DON’T expect this article to tell you who will win — not just yet. What I do want to do here is explore some scenarios that have been thrown up in the unrestrained speculation-feast that typically follows the release of yet another EMRS poll.
Firstly, however, I want to make some general comments about the state of play. At this stage of the campaign Labor has its work cut out to save its majority. It’s far from impossible, but they’re struggling. This much is obvious enough, but these are a few of the factors that may be influencing this state of proceedings:
• Denison is pretty much a goner. Labor shows few signs of being able to win three seats with a team headed by an unremarkable first-termer (Sturges) and a member elected on a recount (Bartlett), who didn’t even have a quota between them in 2002 even after Jim Bacon’s surplus. See my comments below on what might happen to the seat they are almost certain to lose.
• Polls recently show the Greens level with, to up to five points ahead of, their record 2002 result. While Labor is still way ahead of the Liberals in the polls, leads of 10% or so are not enough if the Greens are winning seats in every or nearly electorate. Labor would be in a stronger position if those deserting it were going overwhelmingly to the Liberals rather than the Greens. Yet the four most recent polls (two by EMRS, one by TasPoll and one by Morgan) all suggest otherwise. Given that the Liberals are supposed to be on the way back from one of the most abysmal results in Australian political history, this doesn’t say much for the condition of the Tasmanian Liberal Party (I could explain why it’s in such poor condition, but I would be here all week). The irony is that Labor may lose its majority because the Liberal Opposition is too incompetent to take votes off it instead of those votes going to the Greens.
• Labor continues to suffer badly on that unfashionable matter journalists seldom report on (bar the odd horror-case splash) but that has a far greater impact on actual votes than Salamanca set issues or obscure conflicts of interest, namely health. It has unwisely left this issue — which is looming as a threat to state governments around the nation — in the hands of a plodding and uninspired minister who appears to lack the confidence of the sector.
• Paul Lennon has an intractable perception problem on environmental issues. No matter what steps he takes in a greener direction, such as supporting the phasing out of 1080 and participating in the Recherche Bay buyout, it appears not to lure back those who have already decided to switch from Labor to Green.
• With Labor’s ability to win a majority itself a talking point, Labor is losing the battle to convince the voters that it can win outright as simplistic media reports based on shallow opinion poll interpretations convince more and more that a hung parliament is more or less certain. Since a few per cent of voters regard majority government as a critical issue, the government desperately needs a poll that predicts it will win, but these have been in the minority.
It’s still possible, for instance, that the Greens could lose Lyons to the Liberals
Now, to some individual electorates and the EMRS poll results. 21 of the 25 seats in this election are not that likely to change hands between parties (although it’s still possible, for instance, that the Greens could lose Lyons to the Liberals), so let’s assume the Liberals and Greens retain all of their existing seats and look at the four third seats that Labor is at serious risk of losing: Denison and Braddon to either the Liberals or the Greens, Franklin and Lyons to the Liberals. In all likelihood, Labor retains its majority if it loses none or one of these.
EMRS pollster Tony Hocking has argued that it is very difficult to see the Government holding 12 seats on the basis of his poll results. Quite aside from the difficulties of extrapolating anything from 200-vote samples (as discussed in my article at A premature hanging and in Peter Tucker’s article at Lies, damn lies and polls), I disagree with this view even as an extrapolation from EMRS’ published breakdowns.
Lyons, Braddon
EMRS’ projected result for Lyons (Labor 48, Liberal 30, Green 18) would most likely result in Labor retaining three seats. It appears Hocking has allocated this seat to Labor in his prediction of 11-7-5 and two unclear, so we will move on to Braddon.
Hocking apparently thinks Braddon (Labor 41, Liberal 42, Greens 10) would most likely be won by the Greens, and otherwise most likely by the Liberals. Even without taking the unreliability of 200-vote samples into account, this is a bold prediction. The reason is that when a single candidate short of a quota is trying to win the final seat, unusual things can happen. I mentioned this in the previous piece linked above, in which I wrote:
“Another quirk of the system is that a party can win from an “inferior” quota position if enough of its candidates stay in the cut-up long enough. Suppose in a given seat the Greens have 0.70 quotas and the major parties 2.65 each (something that most journalists would call as an expected seat for the Greens), but each major party has one candidate who has reached a quota and two others left over, each of whom have more than 0.70 of a quota. The Green candidate is excluded and one of the major parties wins three seats despite starting 0.05 of a quota “behind”. (Something like this would have stopped Kim Booth from winning in Bass in 1998 even if he had not been passed by the other parties on Tasmania First preferences.)”
There is a useful historical example in Braddon itself. In 1998 the Greens polled 0.488 quotas compared to the Liberals’ 2.535 and Labor’s 2.643. By the time the final Green, Di Hollister, was eliminated, she had 0.575 quotas to the Liberals’ 2.613 and Labor’s 2.689. So, a relatively narrow defeat, then? Actually, no, because Hollister was on 5378 votes to the following totals for the other candidates still in contention at the time: Best (Labor) 8375, Green (Labor) 7421, Bonde (Liberal) 7776, Cains (Liberal) 7310. While the Greens’ total seemed to be only .038 quotas behind the Liberals when Hollister was excluded, Hollister’s total as a candidate was .168 quotas behind the last Liberal candidate. Even had she had 0.7 of a quota and been notionally “ahead” of both major parties, she still would have lost.
Apply this possibility to the projected quota totals from the EMRS poll (.60 quotas for the Greens, 2.46 for Labor, 2.52 for the Liberals) and it is apparent how a Green candidate being .10-.15 of a quota worse off than they appear because of the distribution of votes between the major party candidates, could have a massive impact on the Greens’ ability to win Braddon. More importantly, it is not just the Liberals who would be in the frame if the Greens failed, since Labor making up a mere 1% of the vote on a large bag of Green preferences would be more likely than not. So even if the EMRS 200-vote sample perfectly reflects voter sentiment in Braddon (something which is somewhat less likely than a thylacine attack in the main street of Smithton), Labor has a fair chance of holding the seat, depending on how the votes of the major party candidates are distributed.
Franklin
The next seat to consider is Franklin, but the sample breakdown for this seat has not been reported to my knowledge, except that the Greens are polling 24. We have seen that Labor can realistically win Lyons and Braddon if the EMRS projections are correct (itself enough to disprove Hocking’s claim about 12 seats being very unlikely) — and if Labor could retain Franklin as well it would win the election. It would be useful to know on what figures the view that Labor would probably lose Franklin to the Liberals is based.
Denison
That leaves Denison. The EMRS distribution (35-28-36) is interpreted as a clear 2-1-2 split. While this is correct, does anyone really believe the Greens will poll 36% in Denison? They may, however, poll 30. A challenge they will face if the Liberals are anywhere near them is that they will be very vulnerable to leakage.
As an illustration of the threat leakage poses to the Greens in Denison, four candidates in the 1998 election statewide and seven in the 2002 election statewide polled a quota. Of these, the average leakage out of the party when a candidate’s surplus was distributed was 8.7%. Only two candidates leaked over 11% of their surplus and these were “name candidate” Kathryn Hay in 2002 (16.7% leaked) and significantly Peg Putt in 2002 (16.8% leaked). If the Greens poll close to two quotas, chances are that Putt will have 1.5-1.6 quotas in her own right, and losses due to leakage from her surplus could affect the Greens’ standing relative to the Liberals by as much as 2.5% of the primary vote, although some portion would be recovered as failed Liberal candidates were eliminated.
In a previous piece I dismissed the chances of the Greens winning two in Denison as “extremely unlikely”. Since then the Greens’ poll results have remained historically strong, and even in a 200-vote sample a 36% result suggests a real support level at least in the mid-high 20s. On this basis I now believe two Greens in Denison is a serious chance — but it is not as likely as the EMRS poll has made some think.
Hung Parliament
Finally, the question of what could happen if there is a hung parliament has been widely discussed by others, but some reporters continue to lazily assume a hung parliament must lead to another Labor-Green accord even though both major parties have ruled out such situations. If Labor does win precisely 12 seats, the chances of someone defecting from the struggling Liberals to sit as an Independent must be substantial. Also, if Labor and the Liberals are really so unable to work with the Greens, it would be easy for them to form a brief grand coalition, amend the electoral system to something less favourable for the Greens (again!) and call a new election. From the Greens’ not-that-flash performance in 1998, it is questionable how much Tasmanians really value their Hare-Clark.
That will do for some preliminary comments, and I’ll be back for more after the next opinion-poll-triggered feeding frenzy.
Kevin Bonham has recently negotiated a non-aggression pact with the seal population of Ile des Phoques, and may invite them to form an interim government should the Tweedlethree fail to sort out their squalid little mess post March 18.
