Madam

We would like to thank the Advertising Standards Authority (ASA) for taking note of our recent article on unhealthy food advertising to children on New Zealand television⁽ Reference Vandevijvere, Soupen and Swinburn ^{1 )}. This is especially important since the ASA’s new self-regulatory Children and Young People’s Advertising Code^{( 2 )} took effect on 3 July 2017 for new advertisements and on 2 October 2017 for existing advertisements. We do want to acknowledge that the new Code is potentially a small step forward compared with the existing Code^{( 3 )}, especially with regard to restricting occasional food advertisements in children’s settings (e.g. schools) and including a definition about ‘targeted to children’ (as outlined in the Letter to the Editor from the ASA⁽ Reference Souter ^{4 )}). However, the main problem is the third part of this definition, which is not well specified and operationalized across different media. For television specifically, the 25 % restriction does not make sense and likely won’t have any impact to reduce exposure of children to unhealthy food advertising. The ASA recommends that an advertisement is ‘deemed to be targeting children or young people if they are likely to comprise 25 % or more of the likely audience’ (although children aged 14 years or younger comprise only about 20 % of the total New Zealand population). For a definition on children’s peak viewing times as part of a children’s Code, it does not make sense to include adults in the denominator. Our paper shows that two other definitions of children’s peak viewing times (including only children’s audience measurement data) would make more sense, with the second definition being the most restrictive but the first one having an impact as well to reduce advertising during some evening hours:

• ∙ times of the day when 25 % of New Zealand children are watching; and

• ∙ times of the day when the child audience is at least 25 % of the maximum child audience of the day.

A review of the Code by seventy-seven health professors in New Zealand⁽ Reference Swinburn, Vandevijvere and Woodward ^{5 )} recommended the first definition but lowering the percentage (e.g. 10–15 % of New Zealand children watching), which will improve the likely impact of the new Code.

Regarding the methodology of our study⁽ Reference Vandevijvere, Soupen and Swinburn ^{1 )}, we acknowledge two of the comments made in the letter⁽ Reference Souter ^{4 )} are correct but are not really limitations of the study.

1. 1. The fact that an advertisement was classified as unhealthy when at least one of the products featured in the advertisement is unhealthy (e.g. the most common example is for supermarkets) is a legitimate methodological choice, although there are alternatives (e.g. looking at the most prominent foods in the advertisement or applying a less strict criterion). There is no gold standard way of dealing with this and the ASA has not specified in the new Code or in the letter how it is going to deal with such cases in practice?

2. 2. The FOUR television channel indeed closed in July 2016, but the study was conducted over the period June–August 2015 (e.g. recording of both advertisements as well as children’s audience data), which means that it would still have been useful to have included FOUR among the sample of channels at that time.

The other limitations mentioned by the author of the letter⁽ Reference Souter ^{4 )} are not correct. We did differentiate between advertisements targeting children and those targeting adults. Although the article does report on the total number and proportions of unhealthy food advertisements with promotional characters and premium offers (including those targeting and not targeting children), we did only take account of those characters and premium offers targeting children to calculate the number and proportions of unhealthy advertisements specifically targeting children. In the paper we explain that advertisements were considered as being targeted to children if they contained promotional strategies or premium offers known to appeal to children, and give some examples of those. In the ‘Results’ section we specifically write that out of the total number of promotional characters identified, only 55·6 % of those specifically appealed to children and out of all the premium offers identified, only 4·2 % were specifically targeted at children. Thus, only the latter were taken into account when calculating the number and proportion of unhealthy food advertisements specifically targeting children. So discounts as such were not considered to be targeting children in our study.

With regard to accounting for whether there is parental or guardian supervision of children or not, this is impossible to do in a study like ours and it is also not relevant. There is no evidence to show that the impact of unhealthy food advertisements is less when children are supervised compared with when they are not. It makes most sense to protect children by simply not airing unhealthy food advertisements specifically targeted at them.

We would be happy to discuss with the ASA how the data and the advertisements collected as part of our study⁽ Reference Vandevijvere, Soupen and Swinburn ^{1 )} can contribute to further specify and improve the new Children and Young People’s Advertising Code.