DAVID FARRAR AGAINST THE DOMINION POST
Case Number: 3343
Council Meeting: OCTOBER 2022
Decision: Not Upheld
Publication: The Dominion Post
Accuracy, Fairness and Balance
Headlines and Captions
Headlines and Captions
Polls and Surveys
1. On 12 September 2022, during the local body election campaign, the Dominion Post published online and on a page three of the newspaper, a story headlined Tory Whanau clear leader in straw poll for Wellington mayoralty. David Farrar has complained the subject of the story was not a poll, but rather an account of a self-selected online survey run by the Dominion Post and Stuff, and that this was not clear from the headline and the reporting. He believes Media Council Principle (1) Accuracy, Fairness and Balance and Principle (6) Headlines and captions were breached. The complaints are not upheld.
2. Stuff asked all three leading candidates for the Wellington mayoralty for a 30 second ‘elevator pitch’ on why they should be mayor. These were posted online on 8 September and readers invited to vote online. In the first sentences of its subsequent story (the subject of this complaint) about the results of the survey, the Dominion Post reported that by Sunday 12 September:
A straw poll of Wellington mayoral candidates has first time runner Tory Whanau with a huge majority over her two main rivals,
Paul Eagle and Andy Foster.
The Stuff poll, which has been open since Thursday, had 5251 votes cast by Sunday morning. It is intended to give an
indication of which candidates are viable shortly before voting documents get sent out.
It only allows one vote per computer but is open to anyone regardless of where they live; it does not account for the single
transferable vote system used in Wellington which means second or last choices for mayor play a big role.
It is just one snapshot in time, rather than a scientific survey. No other known polls have yet been released.”
The article goes on to quote the reservations of several candidates about the reliability of the survey.
3. Mr Farrar complains that the straw poll was not a poll but an unrepresentative online survey. He has numerous issues with the survey and the reporting of the results. Respondents were self rather than randomly selected and he says they were able to vote multiple times. Responses could be dominated by Stuff readers, people active online, repeat respondents or those ineligible to vote in Wellington. A large sample size is not meaningful unless a survey is randomly selected. The survey was not consistent with the New Zealand Political Polling Code (polling industry code of practice). The article was misleading in the way it reported an unrepresentative, self-selected straw survey as a poll. The article’s mentions of the unscientific nature of the survey were insufficient to counter the overall impression of the article, that the results of the self-selected survey were an important indicator that Ms Whanau was in the lead. The print version of the headline was more misleading than the online one, because it omitted the word “straw”. The article was too long and too prominently placed in the print version and could have influenced voters. Mr Farrar also says the statement that only one vote per computer was allowed is factually wrong and that it is “trivial” for internet users to get past restrictions.
4. He believes the article breached Principle (1) as it was inaccurate and misleading as to the significance of the survey and public
support for mayoral candidates. It breached Principle (6) as the headline lacked any reference to the unscientific nature of the survey by
labelling it as a poll. Mr Farrar sought corrective action, i.e. the Dominion Post should commission a scientific poll (Mr Farrar states he
was not seeking the business); or if that was not feasible a follow up story apologising for the headline and the implications of the
original story which claimed to indicate which candidates were viable.
5. The Editor of the Dominion Post, Anna Fifield, responds that she considered carefully before reporting on the results of the Stuff online straw poll of self-selected respondents. Her reasons for proceeding with publication included: the very large size of the poll which collected 5251 responses over a few days; no other polls had been conducted at the time; and the fact that shortcomings and caveats were fully disclosed by writing these into the story, including reservations about the survey expressed by Mr Foster and Mr Eagle. The Editor says the article was not inaccurate or misleading to readers. The online headline and introduction clearly stated that it was a “straw poll”; the print headline used just “poll”, but the introduction made it clear it was a straw poll. The editor says the article was completely transparent with readers about the unscientific nature of the survey, from the start of the story and throughout. The story was one small part of the exhaustive local election coverage by the Dominion Post and Stuff, and in addition the paper hosted a candidate debate.
6. In response to Mr Farrar’s point about ‘poll’ versus ‘survey’ the Editor notes that although this may be an important nuance in the polling industry, readers were likely to use the two terms interchangeably. The Research Association of New Zealand’s (RANZ) political polling code guidelines themselves refer to self-selecting polls and self-selecting surveys, suggesting that both terms can be used. The RANZ guidelines are just that, and news organisations are not obliged to adhere to them. The article was open and transparent about its sources. The editor says people would have been able to see the poll for what it was and treat its results accordingly.
7. In relation to Mr Farrar’s complaints about aspects of the news value, length and placement of the story, the Editor says these are matters for the editor’s and journalists’ news judgement. The article was not a front-page story, but placed on page three of the print edition, among other local government stories on pages 2 and 3. Ms Fifield says Stuff ran numerous stories about the mayoralty over the campaign period of which this article was a small part. Stuff took a creative and transparent approach to its extensive election coverage, which was intended to engage readers and generate interest in democracy.
8. Stuff put online a self-selected survey about three leading candidates for the Wellington mayoralty, and then made a news judgment to publish the results. This was part of their comprehensive coverage of the recent local body elections. The complainant contests Stuff’s presentation of this survey as a poll and believes it may have misled voters. The editor on the other hand, believes the article was transparent and did not mislead readers on its sources. The Council appreciates Mr Farrar ‘s genuine concern, as a pollster, about the use of informal online polling. There is a real difference between a professionally conducted poll, and a self-selected survey. However, it is a fact of life that quick polls and opinion sampling are now a feature of the online environment. They are used by media to liven up coverage and provide talking points.
9. The Council agrees with the editor that the distinction between ‘poll’ and ‘survey’ is one likely to be made by the polling profession, but not by the public. In the Council’s view, many people will now be sufficiently familiar with this kind of quick, self-selected survey to be able to make a judgement about value. People will generally be able to distinguish between the more authoritative results of professionally run random polling and the quick, ‘vox pop’ surveys or online polls commonly used by the media. In this case, the article was clear throughout about the unscientific nature of the survey.
10. The complainant has also singled out the statement that only “one vote per computer” would be allowed, as being factually incorrect. The Council is not qualified to make an expert judgement on this technical issue, nor how easy it would be for general users to evade the control of one vote per computer. We think this arguable statement in the article would have been better avoided, but the Council does not consider it significant enough to support the finding of a breach. The Council considers the article does not breach Principle (1) Accuracy, Fairness and Balance.
11.The complainant also says the print headline which read: Tory Whanau clear leader in poll for Wellington mayoralty” was additionally misleading. It was different from the online headline because it did not contain the descriptor “straw” before the word “poll”. The Council considers the print headline complied with Principle (6) on headlines, in that it conveyed a key element of the story; and we do not consider it to be substantially misleading. The article opened with and had references throughout to the unscientific nature of the survey.
Decision: The complaint on Principles (1) and (6) is not upheld.
Council members considering the complaint were Raynor Asher (Chair), Ben France-Hudson, Jo Cribb, Judi Jones, Marie Shroff, Reina Vaai, Alison Thom, Richard Pamatatau, Hank Schouten, Rosemary Barraclough, Scott Inglis, and Jonathan Mackenzie.