Dr Kevin Bonham
THE FIRST EMRS statewide poll of the year again shows the Bartlett Government in a reasonable position with a stronger chance than many realise of retaining its majority at the 2010 state election. The poll has been flagged by some sources as a poor one for the Greens, but yet again this is a case of commentators making too big a deal out of meaningless variation.

First, the basic numbers. The poll is online at http://www.emrs.com.au/pdfs/State%20Voting%20Intentions%20February%202009%20Report.pdf . Labor is up four points to 34, the Liberals up three to 29, the Greens down three to 15, “Independents” are on two, and 20% are still waiting for EMRS to ask the question in a way that makes them want to answer it. In the mirror-mirror-on-the-wall contest, Premier David Bartlett is up four points to 43, Opposition Leader Will Hodgman is level at 29 and Greens Leader Nick McKim is down three points to 12. There is also a seat-by-seat breakdown of party support and preferred premier, on which a lot more later.

For the required notes of caution in interpreting an EMRS poll, see my previous article at http://oldtt.pixelkey.biz/index.php?/weblog/article/kevie/ (tsk, how dare the editor call me that!) and for an even more comprehensive treatment see http://oldtt.pixelkey.biz/index.php?/weblog/article/poll-worse-for-lennon-than-for-labor/. For an excellent basic wrap of the current poll (and to save me from writing at least 500 extra words) see Peter Tucker’s comments at http://tasmanianpolitics.blogspot.com/

As Peter notes in his article, fluctuations of a few points between polls are generally not statistically meaningful, and it was surprising to hear one radio source describe the Greens as having sunk to a “crushing” 15% (a figure which is actually about their long-term average). There may be some move away from the traditional repositories of soft votes, the Greens and the undecided category, towards the major parties, but another poll will be needed to see if this is real or a one-off. In the previous history of EMRS polls, both major parties have only increased at the same time very close to an election as the undecided voters become committed.

The current poll (34-29-15) is actually extremely similar to results obtained at the same stage of the previous electoral cycle (35-31-12 in Feb 2005, 35-29-14 three months later). At that time Labor support under Lennon was collapsing and virtually no-one believed the party could win outright, but in March 2006 the party not merely won the election but did so easily, almost increasing its majority in the process.

If most of the voters flagged by EMRS as undecided are in fact soft Labor voters, things could be heading that way again. However, while Labor has replaced a divisive Premier with one who is apparently liked, the party has still suffered from losses of some key personnel and scandals affecting others, and it is unlikely that the Liberals and Greens will make the mess of the next campaign that they did in 2006.

EMRS have not published electorate-by-electorate figures for a long time, and it is a little surprising to see the practice return. It is disappointing that EMRS’s survey PDF does not explicitly caution the reader against reading anything much into such tiny samples. It was pleasing to see that Sue Neales (http://www.themercury.com.au/article/2009/02/13/55141_tasmania-news.html) picked up some slack for them in that regard in a report that apart from the usual stray exaggeration (“massive 41 per cent”) and the obligatory misuse of the term “approval rating” was generally reasonable.

However, to say that Labor is “strongly poised” to lose a seat in Braddon based on this sample just isn’t justified.

If the virtually meaningless electorate-by-electorate results from this poll were indeed repeated statewide, Labor would indeed lose its third seat in both Franklin and Braddon, and Denison as well, with Lyons unclear. But the Braddon sample (38-41-16) shows Labor losing the seat not to the Liberals, but the Greens. The Greens have never polled 16 in Braddon at an actual election, and are unlikely to do so anytime in at least the next decade. Their result last time was only 10.2%. I’ll be surprised if they can break 13% in 2010, and at around 12-13% it is still possible for the Greens to win the seat, but it depends on the vote splitting exactly the right way between the major parties. A miniscule sample showing the Greens on 16 does nothing to change my view that the chance of the Greens winning in this electorate is no more than one in five, if that.

Indeed, it’s surprising that so much fuss is being made about Braddon already. Only the semi-disgraced condition of two of the three Labor incumbents justifies it in any way at all, as Braddon is hardly low-hanging fruit if there is a modest uniform swing against the government. A swing of between 4 and 7% against Labor would most likely be needed for Braddon to fall. The lower values in that range apply if the swing is predominantly to the Greens, while the higher values apply if it goes more to the Liberals. Following the recent redistribution, Labor is estimated by Antony Green (http://blogs.abc.net.au/antonygreen/2008/11/final-boundarie.html) to be sitting on a 2006 base vote of 51.2% (which is 3.1 quotas once we redistribute “others”). It is even possible with the right split of votes between different candidates for Labor to lose 8% from that base and still get all three sitting members back. All that is needed to turn EMRS’s Braddon sample into figures that (however weakly) would point to a Labor hold is assume that the undecideds will strongly favour the Government and that EMRS’s polling has overestimated the Green vote. These are hardly radical assumptions.

I meant to supply an assessment of statewide swings needed to achieve particular outcomes at around this time following the redistributions. I have had some difficulties doing this because, in Franklin in particular, I have found it difficult to match up polling booths moved to numbers of voters supposed to be resident in moved areas. Nonetheless the general picture is quite clear and won’t change much based on the exact details of the redistribution.

As a result of the redistribution, Franklin is most likely already notionally gone. A simulated rerun of the 2006 distribution of preferences I did using the new boundaries and Antony Green’s estimate of the impact on raw party votes gave the final seat to Vanessa Goodwin by around 250 votes. The redistribution is the least of Labor’s worries, since it has lost incumbents Paul Lennon and Paula Wriedt and is defending two of the lowest-profile incumbents in recent electoral history (alongside Lara Giddings, who is popular but burdened with the Health portfolio) against the Liberal Party leader and, presumably, the now much higher profile Goodwin. With David O’Byrne expected to bolt in, there will probably not be any serious attempt to re-elect Ross Butler or Daniel Hulme (except from them perhaps). The question of whether David Bartlett can retain his majority therefore most likely hangs on whether he can get away with shedding one seat in Franklin and hold the line at 13-8-4 or 13-9-3.

In terms of a uniform state swing, the next seat to fall after Franklin would be Lyons, which drops off the perch with a straight swing from Labor to Liberal of around 2.8%. There has been virtually no attention given to Lyons, and it is interesting to note that Lyons has a very high “undecided” rate (around 35%), which may well be good news for the Government come election time. (It is difficult to be sure what the undecided rates for all the other electorates are as the claimed sample sizes for Bass and probably Franklin appear to include the undecided voters).

The very slight changes arising from the redistribution have probably made Bass notionally 3-2-0, albeit by a handful of votes. However that is meaningless as the Greens should be able to increase their vote in this electorate substantially on the back of the pulp mill issue (and, I suspect, an increased demographic greening of urban Launceston anyway). Also Labor will lose Jim Cox, a very strong electoral performer. At this stage Bass is the least likely electorate to see any change at the next election, with double-digit swings required for the Government to drop down to a single seat.

Denison is interesting because while modest swings from Labor to either party (3.7% straight to the Liberals or 4.5% to the Greens) could cost Labor its third seat, Labor can withstand a larger swing (up to around 6%) if it is split between the parties. But Labor is very well-placed in Denison with a popular Premier, as one of a diverse group of sitting Labor members, facing a new and lower-profile Green MHA, and a Liberal team with only one incumbent, the veteran Michael Hodgman, who has already brushed off one round of silly-season stuff about the timing of his retirement. Denison has also seen a bizarre squabble involving Hobart City Council alderman Marti Zucco, who has never previously attempted to line up with any political party in almost two decades in local government, and who has now been told he is not welcome to contest Denison preselection for the Liberals.

Zucco would probably not have been a huge vote-getter for the Liberals (probably no more than 2500 votes), given that he polls a modest vote at Council level, and many of those who vote for him at Council level would vote for Michael Hodgman or for Labor at state level, although he would also attract some votes from outside the Hobart council area, from those who do not vote in council elections, and from those who vote for other aldermen. It is likely that the Liberals can find plenty of other candidates who will poll at least the same sort of vote (perhaps more at the expense of each other) and without the risk. But the Liberals have made a mess of their public attempts to justify showing Zucco the door even before the preselection process, and have allowed the situation to look much more like a routine case of one preselection candidate using old dirt and rumours to squeeze out another than it probably is. Even Will Hodgman has been caught out with some curiously stuffy remarks about not letting his name get dragged into it all. Whatever his state ambitions, Zucco will most likely be laughing all the way to the retention of his council seat on the back of the free publicity.

As for the anonymous “Liberal source” talk of encouraging or forcing Michael Hodgman to stand down at or before the next election, this has followed the time-honoured pattern of destabilisation campaigns in which those wishing to undermine a candidate create persistent division and then call for it to be ended (in their favour, naturally). As the only sitting Liberal in Denison at the last election, Michael Hodgman polled 45.5% of his party’s total. This is comparable to the 46.9% of party total polled by Rene Hidding in Lyons, and Hidding was the party leader. (The only other Liberal in a comparable situation, Will Hodgman, polled an excellent 70% of his party’s total in that electorate.) Not only does Hodgman Snr still poll a reasonable vote, but the party needs him to attract the votes of pensioners and battlers across party lines, and would struggle if it had to go to an election with no sitting members at all or one who had only been there for a year.

These are the booths where Michael Hodgman recorded his highest proportions of the total Liberal votes in 2002: Cross Street (38.2%), Newdegate (37.8), Swan (37.7), Goodwood (37.1), Cornelian Bay (36%), Landsdowne Crescent (35.1%), Queenborough (34.5%) and Elizabeth Street/Mt Stuart (34.4%). And these are his best booths for 2006: Collinsvale (61%, but a very small booth with only 41 Liberal voters), Claremont (60.3%), Goodwood again (57.4%), Brent Street (56.7%), Abbotsfield (56.6%), Swan again (54.7%), St Peters (52.8%) and Bowen Road (52.2%). Most of these are not at all wealthy booths.

The argument was put that it is difficult for Will Hodgman to succeed as leader when the presence of his father in the same chamber makes him look like a “boy wonder”. When Will was outpolling Paul Lennon (at least according to EMRS’s dubious readings) shortly before the latter’s resignation, this hardly seemed to be an issue. I suspect the pressure for Hodgman Snr to depart represents a degree of exasperation at the lack of headway that the Liberals are now making. Will Hodgman was an excellent opponent to Paul Lennon because of the steady, cautious and reliable alternative he appeared to present to the controversial Premier. Against the flashy and cocky David Bartlett, that same steady cautious reliable approach is simply making Will and his party look a bit dull.

I suspect that Bartlett knows this, and this is why he is often so publicly respectful of his main opponent, while keen to marginalise and criticise new Greens leader Nick McKim and to paint McKim as an extremist whenever possible.

To summarise, Labor’s chances of retaining majority government are not as low as just about everyone is thinking. Their most likely path to doing so would be holding all of Lyons, Denison and Braddon and losing Franklin. There is no sign of action in Lyons, the Denison Liberals appear to be in a mess, and Braddon requires a rather big swing before anything happens.

However, I doubt that the fear of a minority government that has marked many past Tasmanian elections will be much of a factor in 2010. The voters may well sense that both the current major party leaders could and would work effectively with the Greens in minority government if they had to, and that the Greens may have learnt their lessons of the past should such a situation arise again. If this is indeed the case then the swing of undecided voters to whoever is considered a chance of forming majority government may not be as pronounced as in previous elections – and that could well be Labor’s biggest problem.