Preferential voting for wonks: did the Greens really win Maiwar?

I am a big fan of compulsory preferential voting. I note the informal vote in the Queensland election despite the change from optional preferential was surprisingly low.

The complexity of the current Queensland election outcome reminds me that relevant stuff gets covered by Jordan Ellenberg in How Not To Be Wrong: The Hidden Math of Everyday Life.

Chapter 17: There is no such thing as public opinion. Recommended reading.

What Ellenberg does here is work the numbers differently in a three cornered contest to show that in a close race the eliminated candidate who came third could actually be the preferred candidate over the winner.

Where this could apply in the current Queensland election is the likely Greens win in Maiwar in inner Brisbane. With just a handful of votes splitting them the ALP candidate will likely be eliminated before the Green. The Greens candidate will then surpass the LNP which was a clear winner on first preferences with more than 40% which would have seem them elected easily on the first past the post system in the USA and UK..

However the LNP preferences don’t count at all in this situation. If they did it would be almost certain that the majority of LNP preferences would have gone to ALP over Green as was directed on how to votes. Consequently it is almost certain that the Green is not the most preferred candidate in Maiwar and possibly by quite a long way.

To put this another way. The great advantage of preferential voting is that the votes of minor candidates are not wasted and the result more accurately reflects community opinion. However in this situation it is the preferences of the leading candidate which are actually wasted.

Ellenberg suggests that this situation will tend to discriminate against the more centrist candidate.


Quick election review for FNQ

Well that’s over and a damned good thing too. FNQ was a strong result for the ALP with wins in Mulgrave, Cairns, Barron River and Cook. The election of Cynthia Lui in Cook as the first elected TI representative is welcome and also may challenge the paradigm that a woman can’t win Leichhardt.

When Antony Green says this is the most complex situation he has dealt with it wise to be careful. The One Nation (ONP) vote in the Cairns urban electorates appears to be lower than comparable regions elsewhere such as Townsville and Mackay.

The Cairns electorate is confused by Pyne who has more in common with ONP economic and populist policy and shared misconceptions than supporters of either would likely be prepared to acknowledge. (Note: Rockhampton is also difficult to assess on this with an ex-ALP independent but don’t know the circumstances).

Among the Cairns urban area electorates ONP scored highest in Mulgrave. This was also among those won by ONP back in 1998. Curtis Pitt way too big on primary vote this time to be any contest on preferences. The boundaries have shifted since with Mulgrave moving more into the suburbs and the recent redistribution since the last election.

Notably also Brinsmead has shifted from Barron River into Cairns with the recent redistribution. Consequently extreme caution should be taken with reported swing numbers at the Cairns Post. We used to rely on Antony Green for this stuff. Sadly the new ABS census demographic data by state electorate doesn’t reflect recent boundary changes.

Anyhows. let’s look at Barron River as the cleanest and most marginal of the FNQ complexities and a comparison of One Nation with Greens:

ON v Green

Pre-poll was the largest single booth component by number of voters. LNP outperform at pre-poll but are even outperformed slightly here by ONP relative to electorate result. ALP and particularly Greens underperform at pre-poll. Both ONP and Greens outperformed state averages in Barron River. ONP didn’t contest all Queensland electorates so Barron River would likely be around or an underperformance of those contested. The ONP candidate previously stood in Barron River for Palmer United Party and scored 8.8%.

It can only be assumed that the Greens voters of Machans Beach were too stoned to make it down the road to pre-poll?


Update: This would appear to leave ON vote in the Cairns based electorates below state and regional averages.

@AntonyGreenABC  #qldvotes One Nation polled 13.7%, 9.7 in south-east and 21.4 in rest of state. Looking only at seats One Nation contested, the party polled 20.7%, 18.4% in south-east 23.1% in rest of state.


Leveraging Crocodile Tourism

The Member for Kennedy has colourfully raised concerns about crocodile activity which has spread around the world. Yet again his numeracy appears to be somewhat exaggerated. However this did remind me of a recent 2017 Ignobel Award for: Never Smile at a Crocodile: Betting on Electronic Gaming Machines is Intensified by Reptile-Induced Arousal

Tourists at the Koorana Saltwater Crocodile Farm in Coowonga, Queensland, Australia, including 62 males and 41 females, aged 18–66 (M = 34.2, SD = 13.3), were randomly assigned to play a laptop-simulated Electronic Gaming Machine (EGM) either: (1) prior to entry, or (2) after having held a 1-m saltwater-crocodile. Gambling behavior; including bet-size, speed of betting, final payouts and trials played on the EGM; was investigated with respect to participants’ assigned arousal condition, problem-gambling status, and affective state. At-risk gamblers with few self-reported negative emotions placed higher average bets at the EGM after having held the crocodile when compared to the control. In contrast, at-risk gamblers with many self-reported negative emotions placed lower average bets at the EGM after having held the crocodile. The results suggest that high arousal can intensify gambling in at-risk players, but only if this feeling state is not perceived as a negative emotion.

At least one local operator appears to see some opportunity in this: Have some pots and swim with crocs at iconic Cairns pub. I’m not sure how this salubrious establishment renowned for its discerning clientele earned an “iconic” tag but it could provide an opportunity to test the research in a real world situation.

Regional Housing Analysis

A new paper from the Centre for Social Research & Methods at ANU may challenge some common perceptions: Regional Housing Supply and Demand in Australia

This is really worth reading and so far I have only skimmed it. The appendix table listing the bottom 10 shortage regions is interesting though. Note that these are negative shortages for the bottom 10 so represent surplus based on the methodology.

Housing Shortage SA3

SA3 level splits Cairns in two and Cairns – South also represents the city and inner suburbs and comprises about two thirds the population of the cairns urban area. Cairns – South looks a bit odd there given all the hype we have had on rental vacancy rates and the property market.

Will have to come back for a closer look at this.

Air Freight Drain

Cairns Airport numbers for October were so boring they don’t even deserve a graph. What does deserve a graph is the decline in international freight through the airport.

Cairns Freight

These are BITRE numbers to FY17. Lowest since 1990-91. Guessing this is freight on scheduled flights. There were numbers somewhere at BITRE which I had previously downloaded by destination and indicated Cairns international freight to be overwhelmingly Hong Kong.

Probably worth a further look given media commentary on air freight and agricultural export potential.

Cairns retains Pokie King Crown

Cairns has staved off a challenge from Townsville to hold on to the Pokie King Challenge Mug for October. Townsvillians held back as they saved their pennies to be able to afford tickets to next years Neil Diamond concert.

Pokie kings

This is actually an interesting data set particularly related to recent hours expansions where it seems like all your local suburban pubs are seeking gaming licence expansions for pokies until 4am. I find it difficult to contemplate that the economics of pokies would make 4am gaming in a suburban pub work. But who am I to judge having never ventured from bed at that hour to investigate the local pub.

Will leave this to next month by when Reef Casino will have provided guidance.


What is the rental vacancy rate?

SQM Research provide an excellent range of free real estate data which I frequently use and reference. Michael Matusik has had a spray at their rental vacancy rates data. This is how SQM determine what is a vacant rental from online sources:

The Rental Vacancies component is based on all monitored and unique online listings for the period of a calendar month. The series starts off in January 2005. All listings are taken from online monitoring of major listings sites. Only those properties with unique addresses or a unique listing id are used. Those advertisements with no addresses are excluded from the series. Any addresses repeated between sites are de-duped. Only those listings that have been advertised for three weeks or more (and are still currently advertised as at the time of collation) are used.

I don’t have a problem with that which is intended to identify what is genuinely vacant and not just part of the routine churn of tenants. What is more important is that a methodology is consistent over the historical series. Where I tend to agree with Matusik is what the SQM vacancy rates implies for the rental stock because I have previously taken this up with SQM also.

I never did get a satisfactory explanation and it still doesn’t reconcile when updated for 2016 Census data. This is SQM for postcode Cairns 4870:

SQM1Source: SQM Research

The 208 vacant dwellings and 1.4% vacancy rate implies a rental stock of 14,857 dwellings. This compares with 11,156 rentals in 4870 on census night. You would have to throw in all the census responses with no tenure stated as well as every unoccupied dwelling on census night to get to the implied stock. That pattern is replicated across other Cairns postcodes.

If I aggregate the SQM data for Cairns postcodes (4868, 4869, 4870, 4878, 4879) I get a vacancy rate of 1.6% in October. SQM also provide data on a regional basis although this is not labelled for number of vacancies and doesn’t correlate well with the aggregation for Cairns. Perhaps I have misunderstood something?

The SQM vacancy rate for the Cairns region at 1.9% does though correlate with the most recent HTW CairnsWatch which is based on a rent roll survey at 2%. It isn’t clear what the definition of a vacancy is though.


Source: HTW CairnsWatch

More important than trying to achieve some kind definitional and measurement ideal is a consistent series over time which is understood.  While I have some doubts about this aspect of their data this doesn’t discredit SQM as a useful source. The query I would have on that HTW series is whether the ‘balanced market’ range displayed remains contemporary or requires review. It looks high compared to what now seems to be ‘normal’ in most cities.

SSM: young males lost in the post

Well the same sex marriage results are in with the predicted resounding win for ‘yes’. I’m sure there will be more than enough media commentary for any sane person so will just focus on the response rate where Leichhardt came in last in Queensland.


The bottom four response rates are all the NQ electorates. There was nothing unusual about the result in Leichhardt which was about average at 63.4% yes versus 60.7% for Queensland. Kennedy was one of only a few to vote against with just 46.7% in favour.

The lower response rates came in the younger demographics and particularly males:


All that now needs to be resolved apparently is how to deal with the rights of homophobic bakers and florists *insert irony font here*.