Two pollsters have lowered their colours in recent days with poorly framed questions on the carbon tax. Last week, Roy Morgan conducted a phone poll which, among other many things, asked of respondents: “Australia is only responsible for about 1% of the world’s total carbon dioxide emissions. Are you aware of this or not?” This is essentially a political talking point framed as a question: understandable from a political party engaged in the desperate tactic of push-polling, but quite incomprehensible from a market research firm. Beyond that though, I don’t think the Morgan poll did much harm. As Peter Brent of Mumble rightly points out, results for the aforementioned question and those asked thereafter must be regarded as unreliable, but the nature of these questions was such that this is no great loss. The question on voting intention was presumably asked before the ones on climate change, and the first three climate change questions were usefully framed and produced results consistent with other polling. If the voting intention results from the poll do not seem plausible and they don’t this must be put down to sampling error and perhaps some systemic bias suffered by Morgan phone polls, although this hasn’t been evident in the past.
More troubling is today’s Galaxy poll, which targeted a small sample of 500 respondents on behalf of the Daily Telegraph. For the most part, its results are of genuine concern for the government. Only 28 per cent answered in favour of a carbon tax against 58 per cent opposed, corroborating the 30 per cent and 60 per cent from Newspoll when it last asked the question on April 29-May 1. Even worse for the government, fully 73 per cent said the tax would leave them worse off against only 7 per cent who opted for better off. Less remarkably, the poll found 20 per cent believe the tax would have a major impact on the environment, 46 per cent a minor impact and 29 per cent no impact.
The problem lies with the following: Does the PM have a mandate to introduce the tax or should she call an early election? This gives respondents no outlet for the obvious third alternative: that while the Prime Minister does not have a mandate for a carbon tax (and given her position during the election campaign, it could hardly be argued otherwise), the government should nonetheless govern as it sees fit and face the music at the end of its term. This happens to have been the default position for poll respondents since 1975, when 70 per cent opposed the blocking of supply despite the enormous unpopularity of the Whitlam government. There can be little doubt that many who wished to express a view that no mandate existed opted for the only available alternative. That a substantial proportion would have preferred the option Galaxy did not provide is illustrated by last week’s Essential Research poll, which asked the question the way it should be asked: Do you think the Government should call an early election over the carbon tax? Whereas Galaxy had 24 per cent choosing the has mandate option and 64 per cent should call early election, Essential had it at 42 per cent each way.
Considerably exacerbating the problem is that the poll was conducted for Australia’s most brazenly partisan metropolitan newspaper, the Daily Telegraph (UPDATE: The Herald-Sun is also selling it as its own work). And true to form, the Telegraph has today used the opportunity to run an editorial headlined voters demand a carbon tax ballot, in which it argues that an election now is very necessary. This is not the first time Galaxy has risked taking a hit from its association with the paper. During the election campaign the Telegraph tasked it with assembling an audience for a people’s forum involving the two leaders at Rooty Hill RSL in western Sydney, and there was little doubt that the room was pro-Abbott (sweet revenge, the Coalition might well think, for many a pro-Labor worm in leaders debates of elections past). On that occasion the culprit appeared to be the targeting of undecided voters concentrated in a part of the world which had been notably problematic for Labor. If the Telegraph itself had a hand in either the Rooty Hill methodology or the wording of the early election question, Galaxy might want to be a bit more firm with it in future, lest it jeopardise the reputation its polling record suggests it deserves.
For all that, there can be no denying that the carbon tax debate has so far been all pain and no gain for the government. The initial announcement in March saw the polling trend blow out from 51-49 in the Coalition’s favour to 54-46, and attitudinal questions have generally found the tax to be getting less popular rather than more. However, two caveats need to be added. Firstly, the phone polls conducted by Newspoll, Galaxy and Morgan have painted a more negative picture than the online methodology of Essential Research, which last week found 38 per cent support and 48 per cent opposition. Lest it be inferred that this represents a Labor bias from Essential, their two-party average since the carbon tax was announced has actually been worse for Labor than Newspoll’s (aided in no small measure by Newspoll’s rogue 51-49 to Labor result of March 18-20). Phone polling tends to get taken more seriously purely due to its long track record, but it could be that this disparity illustrates concerns that are increasingly raised about it: that in limiting its catchment to those with landline phones, it misses a more technologically fashionable and, perhaps, environmentally minded section of the electoral market.
Secondly, Essential Research found the idea of a carbon tax to be a great deal more popular when it put it to voters that the money paid by big polluting industries would be used to compensate low and middle income earners and small businesses for increased prices. When the question was asked thus in mid-April, fully 54 per cent expressed support against 30 per cent opposed. However, I’m not inclined to think this offers the government more than limited comfort, unless and until it proves capable of framing the debate in such fortuitous terms.