Tuesday, March 6, 2018

Tasmania 2018: But What Does It All Mean?

Undecided: Franklin - Liberal vs Green - tossup
Undecided: Bass - Labor vs Green vs Liberal - Labor slightly favoured, Liberal chance remote

After the last Tasmanian election, I saw no need to unpack possible meanings of the result, as I thought it was all obvious to anyone who had followed the state's politics through that time.  This one, however, is different, though I certainly won't claim to have all the answers.  A government that seemed to be sleepwalking to a loss of majority has rebounded to the point of suffering just a trivial swing against it.  The Labor opposition did rebound to a degree, but mostly at the cost of the Greens.

Some facts and stats

A few facts about the election first.  For the first time since the 1970s, and the first time for a conservative party since 1912-3, the Liberal Party has topped 50% of the primary vote for a second election in a row.  The charge was led by Premier Hodgman, whose 38.3% is the highest candidate vote since Robin Gray in 1986.  (Hodgman will also break Doug Lowe's 1979 record for the largest number of votes recorded by a candidate, though this record is somewhat meaningless because of population growth. Lowe's record for the highest candidate percentage, 51.2%, may very well never be beaten - and I don't think the fact it was achieved before Robson Rotation really makes much of a difference.)

Contra to Labor leader Rebecca White's claim of having pushed the Government to the edge of defeat (by which she means a hung parliament anyway) this was simply not a close result.  In what turned out unexpectedly to be the most critical seat (Denison), about a 6% swing from Liberal to Labor on the current figures would be needed for the seat to have been lost, making this the Hare-Clark equivalent of a 56-44 pasting at federal level.  Even if the Liberals are reduced to 13 seats, they were still not really close to losing on the night - nor, I suspect, at any time in the final month at least.

The election saw the highest combined major party vote since 1986, and the highest in the era in which the Greens have been a statewide force.  The Greens are currently on 10.03%, though there is a fair chance that absent and other remaining votes will lift them just above their worst-ever score of 10.18% (1998).

In the seats where the Jacqui Lambie Network ran, the swings between Labor and the Greens are very similar to each other: Lyons Labor +5.5 (Greens -5.1), Bass +3.2 (3.6), Braddon +4.1 (-3.7).  Labor gained much more than the Greens lost in only Denison (+8.6, -4.0) and Franklin (+5.9, -2.5).  These are both seats in which the Network did not contest.  Thus the overall results (including a swing of 1.7% to the Liberals in Bass) suggest that net vote transfer between the majors was virtually zero, and that what variation exists in the Liberal and Labor swings by seat was largely determined by differences in the array and performance of Others.

At candidate level, the Liberals have elected Sue Hickey as a new MP replacing their sole retirement, Matthew Groom.  They will probably lose Joan Rylah (if she holds they will instead lose Roger Jaensch) and may lose Nic Street.  Labor will gain Anita Dow, Alison Standen, a new Lyons MP to be determined, and could well gain Jennifer Houston in Bass.  They have also regained David O'Byrne. The only Labor MP in trouble is Madeleine Ogilvie who is presently trailing Ella Haddad by a significant amount.  For the Greens, leader Cassy O'Connor is back, but both Rosalie Woodruff and Andrea Dawkins are uncertain, with Woodruff a 50-50 chance and Dawkins' position more difficult.

Fourth parties have again failed to get near winning a seat, though at times the Jacqui Lambie Network was on track for a seat or two in polling.  The highest vote by a candidate outside the big three was recorded by independent fisherman Craig Garland (Braddon) who may yet break 2000 votes in a campaign that he says cost him $800.

The History

This election was very similar to 2006.  In that election, the Lennon Labor government was widely being completely written off on the basis of bad polling, and was at odds of $9 to retain its majority, but beat the Liberals by 17.5 points having been ahead by as little as three points in an EMRS poll a few months out.  This time the odds were even sillier and many informed observers helped themselves to the $15 available on Liberal majority government.  The 2006 campaign - which was nastier than this one - also featured claims about the result being "bought" by business interests campaigning for majority government.

Media coverage of the election included a few hung-parliament nostalgia pieces dealing with minority governments past, in which key players from those days trumped up their legacies and sanitised the cross-party tensions that existed.  A recap of the 2006 election would have given voters a much better idea of what to expect.

The Mystery

The one thing that is hardest to explain about this election is why Labor painted a target on itself by supporting the complete removal of poker machines from pubs and clubs.  As a strategy, it was successful in removing votes from the Greens, but it failed dismally to connect with working-class voters especially in the north of the state.  Whether the policy was even capable of being sold to such voters is dubious but in any case I think Labor had no idea how to sell it.

Frequently this sort of thing happens when a party's stated reasons for holding a policy aren't their real reasons.  Whatever the reasons, Labor's desire to vastly restrict the pokies industry made the party a sitting duck when it came to the argument that Labor didn't care about working-class jobs, really had not changed from the party that governed with the Greens in the parliament before, and would probably do so again.   The policy seems to have also generated perceptions of class contempt (middle-class for lower-class) and (in my view valid) concerns that Labor and Green ideologies in the state don't take personal freedom seriously.  (The same charge can be levelled against the Liberals, but at least they seem to understand the concept, even when they selectively ignore it.)

It is possible Labor was looking at internal poll numbers that suggested the EMRS polls of late last year were wrong and that the Liberals had a large voting lead locked in, so that attacking the Greens may have made sense.  However unless Labor knew they were already losing, it is difficult to see why this high-risk gamble was taken. As well as the money thrown at the issue by pro-pokies forces, media focus on the issue created an impression that the election was all about banning pokies, preventing Labor from making health their central message (though I am not sure they would have owned that issue, or education, anyway.) I suspect there is a lot more to come to light about this in the future.

It is interesting to see ALP Right sources blaming the fact that none of their candidates have had straightforward wins on the pokies issue.  The charge is that the policy scared off conservative voters.  In fact candidate factors (in the case of Ogilvie), campaign factors (Butler) and being up against a higher-profile candidate who also had a union background (Midson) probably contributed to the Right's difficulties, but even if the pokies policy was the problem, why then were Ogilvie and Butler early movers on the issue?

This was not the only cause of Labor's defeat.  The legacy of the 2014 election left them with a shortage of incumbents and the campaign generally appeared lethargic and lackadaisical, especially when it came to refuting government claims in the media and advertising.  As a result the government was able to get away even with brazen lies like their one about Labor having always done deals with the Greens in the past.  Person-to-person communication as a campaign strategy is all very well, but the same goes for it as goes for billboards: the voter has to be receptive to what you are selling.

The result is also an object lesson in the value of securing your position while you have the power.  The previous Labor-Greens government did not make a serious effort at reforming donation laws.  Had they done so, they could have been much better placed to take on the pokies issue in a way that at least forced transparency from their opponents.

It is not enough to attribute the result solely to the Liberals' resourcing advantages.  The Liberals displayed impressive unity in office with internal tensions virtually never on public display.  The Liberal campaign was well co-ordinated and responsive to almost every issue that was raised.  There was the last-minute issue of the secretive approach to gun policies, but this had little impact, in part because the changes being proposed are mostly minor anyway.  (In fact, while the media mostly saw this as a massive story, voters may have seen it as a beat-up.)

The Polling

This was one of the more sparsely polled Tasmanian elections, with only a ReachTEL and an EMRS by way of public polls in the last few months, and the fieldwork for those was nine days and 5-7 days prior to the poll respectively.

On current statewide numbers (Liberal 50.5 Labor 32.8 Green 10.0 JLN 3.2 Others 3.6) the Mercury ReachTEL (Lib 48 ALP 32.2 Green 12.5 JLN 5.3 Other 2.1) had no errors greater than 2.5 points, which is a strong performance by the standards of polling in this state.  Also, the JLN vote becomes more accurate (4.1%) if adjusted for them not running in all seats.  The EMRS poll (46-34-12-4-3) had the Liberals 4.5 points too low.  It is possible that there were actually shifts in the Liberals' favour during the final week.

Both polls, and also all internal polls I have seen, overestimated the Greens vote, but only by a couple of points rather than the 4+ points sometimes seen in the past.  Unlike in recent previous elections, these polls got the Labor vote right, while it's possible that EMRS polling tends to underestimate incumbent governments rather than the Liberals specifically.

Polling was extensively and adroitly used by the Liberals during the lead-up to the election to convince the public that only they could win majority government.  This then helped generate the same bandwagon effect seen in poll tracking in 2005-6 (though in that case the evidence that the Lennon government was in trouble was actually much spottier than people remember it as being).  Although the company involved (MediaReach) was virtually unknown, Labor totally failed to counter this polling narrative - probably because they couldn't.  The youngest published MediaReach data (5 and 8 Feb with average results of 46.5-30-12-5-6.5) were close enough to suggest the MediaReach polls were broadly accurate at the time they were taken.  That said, there could have been a self-fulfilling prophecy component in this.

Questions might be raised about EMRS' tracking performance.  When the government appeared to be faltering in the EMRS polls late in 2017, was its standing actually that bad?

Hung Parliament Club

In Tasmania there are still too many commentators and politics junkies who are clearly in the tank for hung parliaments.  This mindset involves not only arguing that minority government is the ideal situation for Tasmania but also overestimating how likely it is to occur.  Hung parliament club no longer dominates the commentariat in the way that it did circa the mid-2000s, but it is still far too prevalent.  The fact is that while hung parliaments in the Tasmanian context may be more open compared to majority parliaments, and often deliver reforming outcomes, they are also unstable.  The instability arises from the Greens' desire to deliver for their supporters in cases where the majors are in lock step - the only way Greens can get outcomes in those cases is through the implied threat to bring down the government.  Meanwhile while major parties sometimes want to govern with Greens support, they also realise that the electorate dislikes it, and this can cause internal instability.

This campaign provided a fine example of the instability of Tasmanian hung parliaments when the Greens threatened to move a no confidence motion against the government because it had not declared its campaign funding sources.  Weeks earlier Cassy O'Connor had been saying that minority government was government for grownups and that the Greens wanted to work for stability.  There is a long history of the Greens claiming to support co-operative stability but only actually supporting it until the major parties trip one of their wires.  Logging quotas, the size of parliament, the performance of the Accord era education minister Peter Patmore, and now pokies donations - these are all examples of issues that have caused the Greens to lose interest in working together over the past few decades, and there will always be more.

I think the one-eyed barracking for hung parliaments by some academics and locked-in Greens supporters simply alienates other voters.  (That's not to say the barracking for majority government is any less one-eyed or any more credible.) At least, if trying to convince voters that hung parliaments work, be honest about their limitations and look for ways that those can be addressed.  Ideally though, I think those who like hung parliaments will most increase their chances of occurring if they just shut up about it, permanently.  The first rule of hung parliament club is that you do not talk about hung parliaments.

The Greens

The Greens have suffered a 3.8% swing against them, following a 7.7% swing against them in the 2014 election, and as a result have now lost more than half their record 2010 vote in two elections.  In this election the party lost more than half its already diminished 2014 vote in Braddon, and close to half of it in Lyons.  Fraser Brindley's weakness as a lead candidate was illustrated by him securing just 43.5% of his party's vote so far, the lowest for a Greens candidate ever.  (The previous record low was 53.3%, and Scott Jordan in Braddon is bobbing around that level.)  But stupid candidate preselections in two electorates are far from the only problem.

The Greens were always facing serious challenges at this election.  Having two out of three MPs elected mid-term is not ideal for them in terms of profile-building, but primarily the party does not do well at elections where there is policy daylight between the major parties.  Their best elections have been 1989 (where the Gray government was a beacon for environmental opposition), 2002 (where the Liberal Opposition was utterly hopeless) and 2010 (where both major parties were boring and there was a feeling of nothing at stake.)  At this election, some Greens voters switched back to Labor (in cases for the first time in decades) because they were impressed with Labor's position on poker machines.  Other Green voters would have switched to reward the party for taking a chance by choosing White as leader.  It shows once again that the Greens voters wander off at the slightest distraction and perhaps a feeling that voting Green is some kind of stern moral duty, not something fresh and exciting, has something to do with this.

The Greens are complaining about the pro-pokies campaign "buying" votes away from them, but this is nonsense - those who were still voting Green at least in 2014 are very unlikely to be susceptible to claims made by the gambling lobby.  Also, the evidence is that Greens voters switched to the other anti-pokies party (Labor), not the Liberals.

Another strange excuse came from several Greens, including Bob Brown for whom no Greens result is ever bad (unless it is recorded by the Rhiannon faction of course.)  This is that there were no major  environmental issues, which the authors of this excuse are delusionally blaming on the major parties for not releasing any destructive environmental policies. This is basically either an admission that over the past four years a Liberal majority government has managed Tasmania's environment very well or an admission that to the extent it hasn't, the Greens are unable to communicate this effectively to voters.  In a time where, for instance, the endangered and remarkable Maugean skate is considered by experts to be probably doomed because of mismanagement of Macquarie Harbour, it's a bit hard to argue it's entirely the former.

There's a widespread perception that the Greens' campaign was more or less invisible, but I think that it was also bad in other ways.  For instance, I saw one massive billboard declaring that the party "CAN'T BE BOUGHT. WILL NEVER SELL OUT", but plenty of previously diehard Greens saw their 2010 deal with Labor as exactly that.

It's clear that the demographic factors that propel the Greens vote ever higher in inner city Melbourne, Sydney and Brisbane don't operate to anything like the same extent in Tasmania.  Thus while environmental battles made this their birthplace, they don't have a natural vote floor here, and their vote could in theory sink even lower.  I expect that they'll be back, but if they are reduced to a single seat (that of their leader, who will be widely blamed for this result) then the path back will be more painful than it was in 1998.

The Elephant On The Red Couches

It's worth remembering that this government has had a lot of trouble getting legislation through the Legislative Council, where it had to run the gauntlet of an increasing number of Labor and/or indepedent left-wing MLCs. Currently there are four of each giving the left 8 of 15 seats, and moreover one of the non-left MLCs, Jim Wilkinson, is the President and doesn't normally vote.

This situation isn't very likely to improve for the government any time soon unless there is an unplanned resignation.  Conservative independent Greg Hall retires this year with a vacancy in the new seat of Prosser, so the best hope there is to break even, and one would expect Rob Valentine to be re-elected in Hobart (or if he doesn't recontest, probably someone of similar views.)  Wilkinson is expected to retire in 2019 which will spark a big fight for his seat. The government has already reached out to indicate it will try harder to work with the upper house, but I expect that its more controversial bills will continue to strike trouble.


  1. Labor ran a poor campaign on a lot of things, the pokies being at the top plus having a leader who has only been in the job for 12 months or so and a lack of incumbents. Also a thing that doesn't seem to be factored in is that in areas of Tasmania there is a visceral hatred of the Greens. Due to this the hang over from the 2010 deal between Labor and Greens still persists. The slightest whiff of another of those just turn people off Labor and until they expunge the electorates memories of that will continue to struggle.

  2. Thanks Kevin. I'm wondering whether there are any data available that shed light on conjecture around the dollar value of advertising and its voting outcomes? Thinking the Howard Family Bonuses, Big Pokie etc.?

    1. I'm not an expert in this area but I know studies in the US on things like corflute visibility have found it very difficult to prove a link between certain aspects of campaign spending and success. There is also a lot of cynicism about whether TV advertising is actually still effective - at all. I suspect the difficulty in proving how effective campaign spending is comes partly from the fact that a dud message can't be sold whatever you spend on it, but that doesn't stop people trying. Tasmania federal 2016 - Liberals carpeted the countryside with signs to no effect.

  3. Fantastic summary. As I read it, your advice for the Greens would be as follows:
    1. Focus on issues where Lib/Lab are in lock-step
    2. Be more fresh and exciting
    3. Be open about the pros *and cons* of hung parliaments (and what you'd do to try and improve the latter) and don't talk about them more than you have to

  4. I think you're missing a rationale for the anti-pokies policy there.. that they actually believed (because they broke from tradition and looked at evidence rather than a pile of cash delivered to their party room) that it was the right thing to do! Not adequately selling the evidence to the public is of course still a failing politically.

    Surely you're not suggesting that the only job of politicians is just to follow what's tracking well in the polls and never take a lead on making the state/country a better place to live?

    I'd love to see much stronger laws akin to the legislative council on donations and corporate spending - it might help focus discussion on policy differences and less on billboard slogans.

    1. It is right to do *something* - and something more than token - about pokie proliferation and its consequences. But if Labor's decision was based on a belief that this particular solution and not some different solution that showed even passing respect for personal freedom concerns was right, how did they come to that decision? I addressed this briefly in a previous article (http://kevinbonham.blogspot.com.au/2018/03/tasmania-2018-commissioned-pokies.html) - a decision may be "evidence-based" (in the term of this policy being the best way to achieve a stated desired outcome) but when that label is invoked in ignorance of a philosophical dimension, you get problems.

      There is a saying that politics is the art of the possible. Some things might be desirable but unachievable and it's harder to make the state a better place to live from Opposition.

    2. I agree, Rowan. I got the impression Labor genuinely believed that removing pokies from pubs and clubs would be in the State's best interest. There are some very good economic arguments in favour of doing so but they didn't get that message across well enough (it was always going to be a challenge).

      As for the personal freedoms argument kevin, I'm not that sold. There always has to be a line drawn somewhere between personal freedoms and the wellbeing of society; the argument is about where the line is drawn.

      Pokies are an interesting topic in that regard because I think there is a very good argument that they are designed to take away, or at least push to one side, freedom of choice, in the sense that they rely upon creating compulsive users in order to be profitable. Once someone is hooked, freedom doesn't come into it.

      There's a good reason these things are restricted to casinos in most other developed nations. Speaking of which, freedom to play pokies isn't truly taken away if they are available at casinos.

      They are also concentrated in lower socio-economic areas, the supposed rationale being that less educated people are more likely to get hooked. I don't know just how much more likely, but it's a truly awful concept.

      Possibly these issues could have been addressed by less extreme regulatory measures. Still, a ban would have been better for the state than no action at all.

    3. I think the "freedom to play pokies isn't truly taken away if they are available at casinos" line is the weakest line used by the anti-pokies campaigners and is nearly as bad as some of those used by pro-pokies campaigners. Try telling that, for instance, to a pensioner in Burnie who doesn't drive and whose nearest casino is in Launceston. Getting to and from the casino will probably cost her more than she will lose on the pokies. Even if one doesn't see this as compromising her freedom to play the pokies fullstop, it takes away her pre-existing freedom to play them in her area. Whether she should have that freedom is another question.

      Pokies do have addictive design components but nowhere near everyone who uses them becomes addicted. There are plenty of other designs that use lures to hook people in and keep them hooked - this even includes the design of social media platforms. People become addicted to smartphones to the extent that they use them while driving and crash. Ban or very heavily restrict social media? Ditto smartphones? Ditto alcohol? I don't think the fact that some people become addicted to something is generally reason to ban it or very heavily restrict it for everyone else. I think it's more a reason to look, at a first resort, at how addictive features can be regulated to minimise harm and consumer education and warnings increased. Also, the interplay between pokie use and alcohol needs to be addressed.

      There's no freedom once someone gets addicted? Sure, unless the addiction is fixed. But that's irrelevant when it comes to users who don't get addicted.

      Sure, less educated people are more likely to get hooked, but they're also more likely to genuinely enjoy the product (this also applies to much less addictive forms of luck-based gambling too).

      As I've mentioned before, I think pokies are junk; I can't relate to the desire to play the things at all. But if I dismiss some people's freedom to play the pokies lightly then I invite the same approach to whatever I enjoy. I'd rather defend someone's freedom to do something I can't relate to than give the impression - rampant across the whole of Australian politics - that only the freedoms one's own side values matter.

    4. I take your points, Kevin, but I think you're glossing over the fact that addiction is ancillary to those other things you've mentioned - drinking, smartphones, even horse-racing (at a stretch) can stand alone as a form of entertainment. Addiction is a risk inherent in doing (or betting on) those things.

      Pokies are there purely to get people hooked on playing them. They are addiction-by-design. There is literally no other purpose to them. Most people who use pokies don't get hooked, but an alarmingly large percentage (20%, depending on who you listen to) of people do. Even if that estimate is on the high side it is still an extraordinary strike rate.

      As I said, it's possible that these issues could have been addressed by tighter regulation and better assistance services. I just think a ban is better than no action at all.

  5. There are certainly lots of ways of reducing the damage from poker machines without banning them, and many of these ways respect the freedom issue for those not addicted.
    I find it interesting that Tasmania has the problems it does with minority/coalition government. Victoria, NSW and SA seem to manage OK when there are minority situations. Gillard managed her minority government pretty well in terms of getting things done, though the media were hysterical about it. With Hare-Clark you have to expect minority Government most of the time, and in the ACT it works OK having an agreement between the ALP and the Greens with regard to how the Government works. This time we in the ACT have a Green as a Minister. I suspect the fact that the ACT is almost entirely urban makes it easier for us.

    1. The problems in Tasmania tend to stem from there being environmental issues on which the major parties agree with each other but the Greens don't agree with either, and these issues being very important to the Greens support base. I suspect this problem doesn't arise in the ACT not only because the ACT doesn't have those issues but also because the ACT has a different range of ALP voters. Independents as supporters of minority governments tend to have desires that major parties can work with without burning off their support bases.

    2. The hours that pokies venues are open is of some relevance. The Elwick in Glenorchy has a poker machine lounge that stops serving alcohol but remains open well after that, possibly as late as 4AM and re-opens at 9AM. This targets people with an addiction. I really doubt people are playing for a bit of light amusement at 9:30 on a Tuesday morning. I think that needs looking at, at the very least.

    3. That should certainly be dealt with. There is no reason the things should be allowed to be open very late at night.

  6. Greens do tend to be as inflexible as a steel bar on environmental issues. Very ideological. There is a strange similarity between Bob Brown, Eric Abetz and Christine Milne.

  7. Hi Kevin

    Thanks for your insights. This is the first time I've followed a Tasmanian election and I was wondering whether there will be updates over the next few days or whether we have to wait until they push the big red button (presuming there is a button and it is, in fact, red).

    Also wondering whether the crash in the Greens vote has Federal Senate parallels? They've lost more than half their vote in two cycles and their haemorrhaging votes in the Senate as well. Its particularly bad in SA, where more than half their vote seems to have gone to Xenophon. Given that on his best day he's a wet Liberal, I'm not sure what that says about the strength and resilience of their vote.

    1. There's no button - Tasmanian House of Assembly elections are still done manually. There will be updates to the primary figures, expected on Wednesday and Friday, then through next week there will be a distribution of preferences running from Tuesday to Thursday or Friday. In the past these distributions have been updated frequently so we can follow them online by checking for new figures every few hours.

      Tasmanian local council elections now use data entry with a button press (at least, about half did in 2014; not sure if it will be upscaled to all this year.)

      Yes the Senate is a similar thing. The Tasmanian Greens Senate vote tends to ebb and flow on a similar sort of basis to their state vote. At the 2016 Senate election in Tasmania their vote was harmed by competition from Labor Senator Lisa Singh.

  8. Surely the collapse of Labor and Green support after the 2014 election was due in most part to the GFC induced recession which was a short-term thing anyway... and people's perception it was the fault of the Greens for the cutbacks and decline in economic activity, with Labor getting hit equally hard by those circumstances. The Tasmania Forest Agreement was largely independent from politics too although I can see it would have created an unstable atmosphere in that sector whilst the talks about the long term future of the industry was ongoing. The idea that the Greens cause problems when in government doesn't wash with me but its become a clever mythology since 2010 ramped up by conservative politics and certain business interests. I agree that the Greens ran a terrible campaign, bad candidates, no focus, poor presentation from the top.

  9. @Matt I think if you scarctch below the surface you will find politics, particularly green tinged ones, the primary driver behind the Forest Agreement. Gillard had done a deal with Brown to extend theTWWHA in return for support. This was a major driver behind TFA outcomes - much more than ever reported in the media. Local Greens were in on it too.

    1. Really chris. Why did all the signatories of the TFA - the Unions and major forestry industry players - arch enemies of the environmental movement - stick with the agreement, and pleaded with the Liberals to honor it, including in front of Leg Co committies, well after the Liberals 2014 election win? Were the Greens dictating their views even after their defeat from government? hmmm I don't think so. I am amazed at how many people believe greens and Greens dictate the states economic progress and development, on everything. They are a convenient scapegoats for the establishments incompetence...

  10. "The Greens are currently on 10.03%, though there is a fair chance that absent and other remaining votes will lift them just above their worst-ever score of 10.18% (1989)."

    That should be 1998 rather than 1989. The Greens did very well in 1989, winning over 17%.