Wow

Last night I was asked if I was surprised by the result. The short answer is yes! Not because Labor won (that was my working assumption) but that it won with such a thumping majority. The national two-party preferred vote is currently sitting at 55.37 per cent for Labor, with some 70 per cent of the vote counted. While I expect that total will come back a touch as the remaining vote is counted (postal votes often favour the Coalition), it will remain an historic win for Labor. Right now the ABC is projecting the following seat count in the new 150 seat Parliament:

• Labor - 85 seats (they won 77 seats in 2022, in a 151 seat Parliament)
• Coalition - 36 seats (previously 58)
• Greens - 0 seats (previously 4)
• Others - 10 seats (previously 12)
• Undecided - 19 seats

Polling fail?

Was it a polling fail? My initial thought is maybe. Again, its not that the polls got the winner wrong, but that they did not capture the scale of Labor's win. The average poll estimate since 1 April had Labor on 52.5 per cent of the two-party preferred vote (better than my Bayesian aggregation of 52.1 percent). The interim result is 2.9 percentage points off that mark. On the same metric, the 2019 polls missed by 2.9 percentage points. So far, only one poll since 1 April was better than the interim result for Labor. Nonetheless, because the postal and absentee votes can change things, we will need to wait until all the votes are counted before we come to the question of whether there was a polling fail. 

To be fair, I did warn over the past week that the under-dispersion in the polling data meant that the risk of a polling error was heightened. When I was asked on X/Twitter whether it is possible for the polls to “herd” in a way that underestimates support for the ALP, in contrast to 2019, when polls herded to overestimate it; I simply responded "yes".

Election day - betting markets

There are two posts this morning. This is the betting market post, and separately there is a polls update.

National Winner

At 6.30am this morning, the betting market at Sportsbet was firmly of the view that the Labor Party will provide the Prime Minister in the new Parliament. The odds for a Coalition win were \$9.30 (down from \$11.00 yesterday). The odds for a Labor win were \$1.08 (up from \$1.06 yesterday).  The probability of a Labor win was 89.6 per cent - down from 91.2 per cent yesterday.

Election day - polls

There are two posts this morning. This is the polls post, and separately there is a betting market update.

Two-party preferred

There is an abundance of new polls today, each with the following two-party preferred (2pp) population estimate:

• YouGov - 52.2 per cent to Labor
• Newspoll - 52.5 per cent to Labor
• Freshwater Strategy - 51.5 per cent to Labor, and
• Roy Morgan - 53 per cent to Labor.

Which results in my national aggregation of 52.1 per cent for Labor, which means that if the polls in aggregate are reasonably correct I expect the final result to be somewhere in the range 51 to 53 per cent for Labor, which would see either a Labor minority or majority government. My central estimate is that the new Parliament will be much like the previous Parliament. 

Day 36 - 2025 Federal Election

The election is tomorrow! Today we have two more polls:  Freshwater strategy 51.5/48.5 to Labor, and DemosAU 51/49 to Labor. The aggregation remains unchanged.

Day 35 - 2025 Federal Electon

It looks like we are seeing some poll-herding in the lead-up to 2025 Election. Since yesterday's blog post, two more polls were released in the 52 to 53 per cent range for Labor. It is clear that most polls since the beginning of April are narrowly dispersed in the 52 to 53 per cent range on the two-party preferred metric for Labor. 

One outlier from the narrow 52/53 trend was a Freshwater poll on 15 April which had the race at 50/50. Freshwater Strategy polls typically have the Labor vote lower than the other polls. Two outliers come from Roy Morgan: 54.5 on 10 April and 55.5 on 17 April. In recent months, Roy Morgan has typically had Labor above the trend of the other polls. A further outlier comes from Resolve Strategic with Labor on 53.5 on 11 April. And a final outlier comes from YouGov with 53.5 on 19 April. 

This narrow distribution is not what you would expect mathematically. Let's imagine a world where every poll is based on a random sample of (say) 1200 people, and the underlying voting intention was (for the sake of the argument) 52.8 per cent (the mean poll result since 1 April). We would expect to see a distribution of poll results that was normally distributed around that mean. 

We can use a kernel density estimate to see what we actually have, and that is a distribution that is not normal, We actually have a leptokurtic distribution. It is too peaked around the mean, with thin tails. We can compare both curves, as both of them have an area of 1 square unit under the curve.