Well-meaning online activist group GetUp has been criticised by Crikey and others in the past for being either too close to the ALP, too close to the parliamentary system, or a front for its employees’ burning career ambitions, but a survey released late last year by the group verges on deliberate deception.

As debate raged over the Government’s carbon pollution reduction scheme in October, GetUp decided to query its members in inner city federal electorates on whether climate change had the potential to swing votes. The “2 minute” SurveyMonkey poll asked three short questions as you can see here.

Number 3 was the standout:

If the Federal Labor Government weakens its position or fails to deliver strong action on climate change, how likely would you be to vote for a party with a stronger climate change policy at the next federal election?

Fair enough you might think — the ALP has expressed concerns about the rising levels of support for the Greens among the kind of inner-urban latte sippers that constitute GetUp’s core membership. Presumably acting on a hunch, the group decided to poll three electorates traditionally held by the ALP with a rising Greens vote: Tanya Plibersek’s seat of Sydney, Anthony Albanese’s Grayndler, and Lindsay Tanner’s seat of Melbourne.

But some of the conclusions drawn in the press release issued by union spin doctors EMC were startling.

Take this par:

The poll of members in three important Federal Labor seats, including Melbourne, Sydney and Grayndler, found that a majority of the party’s traditional supporters were prepared to abandon the Rudd Government over the issue of climate change.

But the survey found nothing of the sort.

It was not the party’s “traditional supporters” but a small cross-section of GetUp members who opted to complete the survey (and also identified as “traditional supporters” of the ALP) that in turn claimed they were likely to shift their votes. That’s three or four levels of abstraction from the real mood among Labor’s rusted-on base, but you’d have trouble deciphering that from a casual read of the release. Ironically, it is precisely the influx of the type of gentrified elites disproportionately represented by GetUp’s membership that is displacing traditional ALP voters in the inner city — the climate change connection, if any, is purely coincidental.

And then this par:

More than 80 per cent of Labor supporters surveyed (more than 600 per electorate) in the poll indicated they would rather pledge their support to another party with a strong climate change policy rather than campaign for weak or failed Rudd Government carbon reduction measures.

But the survey didn’t ask whether respondents would “campaign” for either the Rudd Government or anyone else. This seems to have been inferred via GetUp’s self-assessment of its members as “campaigners”. In any case, it’s questionable whether the majority of GetUp “members” are politically active — many are are on GetUp’s email list and not much else.

Interestingly, the media seem to have smelled a rat. Such a shoddy survey should never have been published and the Sydney and Grayndler surveys received zero response. But the next day, on 1 November, the Melbourne angle was given oxygen by Michelle Grattan who produced this Age write-up giving credence to the hatchet job.

Grattan was working off a separate polling report emailed to her by GetUp and not immediately accessible to the public. But a section of the same report that Grattan used is explicit in its assessment of the survey’s flaws:

The data for this report was gathered from an online survey conducted by GetUp.org.au. All respondents have elected to receive emails from GetUp; they are not a representative section of the broader community and have not been weighted to reflect Australian Bureau of Statistic (ABS) data.

Basically, on GetUp’s own admission, the survey is meaningless.

According to resident Crikey psephologist William Bowe, most voluntary survey respondents with an already-existing issues bias will, when presented with a choice between different positions, generally trend towards the most severe option. The survey’s non-randomised nature and the fact it was completed exclusively by a small core of GetUp members renders it statistically useless. Bowe says the poll appeared to be an attempt to generate an easy headline among the heat and light of the ETS debate.

But it also fits the group’s pattern of political engagement — appropriate an issue of genuine concern to the Australian public in the interests of maintaining a monopoly on progressive discussion and debate. Rather than endlessly surveying its members, GetUp would be better off deferring to the grass roots social movements that have been fighting these battles for decades.