had 14 candidates, and electorates with more than 12 candidates presented a counting problem for the VEC. After election night, all votes are manually check counted in bundles of 50 ballot papers, each bundle made up of first preferences for one candidate. After the counting of each bundle, bank note counters are used to quickly double check the number of counted papers in each bundle. The note counters verify the hand tally, NOT what’s written on the ballot papers.

Ballot papers with more than 12 candidates can’t fit into note counters. Hence why Melton (14 candidates), Point Cook (15) and Werribee (15) were data entered. The inability to use note counters would have substantially slowed down the amalgamation count where ballot papers are broken out of their polling place bundles and amalgamated ahead of the distribution of preferences. The distribution of preferences itself would also have been slowed.

Melton was a closely watched seat in 2022. It was the only seat in the state that had swung against Labor at the two previous state elections, and there had also been unusual swings against Labor in the area at May’s Federal election. Melton also attracted two noteworthy Independents in Dr Ian Birchall, who polled well in 2018, and Jarrod Bingham. When rolled in, it turned out neither had polled significantly, Birchall finishing third with 9.0% and Bingham fourth on 5.8%.

The two major parties attracted only 62% of the vote between them, Labor 37.7% and the Liberal Party 24.3%. But the other 38% was split across 12 candidates. In a seat with a low Green vote and with Independents campaigning against the government, Melton was a rare seat where preferences favoured the Liberal Party.

But the main threat to Labor in Melton was an Independent finishing second and sweeping up Liberal preferences, which in the end did not happen. That Bingham recommended preferences for the Liberal candidate ahead of Birchall further reduced the chances of the minor party vote coalescing. Too many candidates resulted in the anti-Labor vote dissipating in the distribution of preferences, as had happened in 2018.

As with my previous posts on Northcote, Preston, Hawthorn and Brighton, this post will use of the electronic ballot papers to analyse preference flow statistics and also to look at the influence of candidate how-to-votes.

The key findings for Melton are –

  • The Liberal Party how-to-vote, which switched in 2022 to recommend preference for the Greens ahead of Labor, resulted in 73.9% of Liberal voters following the recommendation and putting the Greens ahead of Labor. Note that Liberal preferences were not distributed.
  • For the 12 Melton candidates that registered how-to-votes indicating preferences, 27.0% of ballot papers exactly matched the how-to-vote of the chosen first preference party. The Liberals at 36.7% and Labor at 33.4% had the highest rate of ballot paper concordance with how-to-vote recommendations. The plethora of minor parties showed very little how-to-vote discipline.
  • In the two-party preferred count, a rare minority 44.5% of preferences favoured Labor over Liberal. The Greens polled only 4.6% of the vote but still delivered 80.3% of preferences to Labor despite fewer than one in ten Green ballot papers having the recommended how-to-vote sequence.

More detail with tables inside the post.Read More »VIC22 – Melton – Analysis of Preferences

VIC22 – Brighton – Analysis of Preferences

Brighton was another of seven districts where the Victorian Electoral Commission (VEC) chose to conduct its re-check of election night votes by data entering ballot paper preferences into a computer system.

Brighton has never been won by Labor, but new Liberal candidate James Newbury came close to defeat in 2018. Newbury withstood a Liberal pre-selection challenge by former Bayside City Council Mayor Felicity Frederico ahead of the 2022 election. Frederico contested Brighton as an Independent but she finished fourth with 9.1% of the vote.

The VEC conducted a Liberal versus Frederico indicative preference count on election night but switched to Liberal versus Labor on the Monday after. The VEC also decided to data enter ballot papers for the check count.

The VEC reports Newbury as having won Brighton with 54.2% after preferences based on an indicative preference count. Despite using electronic ballot papers, the VEC stopped the distribution of preferences when Newbury passed 50% with three candidates remaining in the count.

Analysing the electronic ballot papers for this post has allowed me to calculate the correct two-candidate preferred results. The final Brighton result is Newbury 22710 votes (55.1%), Labor 18,486 (44.9%).

As with my previous posts on Northcote, Preston ahd Hawthorn, this post will use of the electronic ballot papers to analyse preference flow statistics and also to look at the influence of candidate how-to-votes.

The key findings for Brighton are –

  • The Liberal Party how-to-vote, which switched in 2022 to recommend preference for the Greens ahead of Labor, resulted in a very high 85.7% of Liberal voters following the recommendation and putting the Greens ahead of Labor. At last May’s Federal election by comparison, when the Liberal Party still listed Labor ahead of the Greens on how-to-votes, Liberal preference flows to the Greens were a much lower 31.7% in Cooper, 29.8% in Melbourne and 26.7% in Wills.
  • For the seven Brighton candidates that registered how-to-votes indicating preferences, including three versions for Frederico, 39.1% of ballot papers exactly matched the how-to-vote of the chosen first preference party. A high 57.0% of Liberal voters completed the same sequence of of preferences as listed on the Liberal how-to-vote. This helps explain the strong preference flows to the Greens. A lower 31.8% of Labor voters exactly followed the party’s how-to-vote sequence.
  • In the two-party preferred count, 68.0% of preferences favoured Labor over Liberal, including 84.6% from the Greens, 60.7% from Frederico, 59.5% Animal Justice and 55.4% from Independent Gibson.
  • As I have noted in my previous posts, it is clear that the Liberal Party’s decision on how-to-vote recommendations has a major impact on whether Labor or the Greens receive the majority of Liberal preferences.

More detail with tables inside the post.Read More »VIC22 – Brighton – Analysis of Preferences

VIC22 Election – Hawthorn – Analysis of Preferences

Hawthorn was one of seven districts where the Victorian Electoral Commission (VEC) chose to conduct its re-check of election night votes by data entering ballot paper preferences into a computer system.

The seat was a three-way contest involving Labor, Liberal and Independent candidates. It was a very different electorate and contest to the Labor-Green battles in Northcote and Preston, the subject of my two previous posts in this series.

At the 2018 election the Liberal Party lost Hawthorn to Labor in a major upset. It was a very public defeat for Liberal MP John Pesutto, the result unfolding while he appeared on the ABC’s election night panel. The seat was won by retiree and little known Labor candidate John Kennedy.

Pesutto returned as the Liberal candidate at the 2022 election in a three-way contest against Kennedy and ‘teal’ Independent Melissa Lowe. On first preferences Pesutto polled 42.3%, Kennedy 22.1%, Lowe 20.0%, the Greens 11.1% with another 4.5% divided between four candidates.

The question to be resolved by the distribution of preferences was whether Lowe would pass Kennedy on Green preferences to finish in the final pairing. Kennedy’s initial lead of 948 votes narrowed after preferences but he still led Lowe by 106 votes at the crucial point when only three candidates remained in the count.

After preferences, Pesutto defeated Kennedy by 1,544 votes with 51.7% of the two-party preferred vote. Using the VEC’s ballot paper, it can be calculated that Pesutto would have defeated Lower by 1,296 votes with 51.5% of the two-candidate preferred vote. Pesutto’s victory was followed by his election as the new leader of the Liberal Party.

As with my previous posts on Northcote and Preston, this post on Hawthorn will use of the electronic ballot papers to analyse preference flow statistics and also to look at the influence of candidate how-to-votes.

The key findings for Hawthorn are –

  • The Liberal Party how-to-vote, which switched in 2022 to recommend preference for the Greens ahead of Labor, resulted in a very high 79.5% of Liberal voters following the recommendation and putting the Greens ahead of Labor. At last May’s Federal election by comparison, when the Liberal Party still listed Labor ahead of the Greens on how-to-votes, Liberal preference flows to the Greens were a much lower 31.7% in Cooper, 29.8% in Melbourne and 26.7% in Wills.
  • For the seven Hawthorn candidates that registered how-to-votes indicating preferences, 39.3% of ballot papers exactly matched the how-to-vote of the chosen first preference party. A high 53.9% of Liberal voters completed the same sequence of of preferences as listed on the Liberal how-to-vote. This helps explain the strong preference flows to the Greens. A lower 29.6% of Labor voters exactly followed the party’s how-to-vote sequence.
  • In the two-party preferred count, 73.4% of preferences favoured Labor over Liberal, including 85.9% for the Greens and 74.1% for Independent Melissa Lowe despite Lowe not recommending any preferences on her how-to-vote. By comparison, the flows to Labor in the local Federal seat of Kooyong last May were 77.2% overall, 83.5% for the Greens and 80.5% for Independent Monique Ryan.
  • Calculating preferences for the Liberal versus Melissa Lowe contest, the overall preference flows were 75.6% to Lowe including 79.2% from Labor and 82.4% for the Greens.
  • As I noted for both Northcote and Preston, it is clear that the Liberal Party’s decision on how-to-vote recommendations has a major impact on whether Labor or the Greens receive the majority of Liberal preferences.

More detail with tables inside the post.Read More »VIC22 Election – Hawthorn – Analysis of Preferences

VIC22 Election – Preston – Analysis of Preferences

Preston was one of the seven districts where the Victorian Electoral Commission (VEC) conducted its re-check count of ballot papers by data entering ballot paper preferences into a computer system.

In the case of Preston, data entry would also simplify what had looked set to be a complex preference distribution. Labor led on first preferences with 37.6%, with the Liberals second 17.3%, Greens third on 15.1% and Independent Gaetano Greco fourth on 13.7%. Preferences were always going to push the Greens into second place. For a time it looked like Greco would pass the Greens to reach third place, but his first preference vote declined significantly with the counting of absent and out of district pre-polls.

As with my previous post on Northcote, I will make use of the electronic ballot papers to analyse preference flow statistics and also to look at the influence of candidate how-to-votes.

The key findings for Preston are –

  • The Liberal Party switch to recommending preferences for the Greens ahead of Labor flipped the flow of Liberal preferences. From a first preference vote of 17.3%, Liberal preference flowed 72.3% to the Greens. At the Federal election in May, when Labor was the recommended preference, only 31.7% of Liberal preferences flowed to the Greens. Clearly the Liberal Party’s choice of preference recommendation has a significant influence on how its supporters direct preference.
  • In the Labor versus Greens two-candidate preferred count, 69.4% of preferences favoured the Greens over Labor, only voters for Family First favouring Labor. Labor won with 52.1% of the vote after preferences.
  • In the two-party preferred count, 71.0% of preferences favoured Labor over Liberal with massive difference in flows based on the ideological position of each excluded party. After preferences Labor won 69.7% of the two-party preferred vote.
  • From the ballot papers it is possible to calculate an alternative Labor versus Independent Greco preference count. Had Greco reached second, the final count after preferences would have been Labor 57.0% to Greco 43.0%. Greco received only 60.1% of preferences, doing very well with flows from every party except the Greens whose preferences flowed 61.8% to Labor. This was despite Greco being listed above Labor on the Greens’ how-to-vote recommendation.
  • Overall 24.7% of Preston ballot papers exactly matched the how-to-vote of the chosen first preference party. By party the highest rates of how-to-vote concordance were Labor 33.4%, Liberal 29.4%, Freedom Party 24.0%, Victorian Socialists 19.5%, Greco 18.2% and the Greens 15.8%. All other candidates saw rates of how-to-vote concordance under 10%, suggesting voters either ignored the recommendation or, more likely, never saw the candidate’s how-to-vote.

More detail with tables inside the post.Read More »VIC22 Election – Preston – Analysis of Preferences