EMRS May 2011: Liberal 38 Labor 19 Green 17 Ind 4 Undecided 22
A possible interpretation: Liberal 49 Labor 29 Green 18 Ind 4(?)
If a state election was held right now the Liberal Party would almost certainly win outright. The latest EMRS poll confirms the finding of the February poll (No Giddings Bounce - is Labor doomed?) that Labor is now as unpopular as it has been since the party last lost office back in 1992. The EMRS headline figures, with Labor down one, the Liberals up two, the Greens down three and “Independents” up two (more on that later) show little change from the February poll, and for the major parties that change is not statistically significant. But the February poll in itself was horrendous enough for the Government, and now this one is slightly worse. The (misleadingly low) EMRS baseline rate of 19 is the lowest the Government has yet recorded, and the Liberal lead of 19 points (20 once voters “leaning” to a party are included to give a result comparable to that for other Australian pollsters) is the largest yet. Also, for the first time the Liberal vote exceeds the Labor and Green votes combined. While I am not a huge fan of the “preferred premier” indicator it is notable that Lara Giddings’ modest honeymoon on that score lasted just one poll, and her rating is now a point below David Bartlett’s final score. The change of leadership has not helped Labor’s standing, but it would take a remarkable leader to make this Government poll well.
As usual where two consecutive EMRS polls are similar I have batched the figures from the seat-by-seat samples for the two polls, weighted them slightly in favour of the most recent, and scaled the Green vote to compensate for the established fact that “undecided” voters in EMRS polls don’t vote Green in significant numbers. When this is done the combined sample points clearly to an easy Liberal victory if an election was held now. All electorates would split 3-1-1 in the Liberals’ favour except for Denison which would be 2-2-1 (or maybe 2-1-1-1 given a suitable independent) and Braddon which would probably be 3-2-0, for a likely total of 14 Liberal, 7 Labor and 4 Green, in an exact mirror of the standings of the parties at the 2002 and 2006 elections. The combined sample would also give the Liberals a majority under the 35-seat system (something like 19-10-6).
Apart from the massive gap in support between the two major parties, the most unusual feature of the latest poll is a statewide surge in claimed support for “independents”, which has suddenly doubled since the February poll, rising in the mini-samples for all five electorates. At the same time, claimed support for the Greens has fallen by three points. Given the sharpness of this increase and the fact that claimed support for independents was generally low before this (except in Denison), this looks like a contaminating effect from the Legislative Council elections held a few weeks before the poll was taken. At those elections, voters who vote Green because they do not like the major parties seemed to be especially easily peeled off by independents, even anti-Green ones.
It is possible, however, that this vote is genuinely coming from the Greens in terms of Lower House intention too, and that disenchantment with the Greens at the by-default end of their base has suddenly reached a tipping point, and that large numbers of ex-Green voters are now seeking “fourth party” alternatives. A third possibility is that it is coming from all parties and the Greens are losing votes to the Liberals, but I think this is unlikely, especially as Nick McKim’s preferred premier score has gone up slightly (most likely at the expense of Premier Giddings). We will have to wait three months to find out if the indie surge is real – and until the next election to see if any real alternatives exist.
The hope for the Government remains, as noted last time, that votes supposedly lost by it to the undecided column are not being so firmly locked in by the Liberals. The latest poll sees very modest progress towards the locking-in of defecting Labor voters, with the Liberal “support+leaning” score now marginally above their state election outcome.
But Premier Giddings should not take too much comfort from that. The Premier makes a rather big deal of the headline undecided figure being 22 per cent and this being much higher than the 10 points a year ago, and claims this shows that “many Tasmanians have reserved judgement on all three major parties, a sentiment that was also reflected in the recent Upper House elections.” In fact, those familiar with the long and sloppy story of the EMRS undecided rate will know that it is nearly always high, and is magnified by the company’s use of a headline figure not comparable to that of other polling companies (the comparable rate in this case is 15 not 22). Furthermore, other companies operating in the state in the past have not recorded such high undecided rates, even when EMRS did. The reason that the May 2010 undecided rate was so abnormally low was that that poll was taken closely following a state election with a polarising post-poll period.
The Legislative Council elections did indeed show voters reserving judgement on all parties, but a degree of this is normal for the LegCo. Furthermore in the one seat where the Liberals ran a formal campaign their bad result was down to poor candidate choice and high-risk campaign tactics.
A brief roundup of some other media:
ninemsn appear to be taking their readings from the figures with all undecideds redistributed, a bad practice since we know that this overestimates the Green vote. Nonetheless their basic conclusions are sound. The ABC raises the possibility of links in the drop in the Green vote to a policy change on public service redundancies. Sue Neales follows Giddings in reading the undecided figure as a sign of discontent with all parties, when in fact it alone shows no more discontent than normal over the past several years, and the only thing new there is the high indie vote (which may just be a LegCo artefact). Furthermore while locked-in Liberal support is indeed “on a par” with that polled just after the last election, the real level of Liberal support is bound to be higher as the post-election poll had such a low “undecided” rate.
Of course, it is a long time to the next election; polls taken this far out do not have much predictive value beyond what we already know. That is firstly that elderly governments find that, as the saying goes, “friends may come and go but enemies accumulate”. And secondly, that minority government with the Greens in tow is a difficult thing to sell to mainstream Tasmanian voters. It doesn’t seem to matter what form it takes.
Giddings is putting out the message that Labor needs to take hard economic decisions now; by the time of next election, perhaps the benefits of these decisions will be apparent and the short-term pain forgotten. The problem is that the message, whether correct or not, is probably still unsellable, because Labor has been in charge for almost thirteen years and voters won’t believe it couldn’t see it coming. Every piece of positive economic news released by the government raises the question of why on earth there’s so much trouble in the first place. Even the “hard decisions” of the Michael Field years, taken to fix the damage caused by the other party, took many years to filter back in the form of actual electoral goodwill. It seems unlikely Labor will ever get much credit for cleaning up a mess that is its own.
Show Comments
Comments (30)