Methodological challenges to evaluating alcohol marketing bans

The methodological approach of Manthey et al1 has been proven useful in the evaluation of alcohol and other CODH policies and informed the establishment of ‘best buys’. But its application to the alcohol marketing domain is counter-indicated by several characteristics of marketing.5–7

These characteristics suggest that the impacts of alcohol marketing bans may be additional to or different from short-term changes in general population alcohol consumption, the outcome measure used in many evaluations of marketing restrictions. Assessing these impacts requires a broader system-level evaluation approach.8 Alcohol advertising is a complex system and, by extension, a system perspective is needed in any review seeking to understand the effects of alcohol advertising restrictions. Systemic effects occur at the individual, societal and commercial levels. At the commercial level, for example, marketing serves as a barrier to entry by new actors and generates oligopoly profits, which can then be used for further marketing.9 A further counter indication is the cumulative and lagged effects of advertising, sometimes referred to as accumulated advertising goodwill.10 These are considerable and mean that effect sizes in short-term outcome evaluations will be diluted.

Varied functions and effects of alcohol marketing

Marketing performs important roles for commercial interests, and these vary across different contexts, including the stages of market development. Thus, a range of factors can inform the evaluation of the value of a policy intervention.

Within expanding markets, such as in middle-income countries with growing economies and young populations, advertising theory predicts an increase in consumption as a result of marketing.11 Marketing in this context works to recruit new consumers in sectors where the prevalence is low, such as younger people and women. It also encourages the purchase of commercial products, replacing and/or adding to traditional beverages, and promotes the use of alcohol products in settings in which alcohol would not previously have been used.

It is highly likely, however, that marketing will not increase the prevalence of drinking beyond its already high level in the saturated markets found in some high-income countries; nevertheless, it performs important functions. Population replacement requires that cohorts of new potential drinkers have to be supported in valuing alcohol products, as did their elders. Of particular significance are the future heavy drinkers, a very important sector for the profits of the industry.12 Marketing also supports norms around alcohol products as ordinary commodities with the implication, therefore, of not requiring stringent controls on supply, affordability and marketing.

Marketing as normalisation

Marketing is directed towards stakeholders as well as consumers. This function of marketing serves to contribute to the normalisation of alcohol products as part of everyday life and creates powerful positive associations with alcohol brands and products. These norms and attitudes provide a hostile environment for health promotion13 and policy development. The implication of the messaging is that restrictions on the sale and marketing of products are not needed. If a ban on alcohol marketing affects ideas about what policies are seen as appropriate, leading to policy change, this will have potential longer-term effects on consumption.

Lack of case studies of total bans

Evaluation requires good examples. The WHO14 policy recommendation to ban alcohol marketing has been identified as the least well-implemented NCD policies globally and implementation fell between 2015 and 2017.15

The few examples of evaluations of restrictions on alcohol marketing, largely carried out between 1975 and 2000 and primarily drawn from high-income countries, are most often only for partial restrictions.7 The partial nature of the restrictions, for example, on broadcast media at certain times of the day or removing alcohol promotions on a university campus, have considerable implications. It is well documented that marketing is effective via a great many different venues and modes and that when one avenue is restricted, it is very viable for the commercial interests to find alternative approaches to achieve the same ends. This makes it inappropriate to include partial bans when analysing the effectiveness of removing exposure to alcohol marketing.

Difficulties in the implementation of bans on marketing and therefore evaluations of strong cases have been exacerbated by the increasing use of digital marketing. Alcohol producers were early adopters of digital marketing and now use this approach, turbo charged by AI models that profile, track and target consumers, to both promote and sell their products in digital channels. This form of marketing is largely ephemeral: the messages are only visible to recipients, and the automated models that target them are not open to public scrutiny.16 Increasingly the few countries with bans on alcohol marketing in place are reporting the presence of alcohol marketing in social media and on other digital platforms.

Indirect estimates of likely effects of marketing bans

In the absence of evaluations of comprehensive bans on alcohol marketing, a recent review of alcohol policy drew on evidence of the causal impacts of alcohol marketing to conclude that complete bans are a necessary aspect of a policy to reduce alcohol harm, citing bans as ‘best practice’.7 The reviewed studies included evidence of neurobiological impacts, as well as impacts related to exposure as measured with experimental, cross-sectional and longitudinal designs.7

Advertising theory emphasises that the cumulative effect of repeated exposure, coupled with other socio-psychological factors, contributes to increased alcohol consumption among vulnerable populations such as already heavy drinkers (p. 56).17 This may reflect the characteristics of the audience and also disproportionate exposure. The potential for disproportionate exposure has increased exponentially because of highly personalised targeting in which digital data inform the design of and exposure to curated sequences of ads. The power of digital marketing has now also increased due to the incorporation of online purchase and home delivery options, used more by heavier drinkers.18 In some contexts, heavier drinkers have maintained consumption in sectors that have decreased consumption overall.19 The personalised nature of the marketing may account for this.

The benefits of marketing for the alcohol industry underpin the efforts of alcohol producers to avoid bans, often by establishing and promoting ineffective voluntary codes (self-regulation) and regimes and deflecting restrictions on alcohol marketing in other ways. Applying the ‘scream test’20 might suggest the industry’s concern to prevent bans on marketing is stronger than their concern over availability restrictions and alcohol taxes.