Wednesday, March 22, 2017

Poll Roundup: Attack Of The Poll-Shaped Objects

2PP Aggregate: 52.8 to Labor (-0.8 since last week)
Labor would comfortably win election "held right now"

This week has been a confusing week for many poll-watchers, and an amusing week for those of us who watch the antics of the poll-watching partisans.  Newspoll was eagerly anticipated following its 55:45 to Labor three weeks ago, and widely expected to come out with more of the same if not then some, but pulled up at only 52:48.  Oh well, the line went, maybe it was never 55:45 to begin with, after all Essential had only got out to 53s and the odd 54.  But then Essential jumped to 55:45, so two polls not usually noted for volatility had delivered it not just in spades but also in opposite directions.  At the moment we don't have a third opinion, since the others are very inactive lately.

In trying to decide between these competing figures, it is worth bearing in mind that Newspoll is now administered by Galaxy.  Galaxy was the best pollster of the 2013 election, and the Galaxy/Newspoll stable triumphed again in 2016.  Essential was poor in 2013, and while its final poll in 2016 was very good, its tracking performance suggests it was probably lucky or herding.  So my aggregate comes down more on the side of Newspoll, crediting Labor with 52.8% 2PP.

Anyway, that is a recovery of sorts, but we have seen false dawns before.  Here's the smoothed tracking graph:


For anyone with an interest in my input values, this Newspoll went in at 45.1% for the Coalition after considering the primaries, and the last three Essentials have been 46.9 (published 47), 46.7 (47) and 45.2 (45).

The three-point rise in Newspoll was interpreted by the Australian as caused by Prime Minister Turnbull's Snowy River 2.0 hydro power announcement, but it could have been caused partly or entirely by many other things or none.  It might even be random bouncing. In particular, if a policy change announced by the PM is supposedly the major cause of a three-point rise on voting intention, one would have expected a much bigger rise in the PM's own ratings.  This wasn't the case.

Leaderships

This week's Newspoll had Malcolm Turnbull's net satisfaction up three points to -27 (30-57) with Bill Shorten down two to -28 (29-57).  The last time we saw both leaders polling this badly at once was just before Tony Abbott was rolled.  Turnbull has jumped out to a 43-29 lead as "better Prime Minister", his second-highest lead since the election, but that's not much to write home about.

Roy Morgan Research had a leadership poll release that made me think "Morgan poll, hmmm, oh, is that still a thing?" Morgan found Turnbull to have dropped two points since October to -24 (30-54) and Shorten to have dropped ten to -28, the same as Newspoll (28-56).  Morgan has Turnbull up 49-32 as "better Prime Minister".  Finally, last week's Essential also found both leaders in the slop with Turnbull at -17 (33-50), his worst to date, Shorten at -19 (30-49), his worst since a year ago, and Turnbull up 38-26 in the least-ugly contest.

Morgan also had a preferred leader poll that suggests some of the modest-to-begin-with support for making Tony Abbott Liberal leader again has gone over to Peter Dutton, but otherwise didn't report anything new.

I won't cover the other reputable polling results in much detail this week because in the last few weeks they have been outnumbered by polls that were at least badly reported and in cases unsound.   What follows is a selected parade of poll-shaped objects: public releases based on "polls" that have been poorly reported, are badly designed or that fail appropriate disclosure standards.  (The term is by analogy with piano-shaped object, something that appears to be a piano but sounds horrible or is impossible to play - whether for reason of shoddy construction, mistreatment or decay.)

I should make it clear that in many cases below there is nothing evident wrong with the poll, and the issue is the way the media have trumpeted it.  In other cases the polls themselves have problems.  I've left enough information for readers to hopefully work out which is which.

PSO 1: Peter Dutton Losing His Seat? 

Wirrah Award For Fishy Poll Reporting (image source)

I'll start with this one because I think it's one of the worst examples of bad poll reporting I've seen in a while, but it also underlines the importance of clear communication between pollsters, sponsors and the media and immediate publication of full results.  The Australia Institute has done a ReachTEL of Peter Dutton's seat of Dickson.  TAI-commissioned polls frequently purport to show that voters in marginal and even safe Coalition seats actually support ideas espoused by the political left, especially in the field of renewable energy.  These polls are common and I don't have anything new to say about them for now, but what was interesting about this one is the reports on voting intention polling concerning Peter Dutton.

Polling surrounding Dutton is of interest at the moment because there is, for some reason, speculation that he might be the next Liberal leader; indeed he remains the $1.87 favourite to do so with one bookmaker.  That said he has been barely a blip on the radar in smorgasbord polling since the hype began, scoring a massive 2% (the same as Christopher Pyne and Scott Morrison) in Essential's effort before breaking away from these formidable rivals to lead 5% to their 4% each in this week's Morgan.  Malcolm Fraser went from 5% to leader in three months, so there is hope for #Pynementum yet...

Even if he never becomes leader, Dutton is an influential Coalition head-kicker (and heel-dragger) and so the fate of his marginal (1.6%) seat is of much interest.  And this certainly got some attention when the figures first came out on the Guardian website and on Twitter.

The article's original subheading had "Poll says immigration minister would lose election if held today", and then the article went on to say the following, among the other things it still does say:

"The poll also shows a large One Nation primary vote in the seat (16.8%) and Dutton on track to lose his seat by five percentage points if an election were held today.

[..]

The poll found Labor leading the LNP in Dickson on 55.7% to 44.3% in two-party-preferred terms.

[..]

The ReachTel poll suggests Dutton would lose his seat if an election were held today but, as respondents were asked to allocate their own preference between Labor and the LNP, it may understate the flow of minor party preferences to the LNP.

[..]

Some 3.7% said they intended to vote for another candidate and 5.5% were undecided. Exactly half (50%) of undecided voters said they leaned to the LNP."

The problem was that by the primary votes published by the Guardian, and ignoring the information about undecided voters for want of a full breakdown, the poll would come out at 50.2% to Dutton by nationwide last-election preferences.  And, while the use of respondent-allocated preferences can skew polls by a few points, six points seems a bit much.

Questions were asked, and soon all the references to Dutton losing vanished from the article, as did the line about half the undecided voters leaning to the LNP (which was strange since that bit was actually true.)  A note was added to this effect: "This story was amended on 21 March to remove an incorrect reference to two-party-preferred share in the poll."  - a note which went nowhere near describing everything that had been removed.

This raised so many further questions.  Was there ever a 2PP?  If there was, was the 2PP incorrect as supplied by the pollster to TAI, as supplied by TAI to the Guardian, or had the Guardian misinterpreted it on their own accord?  If there was a real 2PP what was it, and why was it not being reported instead?

It has taken much of the day for the facts to become clearer.  The respondent-allocated 2PP was 52% to Peter Dutton, which would be an encouraging result for him if seat polling was more reliable than it is.  This is pretty close to the 51.0% obtained by 2016 preferences using the full primary figures.  The source of the confusion was that some recent ReachTEL results have had a question 1b that asks the respondent who they would preference out of Labor and the Coalition, but this question is only asked of those who did not already choose Labor or the Coalition.  So 55.7:44.3 is not the two-party preferred for the poll - it is the two-party preferred among the non-major party voters.

ReachTEL have stated on Twitter that "The original document did have a 2pp but it sounds as though that wasn’t communicated."  The version that appears on the Australia Institute website has no 2PP and no clear indication that question 1b isn't the actual 2PP.  I do not know what explains this discrepancy, but what is clear at the moment is that the Guardian article made much of the supposed fact that Peter Dutton was trailing badly but there has been no rush from the Guardian to fully correct the article and ensure readers are informed of the real story.

Amusing as it is, poll watchers shouldn't have to be chasing up this nonsense to hold poll reporting to account.  It should be standard practice that once a poll is referred to in a media publication, the full details of the poll are immediately published, either on that publication's website, or as a link from the article.  The howlers that are being caught are the tip of the iceberg.  Among all the reported results of polls that have been seen only by a journalist and whoever showed them the poll, there are bound to be many errors that cannot be checked.

(And by the way, no, I don't support subsidising the Adani coalmine either.  I'd probably be up for paying it to go away.)

PSO2: Public Support For Amending 18C?

On the other side of the coin, we have the Australian's doubtless specially-commissioned Newspoll on proposed changes to Section 18C of the Racial Discrimination Act, an issue that said paper is notoriously focused on.  Here we did get the question wording upfront, and also interesting party breakdowns (showing that the issue is a lot less partisan than culture-warriors would think) but the problem is the preamble is skewed:

"The Racial Discrimination Act makes it an offence to 'insult' or 'offend' someone on the basis of their race. Supporters of the act say it protects people from racial abuse but opponents say it goes too far and limits free speech.  If the words 'insult' and 'offend' are replaced by the word 'harass' in the act, it will set a higher benchmark for complaints. On balance, would you be in favour or opposed to this wording change to the Racial Discrimination Act?"

This preamble is misleading firstly because it creates an impression that insulting or offending someone on the basis of their race is always illegal.  The preamble fails to mention the Section 18D defences which provide protection for insulting and offensive statements under some circumstances. The preamble also fails to explore the distinction between an offence and an unlawful act.

Another sneaky word in this preamble is "but".  The form of words A-says-X-but-B-says-Y makes it sound a lot more like A is wrong than if the "but" is not there.  Then the next sentence is simply flying a flag for a positive quality of the proposed change and blatantly priming the respondent.

Given the level of skewing in this preamble, the overall result (47-39 in favour of the change) is about as useful as an Andrew Bolt attempt to determine somebody's ethnic identity.  Next ...

PSO3: Daniel Andrews Non-Preferred Premier?

If Daniel Andrews' Victorian Labor government is in serious polling pain then it utterly deserves it.  The recent entitlements scandal that saw both the Speaker and the Deputy Speaker quit and the latter go to the crossbench is exactly the sort of thing that makes people disgusted with politicians in general, especially when it seems that pollies of all parties simply rotate whose turn it is to engage in compulsive entitlement rorting.  After Geoff Shaw in the previous parliament this is the last thing public trust in politics needs, and manna from heaven for populist parties crammed with people who given the chance would probably be even worse.

That said, the Herald-Sun's coverage of a recent commissioned ReachTEL supposed to show Labor trailing 46:54 (Labor Party faces Victorian election 2018 wipeout - may be paywalled) was especially concerning.  It published what was clearly somebody's internal/"private" polling of some sort without saying whose it was, and without even making it all that clear that this poll was something different to the normal run of media polling.  All we got by way of disclosure was "The Herald Sun has obtained a ReachTEL poll, the first commissioned since an expenses scandal claimed the scalps of the Speaker and deputy speaker, [..]"

A possible hint within the poll results comes from the commissioned question about the Firefighters Union boss, suggesting that this could be someone who either has an ax to grind over the issues that led to the departure of Jane Garrett, or someone who just thinks the issue is useful for attacking Andrews on.  That doesn't narrow the field all that much.

Anyway, at least one commentator (I don't now have the link) seized on the finding that Andrews was trailing as better premier and noted how unusual this was for an incumbent, but an important point was overlooked here.  The way in which the question was commissioned, giving options of Andrews, Guy, Other and Undecided, is completely irregular by including an unspecified "Other" option, and therefore there is no benchmark to assess it against.  Even a party breakdown would be useful here to say more. My suspicion would be that when a diehard Greens voter hears Other they will often think "That's us! They've given us a question to be special on!", and hence that this method is not kind to Labor.

PSO4: Federal Libs Losing In The West?

Ah yes again we go, the wailing sirens of looming electoral doom as a sensational new poll makes a startling finding about a swing that is caused by a single thing, the GST carveup, and is not in any way contaminated by the state election held a few days earlier.  Dubious timing aside, there's another issue with this one: respondent preferencing.  Based on the converted primaries, the 2PP by last-election preferences comes out to 50.2 to Coalition using national preferences, or 49.9 to Liberal using WA-specific preferences (which includes treating Nationals as separate) but the national figure for One Nation.  Yet the published 2PP was 53:47 to Labor.

While there have been some recent state elections with massive primary vote swings where last-election preferences were all at sea (Queensland and NSW) those were conducted using optional preferential voting.  In WA last-election preferences (using the last federal election or an assumption for One Nation) seem to have worked better than respondent preferences overall despite a substantial swing in Green preferences to Labor, and the track record of last-election preferences at federal elections is very good.  It's useful to have respondent preferences to keep an eye on shifts and on any evidence of potential for last-election preferences to be badly wrong, but it baffles me that anyone would use them as a headline without even publishing a last-election result.  In this case the choice of preferencing method makes about two seats' difference to the projected outcome. Let's see how this goes when the new state government has some time to settle in.

PSO 5: Christian Porter To Lose Seat Over Marriage Equality?

Hang on, wasn't he losing his seat over the GST carve-up as well?

My Twitter profile has a rainbow flag and describes me as "Totally pro-SSM".  This probably looks like virtue signalling to someone out there but it is actually more like a beware-of-the-dog sign.  It's a warning to moral reactionaries that I may rant about marriage equality from time to time and if they expect me to only tweet about psephology, then I expect to flame them to a crisp.

The reality is that I would do anything for equal love, but I won't do that, where that is let people on either side get away with sloppy poll-reporting practices.  Polling data belongs out of the closet, fully out and parading proudly, not just half-out to a few special friends. The findings of a recent campaign-commissioned poll are quite emphatic, especially as there have sometimes been issues with obtaining results fully comparable with other polling methods by robopolling in the past.  I don't see any reason to doubt the primary result that voters in the listed conservative seats support marriage equality.  But when I see a barrage of results to different questions reported together I'd at least like to see what all the questions were and what order they were asked in, just in case the earlier questions might have potential to influence the later ones.

And when it comes to that staple of issues polling, the "would you be more likely to vote for blah if blah takes this position on issue X?" type question, it should be well understood that this type of question has no predictive value whatsoever.  All that style of question shows is that voters will say that almost anything is important and might change their vote in isolation.  It says nothing about the relative salience of different issues when they come to deciding votes.  Today I told a pollster that adult literacy and numeracy rates were a "moderate problem" in Tasmania, but if a charity caller rang up asking me to donate $1 to a literacy and numeracy campaign I'd tell them I wasn't interested.

The lack of salience of marriage equality as a vote-shifter has been frustrating, but it is what it is. Fear not; the Coalition has a vast capacity for self-harm when it comes to gay rights issues, and has shown in the past that it's perfectly capable of destroying itself on such matters without needing any help at the ballot box.

-----------------------------------------------

There are so many more poll-shaped objects out there.  Every few days on average I will come across a new one.  I've even considered setting up a monthly spreadsheet of them all, categorised with which forms of polling or reporting failure each one displays, but I'm not sure I could keep up with the workload!

Friday, March 17, 2017

White New Labor Leader, But Who Will Take Green's Seat?

It's been a huge day in Tasmanian state politics with the resignation of Labor leader Bryan Green and his unopposed (at least within the PLP) replacement by Rebecca White.  White is the youngest ever Tasmanian Labor leader and will be the youngest Premier by a few to several months if she wins the next state election.  (She is not, however, the youngest Labor leader nationally - Chris Watson, later to be PM, was probably one day younger when he became the first federal Labor leader.  There may have been other younger Labor leaders in other states; I haven't checked.  She is also not the youngest Tasmanian major party leader - Liberal Geoff Pearsall was 32 in 1979.)

Bryan Green is the second long-term Labor leader after Neil Batt (leader 1986-88) to not contest an election.  While Green was  uncompetitive in head-to-head matchups with Will Hodgman (even after allowing for the edge to the incumbent on such measures) he oversaw a time in which the parliamentary party was almost always unified in public and bloodletting following a massive loss in 2014 was contained.

Green also oversaw a victory in the Legislative Council set of Elwick and significant rebuilding of the party's standing in state polling.  I estimate that Labor has so far recovered about six points of support since the 2014 election disaster, while the Liberals have lost about ten. His decision to quit is no surprise after questions were raised about his commitment level, and a strikingly poor preferred premier score in the recent EMRS poll.  If there is no more to it than what we know, then to stand aside now will probably be seen as a selfless decision in order to spare the party from prolonged leadership speculation or a badly compromised and easily attacked campaign.  Such speculation was seen at placing the party at risk in the event of an early election.

Rebecca White is popular - she was the standout choice of voters in a ReachTEL poll of the Labor leadership last November. Although an MP (and a Minister for a month) during the Labor-Green coalition years, she is likely to be seen as a fresh choice who is not especially tarnished by the party's past.   The attack on White will be primarily that she is inexperienced, and also perhaps that she is left-leaning and likely to work with the Greens.  However, her union connections are light on compared to other possible targets.

Identity politics is also bound to raise its head at some time, though the Liberals would be well advised to keep clear of it as a line of attack themselves. Female leadership figures are criticised for putting career before family if they have children and also criticised for the same thing if they don't.  Indeed, the change today surprised some journalists who seem to have wrongly assumed White would definitely not take the job so soon after the birth of her first child.

The change is likely to be bad news for the Greens vote.  Another young left-wing female Labor leader is likely to take a lot more votes from them than Bryan Green ever would have done.  On the other hand, if the change increases the chance of a hung parliament it may help them in the long term.

My previous Tasmania piece essentially argued, without directly saying it, that Labor needed to change leaders because of the dynamic of Hare-Clark elections.  Labor would want to be polling well enough in the campaign to at least stop the Liberals from saying that only they could form majority government.  This change removes a liability, but I do not think Green was much of a drag on the party's current primary-vote polling.

Bryan Green's decision to resign from parliament entirely, instead of retiring at the next election. creates what could be a very interesting and perhaps close Hare-Clark recount.

The stakes are major here.  Former MP Brenton Best, who narrowly lost his seat in the 2014 election, is someone Labor would desperately not want to return.  A career backbencher during eighteen years in parliament, Best was not a team player in the previous parliament and could be even less so (perhaps even moving directly to the crossbench) should he return.  Labor would even face the prospect of going into the next state election with no incumbents in Braddon, and the Liberals would have a field day with it all, while Best's potential presence on the ballot paper as an independent incumbent would be the last thing Labor needed. The other serious contender, Shane Broad, an agricultural scientist, would at the very least be a new face in parliament, and quite likely a source of new policy depth. Unusually for a countback winner (many of whom serve a year or two then lose and then are never seen again) Broad would bring an already substantial personal vote.

Who Will Take Green's Seat - If Broad and Best Contest

(Warning: this section is very numbery and wonky (about wonk factor 4/5).  If you just want a quick summary of the answer - it is very hard to say!)

In Hare-Clark recounts for casual vacancies, only the votes the elected member had when they were elected are included.  It makes no difference who polled the most primaries or who came closest to getting elected the first time round, although usually the candidates who poll more primaries also do better on preferences.

The votes that will determine Green's replacement are the 10716 votes he held after crossing the line and being reduced to a quota.  These are each thrown to the highest placed candidate on them who is contesting the recount.  If someone gets more than 50% that person wins, otherwise whoever is last is excluded and their preferences then thrown, and so on, like a single member election.  Because Green was not elected on primaries, these votes come from a wide range of sources and in some cases we know things about where they will go in the recount.  The breakdown of Green's votes is:

6606 (61.7%) Bryan Green (ALP) primaries
1602 (14.9%) preferences from Shane Broad (ALP)
721 (6.7%) preferences from other Labor candidates Justine Keay and Darryl Bessell
637 (5.9%) preferences from Palmer United, mostly Kevin Morgan
614 (5.7%) preferences from the Greens, mostly Paul O'Halloran
245 (2.3%) preferences from the short-lived "Tasmanian Nationals"
232 (2.2%) preferences from Liberals, including the surpluses of Adam Brooks and Jeremy Rockliff
36 (0.3%) preferences from ungrouped independents
23 (0.2%) preferences from Australian Christians

Here a feature of Hare-Clark recounts that I call the Hare-Clark recount bug rears its head.  Because Shane Broad was excluded before Bryan Green was elected in the original election, votes that went from him to Bryan Green (including those that are, say, 1 Broad 2 Green) then return to him in the recount.  However Brenton Best never had his preferences distributed in 2014 (let alone while Green was still shy of quota) and so a vote that is 1 Best 2 Green will never be seen in this recount.

This means that Best is disadvantaged by lasting too long in the original election while Broad is advantaged by having been excluded early.  This is, of course, silly and unfair, but no sound solution to this problem has ever been found.  Holding a full recount of every vote cast with the retiring member removed from that count could have the impact of unelecting some other MP, or of a seat being lost to another party, both of which are problems likely to stop an MP retiring when they otherwise should.

How big is the problem for Best?  Well, at least 1602 votes, because all the votes that came from Broad to Green have Broad higher placed than Best, and will therefore pool with Broad if Best is his only remaining opposition.  So that's 14.9% Broad gets for free in the race to 50%.  Actually, because of exhausting votes, to a little bit less than 50%.

But there's actually more than that, because the 538 votes from Kevin Morgan (PUP) and the 569 votes from Paul O'Halloran (Greens) would have included some votes that would have flowed to Broad ahead of Bryan Green had Broad still been in the count, but would not have included any votes that would have flowed to Best ahead of Green.  That might advantage Broad by another, say, 200 votes, perhaps more.

In general we would expect Best to do better on preferences than Broad because Best was a high-profile incumbent whose independent streak would have given him some cross-ticket appeal.  And indeed, of preferences distributed directly to them from other parties in the original count while both were still in the race, Best received 697 to Broad's 366.  On this basis, Best should recover some votes from that portion of Green's vote that came from other parties.

However,  Justine Keay's votes from the original count flowed very strongly to Broad compared to Best (635-315) and presumably the votes that went from Keay to Green will do much the same thing.  On the other hand, Darryl Bessell's votes from the original count strongly favoured Best over Broad (217-93).

My suspicion is that Best would have to make up something like the 1600 votes mentioned above, with the rest about cancelling out, from Bryan Green's primaries alone.  Assuming almost none of those exhaust without reaching either Best or Broad, that would leave him needing a roughly 63:37 split of the Bryan Green votes over Broad (the split of Labor's primary votes between the two being 58:42).  The question then is whether Bryan Green's voters would have preferenced against Best, as Keay's did (in which case Broad will easily win) or whether they would have gone largely from one incumbent to the next (in which case Best could just get enough.) Or if they just break as per the other primary votes then I have Best falling short.  (A reader who wishes to be anonymous has sent me detailed modeling showing that if various preferences are assumed to break like the primaries or like the known preference distributions, Broad wins comfortably - though these are medium-sized "ifs".)

The 2010 election might provide some useful insight here.  At that election Green made a quota, and his surplus split 219-94 (70:30) to Best over Broad of votes that went to the two of them.  However, at that election Best polled 7087 votes to Broad's 3303, while at this election Broad was much more competitive with Best on primaries (3648 for Best to 2654 for Broad).  It stands to reason Best's share of Green's preferences relative to Broad must fall substantially - but by how much?

My suspicion is probably "enough" and on that basis I consider Broad a slight favourite to win this recount. 

(Update: as of 21 March, Broad is contesting. Best's decision is unknown.)

If Broad and Best don't both contest

If only one of Broad and Best contests, that one will easily beat anybody else.  (Just about certainly - there is an outside chance the third Labor candidate, Darryl Bessell, might beat Best if Best was really on the nose with Green's voters.) If neither contests, Bessell will win if he contests - enough preferences will stay in the Labor ticket that nobody else has a chance.  In theory if none contested, Labor could invoke a never yet used provision that allows for a single-seat by-election, but this won't happen; someone from the three will stand.   There is no  prospect of anyone from outside Labor winning Green's recount.

Best's Party Status

The fact that Best may now not choose to sit as a Labor MP is irrelevant.  He is still entitled to contest the recount and if successful be elected.  He does not require any permission from the party to do so or to make any decision on his party status when he nominates.  He can even decide to contest the recount and make a decision on his party status later if he wins.  His right to contest the recount exists as a candidate of the previous election.  See Section 228.

Some Further Reflections

As the recount saga has unfolded, it has looked more and more like Green's decision to quit his seat as well as just the leadership has been a personal one and not a party tactic.  It would have made sense as a party tactic if it was likely Broad would win the recount, but it is not clear that he will and not clear that he wants it.  The situation with Green brings greater focus to the recent calls for Lara Giddings and David Llewellyn to also resign their seats.  If both retire at the next election, Labor may go into the election with as few as four incumbents, a parlous situation especially considering the Liberals have stated that all fifteen members will recontest.

With the leadership change Labor would be very keen to press to take votes from the Greens and push to increase their polling towards around 40%, at which (given the variety of minor parties likely to split and waste votes between them) majority government might be distantly plausible.  We've seen swings on a similar massive scale in enough other state elections recently.  But I'm wondering why the existing government's polling isn't better.  Federal factors are not helping, and the government does have a few dud performers, but the state's economy appears strong; are things really so bad?  Perhaps some of the polls are simply wrong.

Thursday, March 16, 2017

WA Washup: Another One Bites The Dust

It's a familiar script.  Five conservative state and territory governments have sought re-election since the Coalition took power federally and only one of the five (NSW) has survived.  Western Australia joins Queensland, the Northern Territory and (less emphatically) Victoria as jurisdictions where Coalition incumbents have been given the boot in the last few years.

The outcome in WA has been on the cards for years, and the fact that the Barnett government lost breaks no new ground by itself.  Defeat could be predicted as probable based on a combination of what I call federal drag (being of the same party as in power federally), the age of the state government, and the federal government's poor polling.  Indeed even Colin Barnett's poor personal ratings alone suggested he was already likely to lose this election less than a year out from his very strong 2013 result.  There's a strong case that Barnett should have been removed at least a year ago. Those who failed to do so look quite silly now, but they are geniuses compared to those who wanted to replace Mark McGowan with someone not even in the parliament because McGowan was thought to be too boring to win the election!

In the usual search for explanations, the important thing is that, historically, "it's time" is not enough alone to explain this change of government, although it is a contributing factor.  State governments older than Barnett's have had about a 50% re-election rate in recent decades.  

The real news has been the size of Labor's victory.  Polling (as correctly interpreted) generally pointed to wins in the range from narrow to decisive, but the final 2PP looks like it could finish above 55%, which was slightly higher than the final polls.  Projecting off the average of seats for which 2PP swings can be calculated, I currently have the 2PP at 55.7% (a 13.0% swing), but this may well move by a few or even several tenths of a point as non-classic seats are included and remaining votes counted.  In my view a few points of the margin and several of the seat losses are most likely explained by state and campaign factors not captured in the overall framework above.  

Labor presently leads in 41 seats, with very few still in any real doubt.  At present they look like matching the seat gain predicted by uniform swing (given the swing actually recorded) and this is a good effort given the number of Liberal and National seats that were on margins close to that swing.  This has happened because the 11 seats that have recorded single-figure swings have mostly been very safe Coalition seats or Labor-held, leaving more swing to go around where it matters.  In some cases this has led to spectacular overkills, eg Southern River (target 10.9%, swing 18.9), Darling Range (target 13.1, swing 19.2) and Bunbury (target 12.2, swing 23.1).  But also Labor has won more than its fair share of the closest (<3%) contests - eight if it wins Joondalup to the Liberals' two.

Polling Accuracy

It will be a little while before we get full primary 2PP figures and can say for sure which of the few pollsters in the field was closest in their final poll.  On current primaries it looks like the final Newspoll was slightly more accurate than the final ReachTEL (currently root mean square error for the primaries for six parties is running at 1.61 vs 1.96), and both final polls were reasonably close.  However when it comes to the 2PP, Newspoll's 54:46 was based on last-election preference flows which will probably turn out to have been reasonably accurate for the recontesting parties, while it looks like ReachTEL's respondent preferences will have overestimated the flow by party to Labor, as respondent preferences so often do.  This has cancelled out ReachTEL having the combined Liberal and National primary about three points too high and the combined Labor and Green primary wrong by a similar amount.  Newspoll had the Liberal and National primaries very close to right, but had a larger error on One Nation at the expense of the left.  

WA would not be that easy to poll and the One Nation vote was volatile through the campaign, so I think both these final efforts were pretty good.  One issue is One Nation and whether the pollsters both overestimated their final result by polling support for them in seats they weren't contesting.  The alternative hypothesis is that their vote crashed 2-3 points in the final days, but I don't really buy it.  After all their Upper House vote has held up well compared to the polling.  ReachTEL, at least, had One Nation on the readout in all seats although the party was only contesting about 60% of seats.  ReachTEL's question was also not house-specific.  

However, when it came to tracking, the Galaxy/Newspoll story (which had the vote at 54:46 in all four polls this year) has seemed more credible than the ReachTEL story (52-50-52-54 in their last four polls).  Some bounciness from sample error isn't a problem but I do not believe that the Liberal Party dropped three points in the primary vote in the last two weeks at the same time as One Nation really dropped a similar amount.  There seems to have been some skew in the ReachTEL primary vote polling which, again, was partly cancelled out on the 2PP by using respondent preferences. 

That One Nation's vote did crash during the campaign (thus the low-teens votes for them recorded by pollsters early on were probably true) has been supported by figures tweeted by Antony Green.  One would not think One Nation voters would be the most organised bunch when it came to non-booth voting and they weren't at the last Senate election, where their non-booth vote was only trivially higher than their booth vote.  

It is also notable that both final polls understated the Green vote.  There's a first time for everything! However, through the campaign up to that point polls had, on average, had the Green vote about right.  

What was released of ReachTEL's commissioned polling has scrubbed up better than their statewide polling!  The Parenthood poll of six marginals has probably tipped all six winners (depending on the fate of Joondalup) with an average error of 1.5 points against Labor, which could be easily entirely down to shifts before polling day - none of the six seats were out by more than 4%, which is hugely impressive given some of the seat polling failures we have seen in recent years.  Shame the full details were not published.  The multi-marginals poll seems to have underestimated the swing by about 2.8 points, but some of that could well be down to swing in the last two weeks too.  

One Nation

At the start of this campaign I thought there was an outside chance Pauline Hanson's One Nation might do really, really well, as has happened at times in the past, but in the end their result was poor compared to their early polling.  They are currently on for a couple of Upper House seats, but that is all.  

The One Nation - Liberal preference swap was brand-damaging for One Nation because the party was seen as having a bit both ways about whether it was against the establishment (especially in the unpopular form of Colin Barnett) or part of it.  It contributed to much of the infighting seen in the campaign, although the party was also dogged by inappropriate candidates and dumb comments about vaccinations by its leader. 

It will take some time to get a final figure on the split of One Nation preferences, but even when we have that it will not tell us much about to what extent the split was driven by the deal.  The reason for this is that should One Nation preferences have skewed markedly to the Liberals, a possible explanation will be that the preference deal caused many One Nation voters who would have preferenced Labor to simply not vote 1 for PHON at all.  

It does look like One Nation preferences favoured the Liberals in at least some seats (Wanneroo looks like a pretty clearcut example) but I do not see evidence that they did so all that strongly on a statewide basis.  

There are currently only two seats that the Liberals have won narrowly against Labor (Dawesville and Geraldton) and if the One Nation preference decisions made the difference in those seats, then the preference flow from One Nation to Liberal would have to have been 13-14 points stronger than it would have been without the deal.  That seems unlikely.  What damage a deal that saved at the very most two seats did to the Liberal primary in the eight seats apparently lost by less than 3% is impossible to say.  Four of these seats did not even have One Nation candidates. 

Upper House

There is still a long way to go in the Upper House count.  The counts on the WAEC site are currently mostly showing at between 70-74% complete, with the exception of Mining and Pastoral which is 57% complete.  These figures appear to include informal votes, meaning that they will probably get up to about 92%. (The slightly lower completion figures on the ABC site may exclude informals, in which case they should wind up at around 89%).  

There are two things that can shift the provisional results shown by the ABC.  The first is that even small changes in the relative vote shares of various parties result in different exclusion orders (to which group ticket systems are so sensitive) and perhaps even enable a party to get through a point at which it is currently excluded and win.  If certain kinds of votes, such as non-booth votes, are underrepresented in the counting, then this sort of thing becomes more likely.

The other is that in the case of a very tight result, the doubtless very small rate (I haven't seen figures on it yet) of below-the-line voting may make the difference.  Below the line votes tend to hurt parties that are trying to snowball from a very low primary vote through preference deals.  

In East Metro, for instance, this is currently an important exclusion point to watch:


The Fluoride Free WA preference snowball currently starts on a primary vote of 0.35%, at which point it leads only five independent tickets, of which two are fakely independent tickets created by Flux the System. Fluoride Free is currently shown as overtaking ten other tickets in inflating its vote 25 times by preference harvesting, before finally being taken to the dentist at the point shown above.  If Pauline Hanson's One Nation fall below Fluoride Free at this point, then John Watt goes on to win.  
However, if the ABC model showed Watt getting over PHON here by only a handful of votes, I would ignore it.  The reason is that PHON is starting from 7.8% and much less affected by below-the-line votes that spray everywhere instead of following the tickets of the parties supplying them.  Fluoride Free, however, has to get almost all its votes from other parties, and any vote that is below the line and doesn't flow to it is one less vote it has at the juncture above.

On the current numbers, none of the micro-parties with really negligible vote shares are actually getting up, and I hope it will stay that way.  But micro-parties are still likely to win one or two seats by preference harvesting.  The Shooters, Fishers and Farmers in Agriculture are currently shown beating One Nation despite having 5.5% compared to One Nation's 11.3%.  In Southern Metro, the Liberal Democrats with 4.1% (most of it arising from ballot paper confusion as happens with this party in the absence of logos) are shown as beating the Greens with 8.3% and One Nation with 6.8%.  Neither of these results would happen under a system where genuine voter choice determined the result.  (The other potential Shooters, Fishers and Farmers seat, in Mining and Pastoral, might well have been won anyway.)

As a sign of the severe malapportionment of the WA upper house, Labor still looks like needing 3-4 votes from the crossbench to pass legislation, despite having massively won the election in the Lower House.  That said it is simplistic to lump the Shooters, One Nation and even the Nationals as "conservatives" and assume they will all be obstructive. 

I will probably add some updates on the Upper House count over coming weeks.

Thursday 16th, 6 pm (WA time): Labor has taken the lead in a 16th Upper House seat, as they are now in theory beating the Shooters, Fishers and Farmers in Mining and Pastoral.

Thursday 8:30 pm (WA time): Labor has now lost the notional lead to the Greens in South West, but the holding of the notional lead is meaningless while it remains as close as it now is (11 votes).  Labor would be more vulnerable to losing votes from leakage, but the Greens are slightly more dependent on votes from other parties, so in the absence of detailed modelling this one is likely to be determined by later changes in the vote margins.  (And yes, as of half an hour later, it's flipped back again!)

For those wondering about the majority situation - because the Legislative Council has 36 seats, a government that provides the President needs to have 19 seats onside to pass bills.  If a crossbencher agrees to be President then this reduces by one.  However certain bills, like electoral reform bills, require 19 votes (not 18) on the floor.

Saturday: See William Bowe and Antony Green for informed speculations on how the inclusion of below the line votes in coming days may affect various tipping points.  The chances of Labor and the Greens managing a combined majority on the floor are now not looking too bad.

Monday, March 6, 2017

EMRS: Liberals Crash, But Hodgman Still Clobbering Green

EMRS Feb 2017: Lib 35 ALP 29 Greens 19 Ind/Other 11 One Nation 6 
On these numbers a hung parliament would be inevitable (approx 11 Liberal, 10 Labor, 4 Green, though one more Liberal seat might fall to One Nation or the Greens)

EMRS Nov 2016: Lib 40 ALP 28 Green 18 Ind/Other 13 (One Nation not in readout)
On these numbers most likely result would have been a hung parliament (approx 12 Liberal, 10 Labor, 3 Green)

Current seat aggregate of all polls: Lib 12 ALP 10 Greens 3

Note: EMRS tends to skew to Greens and Others and against ALP.  No evidence on skew for or against One Nation is known.

--------------------------------------------------------------------------------------------------------

Once again, Tasmanian phone pollster EMRS has released two of its quarterly polls, for November and February, in a single release.  See also the trend tracker, which shows that the Liberal vote has been falling for four years now.

For the first time, One Nation has been included in the readout, and immediately the Hodgman Government has lost five points to 35%, its worst position of the term.  Even the Roy Morgan series, which was obviously skewed against the Government (and hasn't been seen since October) never had it below 37.

Patchy Polling In WA As The Final Week Begins

Going into the final week of the WA campaign, not much has changed from where it started a month ago.  The relatively scant and yet surprisingly diverse nature of polling data available leaves poll-watchers free to choose their own adventure, from a comprehensive win for Mark McGowan's Labor opposition to a very close and messy race in which Colin Barnett's Liberals might even cling on in minority by the skin of their teeth.  About the only thing the statewide polls agree on is this: following a shambolic and confused campaign, One Nation may have tanked.

The following table gives all the statewide polling known to me.  The poll marked as "Essential" is known to me only from a Poll Bludger exclusive that describes it as Greens-commissioned Essential robopoll.  That description places it at least two degrees of separation from anything we can place verified trust in but I mention it all the same.  I've also given an average and a time-weighted (but not performance-weighted) aggregate, in each case excluding the Essential.


In this case though, I don't place that much confidence in aggregation methods. There is a major house-effect difference between the ReachTELs and the Galaxy/Newspoll stable, worth at least four points on each major party's primaries, and coming out at two or three on the 2PP only because ReachTEL use respondent preferences (which have been skewing massively to Labor).  There's a pretty good chance here that someone's right and someone's wrong.  We're not just seeing margin of error issues here - there would not be such large and consistent differences in the major party primaries by pollster if we were.  (And no, voting intention doesn't bounce around this much in reality through a campaign either.)

Saturday, March 4, 2017

Groundhog Day: Group Ticket Nonsense Returns To WA

Glenn Druery will not say exactly what he's been paid to help micro-parties use preference-harvesting to win Upper House seats they don't deserve at the WA State Election. If it was only $5,000 per party plus a $50K win bonus as claimed (and denied), then his services have come pretty cheap.  The game is the same as it ever was: to give parties with very little support a chance at winning they don't deserve, by exploiting inflexible voting systems to create preference flows that have nothing to do with the intention of voters.  Druery trollishly describes this as an "outbreak of democracy".  I will bet that he can scarcely believe his luck to still be in this business.

After the debacle that was the 2013 Senate election in WA, one would have thought WA would be the last place on earth that would let Druery still ply his trade.  Alas, it looks like it will instead be the last place on earth that ever stops him.  It was in WA in 2013 that Wayne Dropulich of the Australian Sports Party (whatever that was) surfed from 0.2% of the vote to a Senate seat as a result of preference-harvesting, only for his election to be annulled because the loss of some ballot papers caused an irrelevant tipping point to become irresolvable. (This, in turn, was a product of the group ticket system.)  And it was then WA where the whole state's Senate election had to be rerun from scratch in 2014 at immense cost.

It seems quite a damning indictment on the Barnett government that it has had three and a half years since the 2013 debacle to clean up the state's similar Legislative Council voting system and hasn't even introduced a bill to that effect.  By comparison, the model being considered in South Australia is pretty bad, but at least South Australia's government is trying.  Whether WA's has just had too many other problems to care about democracy, or else has kept the system to deliberately salt the earth for its successor, I don't know.

WA's upper house has the worst state electoral system in the nation.  It is badly malapportioned in favour of rural electorates, it has Group Ticket voting, and it has a ridiculous lack of savings provisions for votes that stray off the narrow path of exact formality.  What we will see in the WA Upper House next weekend is barely even fit to be considered an election.

Wednesday, March 1, 2017

Poll Roundup: One Nation Soars As Liberals Squabble

2PP Aggregate: 53.8 (+1) to ALP
Labor would easily win election "held now"
Coalition's worst position of the current term so far
On current polling One Nation could win at least three lower house seats

Normally I go a couple of Newspolls between poll roundups these days, but this week's has been one of those Newspolls.  Following a conveniently timed "Newspoll bomb" by former Prime Minister Tony Abbott, the Turnbull government has recorded a new worst set of figures, and leadership speculation is rife.  This comes on top of a Fair Work Commission decision to cut penalty rates, which is seen as bad news for the government although the process was set in train and, until its outcome, supported by Labor.

We are still getting very little federal polling apart from the weekly Essential readings and the slightly less than fortnightly Newspolls.  The latest Newspoll came in at 55-45 to Labor, the highest reading for the Opposition since March 2015.  (In total during the Abbott Prime Ministership Labor recorded four 55s and one 57.)  I've aggregated it at 54.8 after processing the primaries.  The last few Essentials were more restrained (typically for Essential) at 52, 52 and 53 for Labor, which I aggregated at 52.4, 52,2 and 53.0.  Overall, largely on the back of the recent Newspoll, my aggregate has for now gone to 53.8% in Labor's favour.  This is the first time in this term that it has exceeded the 53.6 at which Tony Abbott was disposed of in the term before.