Discussion

Overall, the sugar content of beverages available for purchase in Australia decreased by 1·3/100 ml (or 17 %) between 2015 and 2019. However, there was substantial heterogeneity across beverage categories, with large decreases observed in soft drinks and milk-based protein drinks, compensated by minimal change or even increases in other categories, such as fruit and vegetable juices, energy drinks, and ready-to-drink coffees. SSB have accounted for about 56 % of all beverages available for purchase since 2015. The reduction in sugar content among SSB was about 10 % (0·8 g/100 ml) and mainly driven by lowering the sugar content of soft drinks, milks and milk-based protein drinks. Taken together, these findings suggest the overall reduction in sugar content for all beverages was due to a combination of new low- or no-sugar products on the market and reformulation of some categories of SSB. Between 2015 and 2019, there was a 3 % increase in the proportion of beverages containing at least one NNS, more notably for waters, soft drinks and electrolyte/sports drinks. The decrease in sugar content for all beverages and SSB separately was, in general, larger for non-signatories than for signatories of the voluntary industry pledge from 2018, thus suggesting the pledge had little or no impact on lowering sugar content of beverages available for purchase in Australia.

The larger reduction in sugar observed for beverages overall (17 %) v. SSB alone (10 %) suggests at least some of the overall reduction is due to the appearance of new products with low or no sugar, rather than efforts to lower the sugar content of existing products. This is important because consumers may be reluctant to replace their usual beverage choices with low- or no-sugar substitutes, thus limiting the potential impact on population sugar intake^{(Reference Fichera, Mora and Lopez-Valcarcel21)}. In addition, SSB account for over half of all beverages available for purchase in Australia every year since 2015. Mean sugar content of SSB has declined only by 10 % between 2015 and 2019, suggesting only modest reformulation has occurred. However, the almost 25 % lowering in sugar content for SSB in the soft drinks category demonstrates reformulation is feasible and should be applied to other beverage categories. Our findings suggest that reformulation was achieved, at least in part, by adding NNS, as shown by the rise in the percentage of beverages containing NNS. Although NNS have minimal or no calories and are used in minute quantities, their long-term health effects remain controversial^{(Reference Mooradian, Smith and Tokuda22)}, with emerging evidence suggesting they may have detrimental consequences at least for some individuals^{(Reference Suez, Cohen and Valdés-Mas23)}. Therefore, while it is undoubtedly in the interest of public health to reduce intake of added sugars, their replacement by NNS should be cautious and complemented by policies that promote replacement of SSB by healthier alternatives, such as plain water or tea. A detailed analysis of the changes in NNS across the entire food and drink supply and its implications in Australia is available elsewhere^{(Reference Dunford, Coyle and Louie24)}.

We found no evidence that the ABC’s voluntary sugar reduction pledge had led to a significant change in the sugar content of beverages available for purchase in Australia. Trends in sugar content for all drinks and SSB were broadly more favourable for manufacturers that did not sign the pledge than pledge signatories. This may be due to self-selection as the pledge was voluntary and, thus, companies with fewer SSB on their portfolio may have been more likely to sign the pledge than others. Furthermore, the decision by ABC to make the pledge apply retroactively from 2015 appeared to be an attempt to accrue credit from pre-existing reductions in sugar content in some cases, with our results showing that between 2015 and 2018, a substantial reduction in the mean sugar content of products available from pledge signatories had already occurred. It is, thus, germane to ask whether industry should have established a more ambitious lower target (i.e. a larger percentage reduction), or have used 2018 as a baseline rather than 2015. Our study casts doubt on signatories’ commitment to reduce the sugar content of their beverages and invest in reformulation of their SSB. In addition, the effectiveness of the pledge is limited as signatories only account for about 10 % of the beverages available for purchase in Australia, even if they may represent a larger proportion of sales. Indeed, Australia’s two largest non-alcoholic beverage bottlers, Coca-Cola Europacific Partners Australia and Asahi Holdings Australia Pty limited, are pledge signatories and represent 56·3 % of the market share for soft drinks and 85·7 % for bottled water⁽²⁵⁾. Nonetheless, our findings suggest that voluntary sugar reduction initiatives will not drive consistent improvement across the entire beverage supply and are likely to have suboptimal public health impacts due to inherent conflicts between the public health objectives of such initiatives and the beverage industry’s profit imperative^{(Reference Knai, Petticrew and Douglas26)}. Food companies are often less committed to engage with sugar reduction programmes in the absence of a level playing field created by mandatory legislation, complete with sanctions or some form of financial incentive or penalty for lack of progress towards sugar reduction targets^{(Reference Bandy, Scarborough and Harrington27)}. Commercial conflicts of interest are also likely to impair the achievement of the recently adopted government-endorsed voluntary targets for sugar reduction included in Australia’s Healthy Food Partnership initiative⁽¹⁶⁾.

By 2021, the Obesity Evidence Hub estimated that about fifty countries worldwide had implemented some form of tax on SSB despite wide variation in type and design of tax across countries (e.g. tiered v. uniform tax, eligible drinks, sugar limits, etc.)⁽²⁸⁾. The benefits such taxes are fourfold: they introduce a price signal to consumers that a product is unhealthy; they create disincentives to purchase higher priced products; they provide profit incentives for manufacturers to systematically reformulate to reduce sugar content; and they generate revenue that governments can reinvest in public health and healthcare interventions^{(Reference Wright, Smith and Hellowell29)}. Examples from across the world illustrate how those benefits can be achieved. In Mexico, a fixed tax of 1 peso (£0·03; €0·04; $0·04) per litre (approximately 10 % increase in price) was applied to all non-alcoholic drinks with added sugar starting in 2014. Two years after implementation of the tax, household purchases of taxed beverages decreased by an average of 7·6 %, with larger reductions in purchases documented in urban areas, in households with children and adolescents, in low socio-economic households, and among high SSB purchasing households^{(Reference Colchero, Molina and Guerrero-López30,Reference Ng, Rivera and Popkin31)} . In Portugal, the government introduced a consumption tax levied on SSB in 2017, which was divided into two tiers: drinks with sugar contents below 80 g/l of final product (charged at €8·22/100 l) are the lower tier and those above 80 g/l of final product (charged at €16·46/100 l) are the upper tier. During the first year of implementation, this tax collected about 80 million Euros, and all revenue was invested towards the Portuguese National Health Service funding. Although sales of SSB decreased by only about 7 %, reformulation processes led to an 11 % reduction of total energy intake through consumption of SSB at population level^{(Reference Goiana-da-Silva, Nunes and Miraldo32)}. In response to these results, the Portuguese government redesigned the taxation to further encourage reformulation, but its evaluation is yet to be published^{(Reference Goiana-da-Silva, Cruz-e-Silva and Gregório33)}. In the UK, the Soft Drinks Industry Levy (SDIL) was implemented in 2018^{(Reference Bandy, Scarborough and Harrington34)}. This two-tiered levy subjects high-sugar products to a higher tax than lower sugar products and led to a 10 % decrease in sugar consumption from soft drinks per household per week, despite no change in the volume of drinks purchased overall, thus demonstrating the potential to reduce the harmful effects of sugar intake without damaging industry profits^{(Reference Pell, Mytton and Penney35)}. The findings from individual countries have been recently confirmed by a systematic review and meta-analysis, which showed that the overall tax pass-through rate was 82 % and the demand for SSB was highly sensitive to tax-induced price increases, with a mean reduction in SSB sales of 15 %^{(Reference Andreyeva, Marple and Marinello36)}. It did not find evidence of substitution to untaxed beverages or material changes in SSB consumption. Importantly, it demonstrated that tiered taxes resulted in reformulation and reduced sugar content of taxed beverages, whilst local-level taxes incentivised cross-border shopping, which are useful lessons for countries considering the pros and cons of different taxation models.

The mounting evidence supporting effective taxes on SSB elsewhere warrants consideration of the policy’s merits in Australia. Modelling by the Australian Medical Association in 2021 suggested a tax on selected SSB would reduce sugar consumption from soft drinks by 12–18 % and raise annual government revenue of $749 to $814 million⁽³⁷⁾. This builds on prior research suggesting a 12·6 % decline in daily consumption of SSB could result in a decline in obesity of 2·7 % in men and 1·2 % in women over the lifetime^{(Reference Veerman, Sacks and Antonopoulos38)}. It has also been estimated that a 20 % tax could save 1606 lives in 25 years, and over AUD$400 million in annual healthcare costs, almost 50 % of which would accrue in the most disadvantaged groups^{(Reference Lal, Mantilla-Herrera and Veerman39)}. Furthermore, national surveys showed that 77 % of Australians supported a tax on sugary drinks if the proceeds were used to fund obesity prevention^{(Reference Falbe, Thompson and Becker40)}.

Despite its myriad benefits, a tax on SSB is not a silver bullet that will address the excessive sugar intake and obesity epidemic in the Australian population^{(Reference von Philipsborn, Stratil and Burns41)}. Complementary policy-related measures that have been proposed include introducing package size limits^{(Reference Crino, Herrera and Ananthapavan42)}, improving information on added sugars on food labels, introducing mandatory targets or compositional limits on sugar content in SSB⁽⁴³⁾, and restricting sale of SSB in some settings, such as schools, hospitals and workplaces^{(Reference Epel, Hartman and Jacobs44)}. Although a comprehensive suite of measures will likely be required to bolster ongoing progress in reducing sugar consumption in Australia, a tax on SSB would be a key step forward considering the compelling effectiveness of SSB taxes worldwide, which underpins the recommendation as one of the WHO ‘best-buy’ policies.