3.1.1 Discrimination

Discrimination under international trade and investment law incorporates the concepts of both ‘de jure’ discrimination – discrimination on the face of the measure or origin-based measures – and ‘de facto’ discrimination applying to measures that, on their face, appear ‘origin-neutral’ and involve the application of formally identical legal provisions to domestic and imported products or producers, but would, in practice, accord less favourable treatment to certain imported products or investors (US – COOL, 2012). For example, claims may be made that a measure such as an SSB tax that is origin-neutral and applies to both domestic and imported products in effect discriminates in its operation against or between imports from other countries.

Under international trade law, the concept of discrimination in relation to an internal tax on SSBs is governed by the principles of national treatment and most-favoured nation treatment. Internal taxes are those that apply to both domestic and imported products, such as value-added taxes, sales taxes and excise duties, as opposed to purely border measures such as customs duties (Van den Bossche and Zdouc, Reference Van den Bossche and Zdouc2013). National treatment obligations embodied in WTO agreements prohibit a country from discriminating against other WTO Member States. Most-favoured nation treatment provisions prohibit a country from discriminating between other WTO Member States. The broad purpose of these provisions is to avoid protectionism and to ensure equality of competitive conditions.

Article III:2 of the GATT enshrines the national treatment principle in relation to internal taxes. Under GATT Article III:2 first sentence, WTO Member States agree that they will not levy internal taxes on imported products or charges directly or indirectly ‘in excess’ of those applied to ‘like’ domestic products. This requires determination of whether imported and domestic products are ‘like’ products and whether the imported products are taxed ‘in excess’ of domestic products. If both of these questions are answered in the affirmative, the measure is inconsistent with Article III:2 (Canada – Periodicals, 1997). This can be seen in Figure 1.

Figure 1

General Agreement on Tariffs and Trade (GATT) III:2, first sentence

In determining whether domestic and imported products are ‘like’ products, the WTO panel and Appellate Bodies have stated that this should be construed narrowly and requires a determination of the nature and extent of the competitive relationship between and among products, including considerations such as product properties, nature and quality; end uses; consumer tastes and habits; and tariff classification.Footnote ³ In determining whether imported products have been taxed ‘in excess’ of domestic products, the WTO Appellate Body in Japan – Alcoholic Beverages II (1996) established a strict benchmark, with even the smallest amount of excess of tax leading to the conclusion that the tax is inconsistent with GATT III:2, first sentence.

To illustrate how Article III:2 of GATT operates, in the case of Mexico – Taxes on Soft Drinks (2005) the United States challenged Mexico’s tax measures that implemented, among a number of measures, a 20% tax on the transfer or, as applicable, importation of soft drinks and other beverages that used any sweetener [including high-fructose corn syrup (HFCS) and beet sugar] other than cane sugar and a 20% distribution tax on specific services (e.g. brokerage, consignment and distribution) provided for the purpose of transferring products such as soft drinks and other beverages that use any sweetener other than cane sugar. The differential taxation regime was not implemented for public health purposes but rather implemented by Mexico to ‘secure compliance’ by the United States with provisions in the North American Free Trade Agreement (NAFTA) regarding access of Mexican-produced sugar to the US market.

The WTO Panel examined the tax measures as imposed on beet sugar and HFCS as sweeteners, as it was the presence of non-cane-sugar sweeteners that triggered the tax and therefore non-cane-sugar sweeteners were indirectly subject to the soft drink tax when they were used for the production of soft drinks and syrups. The WTO Panel also examined the tax measures as imposed on the soft drinks and syrups that use any sweetener other than cane sugar.

The WTO Panel found that the tax measures, as imposed on beet sugar and on soft drinks and syrups, were inconsistent with Article III:2, first sentence. In examining compliance with GATT Article III:2, first sentence, the WTO Panel compared the physical properties, nature and quality; end uses; consumers’ perceptions and behaviours; and tariff classification of beet sugar (an imported product) and cane sugar (a domestic product). The Panel found these to be ‘like products’. The WTO Panel also compared soft drinks and syrups sweetened with non-cane sugars including HFCS and beet sugar (which were mostly imported) with those sweetened with cane sugar (which were mostly domestic). The Panel also found these to be ‘like products’. The Panel found that beet sugar and imported soft drinks and syrups sweetened with non-cane-sugar sweeteners were subject to internal taxes in excess of those applied to like domestic products. The WTO Appellate Body upheld the findings of the Panel (Mexico – Taxes on Soft Drinks, 2006).

If an internal tax is found to be consistent with GATT Article III:2 first sentence, there is a need to examine whether the measure is consistent with the second sentence of GATT Article III:2, which prohibits internal taxes or charges where ‘directly competitive or substitutable products’ are ‘not similarly taxed’ so as to ‘afford protection to domestic production’. This can be seen in Figure 2. ‘Directly competitive or substitutable products’ have been defined more broadly than ‘like’ products for the purposes of Article III:2, applying to products that are ‘interchangeable’ or offer ‘alternative ways of satisfying a particular need or taste’ (Korea – Alcoholic Beverages, 1999: 33). Determining whether products are ‘directly competitive or substitutable products’ includes consideration of the competitive conditions between the products in the relevant market, in light of their physical properties, their common end uses including evidence of advertising activities, tariff classification, channels of distribution and price relationships including cross-price elasticities without placing an undue emphasis on the quantitative elements of a competitive relationship between products (Mexico – Taxes on Soft Drinks, 2005: 107; Korea – Alcoholic Beverages, 1999: 38). To be inconsistent with GATT III:2, second sentence, the tax burden on imported products must be heavier than on directly competitive or substitutable domestic products, and this burden must exceed a de minimis threshold. The de minimis threshold is determined on a case-by-case basis (Japan – Alcoholic Beverages, 1996: 27). The dissimilar tax must be applied so as to afford protection to domestic production with this protective application discerned from the design, architecture and revealing structure of the measure (Japan – Taxes on Alcoholic Beverages, 1996: 29).

Figure 2

General Agreement on Tariffs and Trade (GATT) III:2, second sentence

In the case of Mexico – Taxes on Soft Drinks and Other Beverages, the WTO Panel found that Mexico’s tax measures, as imposed on HFCS as a sweetener, were inconsistent with Article III:2, second sentence, as the measures imposed dissimilar taxation on directly competitive or substitutable products so as to afford protection to the Mexican domestic production of cane sugar. Mexico subsequently withdrew the measures in compliance with WTO Dispute Settlement Body recommendations (WTO, 2017a).

The concept of most favoured nation treatment with regard to internal taxes is contained in GATT Article I:1, which prohibits a member country from discriminating between countries by granting any advantage, favour, privilege or immunity to any like product from or for another country with respect to all matters referred to in Article III:2 of GATT (Canada – Autos, 2000). Therefore, governments must be cautious in designing SSB taxes to avoid protectionism with regard to not only domestic products but also to products of one WTO member country over another.

Under international investment law, although the terms of treaties vary, the obligation of non-discrimination is also commonly expressed under the principles of ‘national treatment’ where states agree to accord treatment to foreign investors, or its investment, no less favourable than that provided to domestic investors in ‘like’ circumstances and ‘most-favoured nation treatment’ where states agree to treat an investor from a party to an agreement, or its investment, no less favourably than an investor, or its investment, from any third country. The concept of non-discrimination is also an element of the fair and equitable treatment obligation in investment treaties, which will be discussed below. The NAFTA cases of Archer Daniels Midland and Tate & Lyle Ingredients Americas, Inc. v. United Mexican States (2007), Corn Products International, Inc. v. United Mexican States (2008) and Cargill, Incorporated v. United Mexican States (2009) concerned, in an investment law context, the imposition of Mexico’s 20% excise tax on soft drinks and syrups that used any sweetener other than cane sugar. The taxation regime was found to be discriminatory and contrary to the national treatment principle as it unreasonably afforded protection to domestic cane sugar producers, demonstrating an intention on the part of Mexico to treat US investors less favourably because of their nationality.

3.1.2 Legitimate distinctions

It is important to note that under both international trade and investment law, governments may draw regulatory distinctions against and/or between foreign products and investments if these distinctions are based on bona fide, evidence-based public health regulatory reasons.

In the context of international trade law, the Appellate Body has found that, in examining whether products are ‘like’ products for the purposes of discrimination, WTO Panels must consider all relevant evidence including evidence relating to the health risks associated with a product (EC – Asbestos, 2001). The protection of public health would also be relevant to show a measure is not applied ‘so as to afford protection’ for the purposes of GATT Article III:2, second sentence (Mitchell and Voon, 2011).

In addition, measures that are found to effect discrimination will not violate GATT if the measures are ‘necessary’ to protect human health by virtue of an exception under Article XX(b) GATT, provided the measures are designed and implemented in good faith and do not unjustifiably or arbitrarily discriminate or restrict international trade. In determining whether a measure is ‘necessary’ to protect human health, WTO panel and/or Appellate Body will conduct a weighing and balancing exercise, in light of the importance of the objective pursued by the government, of the contribution the measure makes to the objective of protecting human health and the degree of restriction the measure places on international trade. If this weighing and balancing process indicates that the measure is necessary, the decision-making bodies will consider whether there are less trade-restrictive alternative measures that are reasonably available that make an equal or greater contribution to the objective (Brazil – Retreaded Tyres, 2007: 71).

A small number of tobacco control measures have been challenged under WTO law on the grounds that they discriminate against imported products. In the case of US – Clove Cigarettes (2011, 2012), the WTO Panel, upheld by the Appellate Body, found that US measures prohibiting the sale of certain flavoured cigarettes, including clove cigarettes, but exempting menthol or tobacco-flavoured cigarettes, were discriminatory. The Panel found that clove cigarettes and menthol cigarettes were ‘like’ products and that the detrimental impact on competitive opportunities for clove cigarettes reflected discrimination against these ‘like’ products from Indonesia compared to menthol products predominantly produced in the United States.

One of the arguments proposed by the United States was that the regulatory distinction drawn between clove and menthol cigarettes was for legitimate public health reasons. The United States highlighted the potential development of a black market and burden on the health care system associated with treating withdrawal symptoms of the significantly higher numbers of menthol cigarette smokers if menthol cigarettes were also banned. In rejecting this argument, the Appellate Body noted that the addictive ingredient in menthol cigarettes was nicotine, not menthol. While recognising the right of members to enact measures to pursue legitimate public health objectives, the WTO Panel and Appellate Body found that the US measure violated the relevant discrimination provision under WTO law and the detrimental impact on imports did not stem from a legitimate public-health-based regulatory distinction.

In the context of international investment law, tribunal decisions have supported the differential treatment of foreign and domestic investors where this is justified for public policy reasons. Where a government can establish a reasonable or rational connection between the measure and the achievement of its objective, the state may be found not to have acted in a discriminatory manner [Saluka Investments BV (The Netherlands) v Czech Republic, Reference Tsikas2006: 94].

Drawing from these lessons, it is clear that governments seeking to implement an SSB tax need to consider whether the tax discriminates against and/or between imported products or foreign investors in its design, structure and application. Regulatory distinctions, including determining to implement a sugar-based tax rather than a tax on other unhealthy nutrients such as fats and salt, imposing a tax on beverages rather than other sugary products, determining which beverages will be included and excluded from SSB tax coverage, and the threshold levels of sugar required for the tax to apply, should be carefully considered and supported by legitimate, bona fide public health reasoning and/or evidence.

In determining coverage of an SSB tax, governments should consider the evidentiary basis justifying the coverage of products such as national dietary guidelines and/or international documents such as the WHO Guidelines on Sugars Intake for Adults and Children (WHO, 2015).

3.2.1 Measures are evidence-based and tailored to public health objectives

One clear lesson to emerge from tobacco control cases is that it is good practice for countries to ensure that measures, including their objectives, are evidence-based and defined and tailored in a manner that is linked to evidence and public health goals, including, where relevant, commitments under domestic and international instruments. Objectives should be defined in a manner that enables the effects of the measure to be monitored to the extent possible since the contribution the measure makes to its objective may be considered by courts.

This has been observed in the context of claims that tobacco control measures are more trade-restrictive than necessary. Over the last few years, tobacco control measures have been discussed in the WTO’s Technical Barriers to Trade (TBT) CommitteeFootnote ⁴ and Trade-Related Aspects of Intellectual Property Rights (TRIPS) Council,Footnote ⁵ and claims have been brought against WTO Member States that tobacco control measures are more trade-restrictive than necessary to fulfil a legitimate objective (US – Clove Cigarettes, 2011, 2012; Australia – Tobacco Plain Packaging, date pending).

One of the measures currently facing such a claim before the WTO is Australia’s plain packaging measure. On 29 April 2010, the Australian Government announced that it would introduce mandatory plain packaging of tobacco products as part of a comprehensive strategy to reduce smoking rates in Australia. The Tobacco Plain Packaging Act was passed by the Australian Parliament in November 2011 and took full effect from 1 December 2012. The objects of the Tobacco Plain Packaging Act 2011, as outlined in Section 3 of the Act, specifically include improving public health by discouraging people from taking up smoking, or using tobacco products; encouraging people to give up smoking and to stop using tobacco products; discouraging people who have given up smoking, or who have stopped using tobacco products, from relapsing; reducing people’s exposure to smoke from tobacco products; and to give effect to certain obligations that Australia has as a party to the WHO FCTC by regulating the retail packaging and appearance of tobacco products in order to reduce the appeal of tobacco products to consumers, increase the effectiveness of health warnings and reduce the ability of the retail packaging of tobacco products to mislead consumers about the harmful effects of smoking or using tobacco products.

These objects were supported by extensive research evidence set out in the reports of the National Preventative Health Taskforce, a group of Australia’s leading public health experts (Commonwealth of Australia, 2009; Australian Government Department of Health, 2014). Australia’s plain packaging measure has been subject to extensive evaluation, with a Post Implementation Review of tobacco plain packaging released by the Australian Government on 26 February 2016 confirming that the measure had begun achieving the objectives of reducing smoking prevalence and exposure to tobacco smoke in Australia (Chipty, Reference Chipty2016).

While the outcome of the WTO dispute is pending, it has been reported that Australia has been successful in relation to the WTO challenge (ABC News, 2017). As outlined in the Integrated Executive Summary of Australia’s Submissions to the WTO, Australia has a ‘large body’ of qualitative evidence supporting the conclusion that tobacco plain packaging is apt to contribute to its clearly outlined public health objectives (Australia – Tobacco Plain Packaging, 2016: 5).

Countries implementing an SSB tax should similarly ensure that the objectives of the relevant legislation are drafted in a manner that reflects the evidence base.

States can bolster the formulation and design of SSB taxes by gathering the best available evidence base to support the measure. The WHO Fiscal Policies for Diet and Prevention of Noncommunicable Diseases Report states that there is growing evidence in relation to the potential impact of fiscal policies for promoting healthier diets, including evidence of the effects of SSB taxes in reducing consumption of the taxed products. In the Report, WHO recommends that governments conduct a situation analysis ‘using all existing relevant information’ to inform the design of the measure including the relevant objectives, and highlights the importance of monitoring and evaluation of measures in order to document their effectiveness in terms of achieving objectives, and monitoring the impact on purchase, consumption, product reformulation and substitution patterns (WHO, 2016a).

Monitoring and evaluation should be considered and planned in advance of design and implementation of the measure. Prior to introduction of Hungary’s PHPT, laboratory analyses of the salt, sugar and other unhealthy ingredient content of foods were collected to identify ‘unhealthy’ food and, in combination with data on consumption patterns of unhealthy food products, was used to prepare the legislation and served as a baseline for monitoring and evaluation of consumption and product reformulation patterns (WHO Regional Office for Europe, 2015).

One evidence gap highlighted in the WHO Fiscal Policies for Diet and Prevention of Noncommunicable Diseases Report is the quantification of the impact of SSB taxes on improving weight outcomes and other diet-related risk factors (WHO, 2016a). In the interim, there is clearly a need for governments to understand and establish clear indicators along the path to the overall objective of obesity and overweight reduction. Hawkes et al. (Reference Hawkes, Smith, Jewell, Wardle, Hammond, Friel, Thow and Kain2015) state that an assessment of the impact of a measure must be based on a clear comprehension of the mechanisms through which the measure could work, including what the expected outcomes would be and when they could be expected to occur. The authors highlight that this is pertinent in the area of public health due to a time lag or indirect relationship between food consumption patterns and overweight and obesity prevalence and health benefits, with the need for indicators to measure progress along the pathways of change towards reduced prevalence of overweight and obesity.

Tobacco control cases have also recognised that evidence in relation to the effects of public health regulatory measures is complex. Cases before domestic courts, the WTO and international investment tribunals have recognised that the impact of a public health measure may be difficult to determine in the short term, or where the measure is part of a complementary suite of measures (Brazil – Retreaded Tyres, 2007: 59–60; BAT v. Secretary of State for Health, 2016: 258–259; Philip Morris v. Uruguay, 2016: 87–88). The WTO’s Appellate Body has stated that the contribution a measure makes to the achievement of its objective may be demonstrated with quantitative projections or qualitative reasoning that the measure is apt to produce a material contribution to the achievement of its objective in the future.Footnote ⁶

Countries may not have the capacity to undertake additional studies or further research in relation to evidence. In such cases, it may be enough to rely on international experience. The international investment tribunal in the Philip Morris v. Uruguay case noted that states, particularly those with limited technical and financial resources, should be able to rely on international normative instruments, practices, studies and research to demonstrate the reasonableness of measures (Philip Morris v. Uruguay, 2016: 113). It is critical for countries to carefully consider the mounting evidence in relation to SSB taxes and the applicability of the evidence base to their domestic context in the design and administration of the SSB tax, including its objectives.

3.2.2 Less trade-restrictive ‘alternative’ measures considered and a comprehensive strategy adopted

As previously discussed, in assessing the ‘necessity’ of a measure under international trade law, WTO Panels and Appellate Bodies have considered the existence of any potential alternatives to the measure under challenge that are reasonably available, make an equal (or greater) contribution to the objective and are less trade-restrictive than the measure at issue. Governments should therefore consider whether there are less trade-restrictive alternatives to an SSB tax that make an equal or greater contribution to the identified public health objectives, noting that an import ban is “by design as trade-restrictive as can be” (Brazil – Retreaded Tyres, 2007: 59).

The WTO Appellate Body has held that an alternative measure may not be considered reasonably available “where it is merely theoretical in nature, for instance, where the responding Member State is not capable of taking it, or where the measure imposes an undue burden on that Member State, such as prohibitive costs or substantial technical difficulties” (US – Gambling, 2005: 102).

It should also be noted that decision-making bodies have made a distinction between ‘alternative’ measures and ‘complementary’ measures operating as part of a suite of measures designed to achieve a particular objective. The concept of complementarity has been recognised in WTO case law:

In the context of tobacco control, Articles 4.2 and 4.4 of the WHO FCTC clearly recognise that effective tobacco control requires a comprehensive suite of measures. Measures claimed by tobacco industry advocates to be less restrictive ‘alternative’ measures have been successfully argued by governments to be complementary, not alternative, measures (US – Clove Cigarettes, 2011).

The lesson to adopt SSB taxes as part of a comprehensive suite of measures accords with nutrition policy and regulatory frameworks. In the context of nutrition policy, Hawkes et al. state that effective nutrition policies are not implemented as ‘single magic bullets’, but as part of “a combination of complementary and mutually reinforcing actions” (Reference Hawkes, Smith, Jewell, Wardle, Hammond, Friel, Thow and Kain2015: 2415). For example, Hungary’s PHPT was introduced as part of a ‘comprehensive set of measures’ targeting the promotion of healthy eating (WHO Regional Office for Europe, 2015).

The WHO Fiscal Policies for Diet and Prevention of Noncommunicable Diseases Report recognises the importance of policy coherence to ensure the most effective impact, providing examples of placing warning labels on taxed products and limiting marketing of these products (WHO, 2016a). The Lancet Public Health editorial states:

This is embodied in relevant international documents. As outlined previously, WHO’s Comprehensive Implementation Plan on Maternal and Infant and Young Child Nutrition 2012 recognises that taxes and subsidies, when used in conjunction with trade measures, effective nutrition labelling, responsible food marketing and health promotion, are important in facilitating easier access and choice of healthy diets.

Implementing SSB taxes as part of a comprehensive strategy to combat overweight and obesity will ensure that governments are following best practice public health principles and provide support in response to claims that less trade-restrictive alternative measures are reasonably available that could make an equal or greater contribution to the public health objective.