BSA Decisions Ngā Whakatau a te Mana Whanonga Kaipāho

Summary  

[This summary does not form part of the decision.]

The Authority has declined to determine a complaint that the presentation of a 1 News Kantar Public poll concerning support for political parties ahead of the 2023 general election was misleading. The Authority has previously determined that excluding undecided voters from poll figures was not inaccurate, and the issue of poll figures adding to 100% did not require our determination. On this basis the Authority considered it appropriate to decline to determine the complaint.

Declined to determine (section 11(b) of the Broadcasting Act 1989, in all the circumstances): Accuracy

The broadcast

[1]  An item on 1 News, broadcast on 25 May 2023, discussed the results of the latest 1 News Kantar Public Poll of New Zealanders concerning political parties and preferred Prime Minister. The item included graphics with the following numbers:

National 37%

Labour 35%

Act 11%

Green 7%

NZ First 3%

Te Pāti Māori 2%

The Opportunities Party 1%

DemocracyNZ 1%

Aotearoa Legalise Cannabis Party 1%

New Conservative 1%

[2]  The item then included a visual of how these numbers would translate to seats in Parliament, with the following commentary:

So let's take a look at how this translates to seats in Parliament and how a government can be formed on these numbers. National would bring in 47 seats here in blue. Add in 15 from ACT and you've got 62 – enough for the two parties to form a government without needing Te Pāti Māori. On the other side, Labour would bring in 46 MPs here in red, the Greens dropping away with only nine MPs and even with Te Pāti Māori, this group would only get 58 seats – not enough to govern.

[3]  The reporter noted later in the item: ‘In this poll, 12% still don't know how they'll vote or refused to say.’

The complaint

[4]  Paul Eady complained that the broadcast breached the accuracy standard of the Code of Broadcasting Standards in New Zealand for the following key reasons:

• ‘The method of presentation of the poll results gave a confusing and potentially very misleading impression of the number of votes that would be cast for the relevant political parties, and the potential compositions of government if an election were to be held at the time the poll was taken.’
• ‘Firstly, the percentages allocated to each party do not total to 100%’, (only 96%).
• ‘Secondly, the 12% of respondents were "Don't know or refused" were noted but that data was not represented in the allocation of seats,’ or in the polling tables. This was a statistically significant amount, higher than either the Green Party or Act support figures.
• ‘Removing this significant amount of respondents from the discussion and then declaring a journalistic conclusion that certain parties could form a government with a 2 seat majority creates a misleading and statistically incorrect picture of the mood of the electorate’ because 12% of voters represents approximately 14 seats in Parliament, which was clearly enough to sway the outcome of the election.
• Including the ‘don’t know’ and ‘refused’ poll answers in representing the seats in the house would have allowed viewers to consider the potential impact of uncommitted voters and ‘hidden votes’. A separate figure for ‘don’t vote’ or ‘won’t vote’ would also have informed viewers of the level of engagement of the public.
• Additional information on polling processes and methodologies and where that was available online (referred to by TVNZ in its response to the complaint) should have been included in the programme.
• The complainant disagreed the presentation of the results was a matter of their own preference (responding to TVNZ): ‘I have no particular preference for how the data and conclusions are presented, except that the presentation is clear and concise about what IS presented and what relevant aspects ARE NOT presented so that viewers can critically assess the validity of any conclusions drawn by the broadcaster.’

The broadcaster’s response

[5]  TVNZ did not agree the accuracy standard was breached. Its key points in response to the complaint were:

• The numbers in the initial graphic total 99% (and it assumed decimal figures made up the other 1%).
• It is a reasonable approach to exclude undecided voters, non-voters and those who refuse to answer, from data on party support, including when extrapolating poll numbers into seats in parliament.
• The results were accurately reported straight from the Kantar Public poll report (both the figures, and the representation of seats in parliament). The poll report indicates ‘don’t know’ respondents are excluded from party support figures (it was not the broadcaster who excluded the 12% from these results in the initial polling figures). The poll methodology and reporting are compliant with the Research Associations New Zealand Political Polling Code.¹ TVNZ provided a link to the full poll report online.²
• The reporter confirmed in the item that ‘12% of respondents still don’t know how they’ll vote, or refused to say,’ which was sufficient for viewers to understand this outcome of the poll.
• While the complainant may prefer the polls were presented in a different way, this amounts to a personal preference which cannot be addressed under broadcasting standards.

Outcome: Decline to determine

[6]  Section 11(b) of the Broadcasting Act 1989 authorises the Authority to decline to determine a complaint if it considers that, in all the circumstances of the complaint, it should not be determined by the Authority.³

[7]  First, we are satisfied the poll figures displayed onscreen add to 99% and the remaining 1% can reasonably be explained by rounding/decimals – so there is no issue here requiring our determination.

[8]  With respect to the remaining points of complaint, the decisions of the Authority issued over time provide guidance to broadcasters and complainants about what is acceptable under broadcasting standards. The Authority has previously found that the number of undecided voters being excluded from poll results did not amount to a material inaccuracy.⁴ This report clearly did include this figure, and in any case the manner of presentation of poll results – including where in the report to include that figure – is a matter of editorial discretion for the broadcaster. On this basis we see no reason to reconsider the Authority’s previous approach.

[9]  Accordingly we consider it appropriate to exercise the Authority’s s 11(b) discretion in this instance.

For the above reasons the Authority declines to determine the complaint.
Signed for and on behalf of the Authority

Susie Staley
Chair
7 November 2023  

Appendix

The correspondence listed below was received and considered by the Authority when it determined this complaint:

1  Paul Eady’s formal complaint to TVNZ - 27 May 2023

2  TVNZ’s decision on the complaint - 23 June 2023

3  Eady’s referral to the Authority - 24 July 2023

4  TVNZ’s comments on referral – 20 September 2023

5  Eady’s further comments - 20 September 2023

6  TVNZ’s further comments – 21 September 2023

7  Eady’s further comments – 24 September 2023

8  TVNZ confirming no further comments – 29 September 2023

¹ Citing “Frequently asked questions” Kantar Public <kantarpublic.com>
² Felix Desmarais “Poll: National, ACT have numbers to govern, Luxon lags in preferred PM” 1 News (online ed, 25 May 2023)
³ Broadcasting Standards Authority | Te Mana Whanonga Kaipāho “Guidance: BSA power to decline to determine a complaint” <bsa.govt.nz>
⁴ See for example Wright and Discovery NZ Ltd, Decision No. 2020-143 at [13]; and Seymour and Television New Zealand Ltd, Decision No 2012-082 at [12]