Last week I wrote an analysis piece about the different shape of ballot papers in different parts of New South Wales – the urban areas and some large rural councils have above-the-line voting, small rural councils have a single column of ungrouped candidates, and there are some councils in between with a transitional mode.
I wanted to understand a bit more about below-the-line voting, at least in the places where a “line” exists.
This seems particularly relevant this week, with a number of surviving Liberals forced to rely on below-the-line voting to win in places like Hornsby and Pittwater. Labor candidates in Camden and Northern Beaches are in the same boat.
I started by looking at who got elected in each council, and whether any of them were either running in an ungrouped column or a group that doesn’t qualify for above-the-line voting. Incidentally I did not find a single case of a candidate running in a group with above-the-line voting who won while a candidate ranked above them lost.
It turns out that most councils with above-the-line voting consist entirely of ATL-eligible candidates. In the entire Greater Sydney region, along with the Illawarra and lower Hunter regions, every single councillor had run for a group receiving ATL votes. This map shows these councils in blue. The small number of councils where some councillors ran above the line and some ran below the line are marked in red. Those without above-the-line voting are marked in green. There wasn’t a single council where above-the-line voting was available but no-one was elected from it.
So what makes those red councils stand out from the blue councils? From looking at the map, it obviously has a relationship with geography – they tend to be councils further away from Sydney. But what element of the electoral system actually causes this outcome?
If you toggle the map onto the other layer, it shows the proportion of votes cast below the line in councils with above the line voting as an option. Those councils with some councils elected below the line have much higher below the line voting rates. The nine transitional councils have below-the-line rates ranging from 42.9% to 75.3%, whereas the other 57 councils with above-the-line voting had a rate ranging from 7.3% to 31%. There is no overlap between the two. And within that group of transitional councils, there is a strong relationship where a higher below-the-line vote produces more councillors elected below the line.
It makes sense that voters casting votes below the line elect more councillors who rely on those votes, but why? The next chart makes it clear – the particular choices of candidates in who runs influences the vote.
If you look at what proportion of the candidates were running ungrouped or for a non-ATL group, that proportion is under 20% for all of the councils with no below-the-line councillors (marked in blue), and the below the line rate only cracked 30% in one council.
For those with higher below the line rates, the proportion of candidates running below the line creeps up, with Snowy Monaro, Kempsey, Griffith and Narrabri having a ballot with more than half of the candidates unable to access above the line votes.
As more candidates run above the line, rates of voting below the line fall, and rates of candidates getting elected without an above-the-line box collapse to almost zero. There is little evidence of successful below-the-line campaigns in places where groups predominate.
Unfortunately the second layer of the map doesn’t give a particularly informative sense of what is going on within the urban parts of the state, but there is some variation in rates of below the line voting.
This next chart just shows urban councils (rural councils show a different trend), comparing the average number of councillors per ward and the proportion of votes cast below the line.
Below-the-line and above-the-line voting are not created equal. An ATL vote only requires one preference, a BTL vote requires multiple preferences to be formal, and generally requires more numbering to have the same effective impact. That assymetry gets worse as the magnitude rises. Indeed for 2-member wards, either system only requires a single ‘1’. But in Campbelltown, below-the-line voting requires eight preferences, and above-the-line still just requires one.
And this understandably is reflected in the above chart – below the line rates drop as the magnitude goes up.
Finally, I think I can make some conclusions about implications for 2024.
Firstly, it is going to be hard for someone to get elected from below the line, but possibly achievable in a 3-member ward for a party with a strong base of support. I can’t see it being possible in 15-member Campbelltown.
Below the line voting remains a mostly vestigial element of the electoral system in big urban councils – when the ballot structure switches to include above-the-line voting, they predominate in most cases.