Two months on from the May budget, opinion polls are providing no indication that the public backlash against the Abbott government is wearing off. The real winner is Clive Palmer, whose Palmer United Party is enjoying unprecedented levels of support -- depending on which measures you believe.
All three poll results published this week show Labor well ahead on two-party preferred, by margins ranging from 53-47 in Essential Research
to 56.5-43.5 in Roy Morgan
According to the BludgerTrack poll aggregate, updated weekly at my blog The Poll Bludger
, Labor's two-party vote has increased by 0.3% on last week to put its lead at 53.5-47.5 -- an exact reversal of the result at the September election.
The other notable feature of BludgerTrack this week is the result for Palmer United, which has tested the bounds of the error margin with an increase from 6.3% to new high of 7.5%.
Here we encounter a complication, in that the model does not account for the very different picture of Palmer United support that is apparently emerging from Newspoll. Newspoll's published tables
do not provide a dedicated result for Palmer United, but according to The Australian's reporting of the poll
, the 16% "others" vote includes a Palmer United component amounting to "about 3.5%".
However, the lack of "official" status for this figure means it does not enter into the calculations for BludgerTrack, which instead draws upon pollsters who are rating Palmer United roughly twice as high. This variability points to the challenge pollsters face in calibrating support for minor parties that feed off capricious anti-politics sentiment.
In the past year, television advertising campaigns have inflated Palmer United support from practically nothing to 5.5% nationally in the last two weeks of the federal campaign, and doubled it again when applied with even greater intensity before the Western Australian Senate election in April.
With no such campaign currently running, a distinction might be said to have opened between the party's latent and actual support, causing its poll ratings to be highly sensitive to the way the voting intention question is asked.
Respondents to automated phone polls find themselves presented with a single list of possible responses, in which Palmer United features as one among a fairly small range of options. This serves to promote the salience of the party in the minds of undecided respondents, creating a potential for its support to be overstated. Newspoll negates this effect, perhaps excessively, with a first question that offers a choice between the Coalition parties, Labor, the Greens and "others". Those who choose the latter option are given a follow-up question in which they can choose from a list of smaller parties or an unspecified independent.
Faced with this secondary question, those looking to express a general sense of alienation from the established parties seem to prefer the independent option to Palmer United, notwithstanding that there may not be a viable candidate in their electorate to vote for.
The range of results which have emerged from the three pollsters providing regular numbers for Palmer United is illustrated in the chart below, together with the trend result from BludgerTrack:
The three pollsters shown account for a range of methodologies -- automated phone polling in the case of ReachTEL, which has tracked above the overall trend; online polling for Essential Research, which has usually been below it; and a combination of face-to-face and SMS polling from Morgan, which has followed the middle course.
Only in mid-January did the trend reach the zone currently indicated by Newspoll, since which time the party's stocks appear to have been boosted by the WA election in April, the adverse response to the budget, and most recently the publicity which has attended the party’s early endeavours in the Senate.
This serves to reinforce a lesson that was already evident from the impact of its television advertising -- that the party’s electoral potential is considerable, but heavily dependent on it maintaining the oxygen of publicity.