Curia exited RANZ because Family First are unbalanced

There has been some speculation on exactly why David Farrar’s company Curia abruptly exited the Research Association of New Zealand (RANZ). It appears to have been triggered by a complaint about the fairness of the wording of a question in a April poll. The finding from that complaint has been released from RANZ, and I have published a copy of it in full below.

This was a really bad poll, and this question a particularly egregious breach of fairness. It was paid for by Family First. To my eye, and evidently to the professionals who reviewed it, the question seems designed to elicit a specific completely unfair and unbalanced headline.

The question was checked by the RANZ Professional Conduct Officer (PCO) and Professional Standards Group (PSG), who recommended that Curia be suspended from the RANZ. The board then appointed a three member independent subcommittee to evaluate both the complaint, their view on it, and if the procedures for handling complaints needed needed changes.

The question was :-

The UK health service (the NHS) has stopped the use of puberty blockers, which begin the gender transition process, for children under 16 as it deemed they are too young to consent. Do you support or oppose a similar ban in New Zealand on the use of puberty blockers for young people 16 or younger?

The complaint about the question was

14. The complaint is that this is “significantly biased…… The question was whether there should be a ban on the use of puberty blockers in New Zealand. By prefacing the question with the example of one country that has banned these, rather than an example of a country that has not banned them, the answers to this question were inevitably skewed.”

David Farrar claimed previous practice of similar questions, which was rightly rejected. The complaint was just on this question against the standards set by the RANZ. (point 21). Complaints against similarly framed questions will be guided by the precedence on this one. I’d also point out that there have been numerous informal complaints about framing against Curia’s questions in the past visible on this site and others.

The other two defences by David Farrar and the reason their rejection were:-

22. Mr. Farrar stated that reference to the UK legislation provided ‘context’. We rejected this argument because it is good research practice to ensure that any contextual explanations for questions like this are completely neutral. For example, it would be acceptable:
    a. To provide a factual explanation about what puberty blockers are as context,
    and/or
    b. To state that one country has supported puberty blockers while another has not as context, if it was true. It cannot be context if it is one-sided.
23. Mr. Farrar stated that it is normal research practice to test ‘framing’ of a question, which we agree with. It is common research practice to detect whether one way of framing an issue gains more public support than other ways of framing. To do this, best practice is to ask the question at least twice – once without framing and then with a frame. The Curia survey did not do this. It asked the question once only, framed. It is therefore impossible to tell whether or not the reference to the UK affected respondents’ answers. We note that the example that Mr. Farrar supplied of a question used for framing did provide multiple wordings. We conclude from this, that this survey did not test framing.

Point 24 gets to the meat of the complaint. Curia’s client led and insisted on the framing of a poll question designed to elicit a specific response.

24. Mr. Farrar also states that although he advises clients on the relative merits of questions, ultimately he will accept the question if the client insists. This is a particular risk where the client is lobbying for a particular outcome. We believe that this situation requires particular attention to question design as the results are more likely to be published and scrutinised for robustness. By signing up to the Code of Practice, there is an obligation on RANZ members not to simply ask a question in a particular way if the client insists, even if they have made efforts to dissuade the client from asking in in a particular way, if it could be considered leading or bias the response. The purpose of the code is to ensure best practice is followed. We do however note that Curia Market Research publishes the question asked, which is an explicit requirement of the New Zealand Political Polling Code but not the Code of Practice under which this complaint is being investigated, which does allow the public to decide if the question is biased or leading.

This was a poll question commissioned by Family First and summarised with similar unbalanced questions in a press release on the 25th of April 2024 (PDF), and based on the poll results from Curia (PDF). I have put local copies on this site, so you don’t have to click to the Family First site.

How Family First presented the information to the public and media quite clearly shows their intent. It also illustrates why they must have insisted on the poor methodology of the way that the question was asked. Family First presented the results of their commissioned poll questions in a egregiously unbalanced fashion, in my view designed to obfuscate the questions asked of respondents.

The questions on the summary below show abbreviated questions. Family First only highlighted half of the questions, presumably to direct attention to what they wanted to highlight. The press release was, in my view, obviously carefully made to provide clickbait designed for the type of lazy media who are only interested in the headline or simple slogan for TV or radio.

The remainder of the press release did publish the actual questions. But even here the questions were not highlighted in anyway and were looked exactly the same as cherry-picked, out-of-context and completely unlinked quotes from other sources from the NHS in the UK, and the FDA in the US. This presentation technique was clearly designed to obscure the actual questions asked of respondents. A classic technique to limit questions about methodology.

But I expect nothing less of Family First, who are conservative lobby group who occasionally try to masquerade as a serious political party. There is an extensive wikipedia page on their activities since 2006.

Its 2006 stated objectives were to “seek to influence public policy affecting the rights and protection of families and promote a culture that values the family”

I usually refer to Family First as being the Family Fist, because the aim of their proclaimed policies appears to be the suppression of any social progress since the 1950s, i.e. the ‘good times’ for the Family Fist. When women were expected to be confined to the home, contraception was frowned on and largely unavailable, the beating of women and children in families in the pursuit of patriarchal authority largely accepted, and any ‘deviants’ from a sexual and societal norm were likely to be prosecuted.

While a child and adolescent, I saw the trauma and agony of the Family Fist’s preferred social norms as young pregnant women in their teens from provincial areas would stay in our Auckland house for periods of time before fostering a child. It made an strong impact that ‘righteous’ bigots like the members of Family Fist wouldn’t celebrate the birth of child, but instead preferred to torture teenage mothers for their human behaviour, while forcing them to give up a child.

The highlight for me in the Family First wikipedia page was reading the legal battles over retaining their charitable status, which went on from 2013 until it got to the Supreme Court in 2022.

On 28 June 2022 the Supreme Court ruled that Family First did not qualify for charitable status, concluding that its research lacked the balance needed to be educational. The Supreme Court also held that Family First’s conservative family advocacy was not charitable on the grounds that it lacked “fairness, balance, and respect.”

Which really just goes to show Family First has form in their systematic lack of balance and fairness.

The Women’s Rights Party shows similar traits and seems to be covering the same tired tracks of selective victimisation almost word for word. But I will save that for a different post.

Feel free to offer well-written guest posts and commentary about any polling that probably doesn’t meet the RANZ code of practice. It looks like we may finally have a industry body that doesn’t seem to be just acting like a rubber stamp for poor behaviour.

I’d like to specially mention Ben Gray, who laid the complaint for their success at helping to curb the excesses of the polling industry.