(1)

On 9 June 2011 the Commission received a complaint from Margarineforeningen (MIFU) regarding the Danish Act on taxation of saturated fat in certain food products, Law No 247 of 30 March 2011 (3) (‘the Act’). The law entered into force on 1 October 2011 and was later abolished as of 1 January 2013. The complaint concerned possible aid to producers of non-taxed products.

(2)

In order to seek clarification, an information request was sent to Denmark on 4 July 2011. The Commission forwarded the original complaint to the Danish authorities and asked them to submit their own summary of the facts in order to address the claims made by the complainant within 20 working days or to notify the measure as State aid to the Commission for approval. Denmark asked for an extension of the deadline on 18 July 2011 which was granted until 31 August 2011. On 7 September 2011 Denmark submitted its response to the Commission’s letter in which it presented its views on the issues raised by the complainant.

(3)

On 13 October 2011 the Commission forwarded Denmark’s comments to the complainant. On 27 June, 28 July, 12 September and 15 November 2011 the complainant submitted additional information in support of the complaint.

(4)

On 14 December 2011 the Commission informed Denmark that, as the Act had entered into force on 1 October 2011 without having been notified pursuant to Article 108(3) TFEU, the measure in question had been registered as non-notified aid under the number SA.33159 (2011/NN).

(5)

On 13 March 2012 the complainant submitted further information in support of the complaint.

(6)

On 16 March 2012 a meeting was held between the Commission and Denmark. On 12 April 2012 the Commission forwarded the full complaint to the Danish authorities and asked them to address and clarify the issues put forward by the complainant. In case the Danish authorities would be of the opinion that the measure does not give rise to State aid concerns, they were asked to provide their own summary of the facts, as well as a substantiated reasoning of their view. On 13 September 2012, after a request to extend the deadline for answering, Denmark replied to the Commission’s letter of 12 April 2012.

(7)

On 30 October 2012 the Commission forwarded the Danish authorities’ comments to the complainant. On 30 November 2012 the complainant submitted additional information in support of the complaint.

(8)

On 16 May 2013, after the Act ceased to be in force on 1 January 2013, a meeting was held between the Commission and the complainant.

(9)

By letter of 4 February 2015 (C(2015) 535 final) the Commission informed Denmark that it had decided to initiate the procedure laid down in Article 108(2) TFEU with regard to the Act (‘the opening decision’).

(10)

In the opening decision the Commission stated that the measures imposed by the Act appeared to have all the characteristics of State aid and requested Denmark to submit its comments and to provide all information which could be helpful for the assessment of the aid.

(11)

By letter dated 26 May 2015, Denmark submitted comments concerning the opening decision.

(12)

The opening decision was published in the Official Journal of the European Union (4). The Commission invited interested parties to submit their comments within one month. The Commission received comments from three interested parties on the measures mentioned in the opening decision. The comments received were transmitted to Denmark by letters of 22 October and 9 November 2015. Denmark responded to those comments on 21 December 2015 and 29 January 2016. By letter of 10 August 2017 the Commission requested further information from Denmark. By letter dated 15 December 2017 Denmark submitted the further information.

(13)

The Danish Act on taxation of saturated fat in certain food products had as simultaneous objectives to promote better diets and to improve the health of the Danish population (5). These were to be achieved by taxing the primary sources of intake of saturated fat in Denmark (see recital 37), while at the same time taking into account official dietary guidelines (6) (see recital 47) and other considerations (see e.g. recital 34), which led to exempting certain products from the scope of the tax.

(14)

The Act was in force between 1 October 2011 and 1 January 2013, which is therefore the relevant time frame for this decision.

(15)

The Act levied a tax on the weight of saturated fat in the food products listed in Article 1 of the Act, and identified by the Union combined nomenclature code, if the saturated fat content in the food product exceeded 2,3 percent of the weight of the product. The food products listed were:

(16)

According to Article 2 of the Act, the tax was fixed at DKK 16 per kg of saturated fat in the respective food product.

(17)

Article 3 of the Act defined operators liable to pay the tax as those who for commercial purposes

(18)

Exempted from paying the tax (and from registration under the Act) were those businesses with annual sales of taxable food products totalling DKK 50 000 (exclusive of tax) or less (Article 4(6) of the Act).

(19)

The amount of saturated fat in the food products listed in Article 1 (2-7) of the Act was determined based on:

(20)

If the base amount on which the tax was to be paid could not be set using one of the above methods, the tax was to be paid based on the food product’s total fat content or ultimately on the food product’s net weight.

(21)

According to Article 8 of the Act, a coverage tax was to be paid on the constituents used in the production of food, where these constituents originated from any taxable food product specified in Article 1 of the Act and where such food was received, imported or sold by distance selling in Denmark. The coverage tax was only imposed on food received, imported or sold by distance selling in Denmark, as a taxable food product produced in Denmark was always taxed in its ‘pure’ form at the first point of sale/use and therefore there was no need for a coverage tax on food produced in Denmark.

(22)

According to Article 12 of the Act, the tax basis was the weight of saturated fat in the constituents which originate from the taxable food product specified in Article 1 and which were used in the production of food received, imported or sold by distance selling.

(23)

According to the second paragraph of Article 1 of the Act, taxable businesses had to be able to document the weight of saturated fat in the constituents used in the production of food and were, on request, to present a declaration thereof from the producer. For constituents subject to a coverage tax as specified in Article 1(1), i.e. meat, the tax basis was to be determined on the basis of the average rates provided for in Annex I.

(24)

For other constituents on which coverage tax was to be paid, the producer could, when issuing the producer declaration, make use of publicly available food product information, which provided average standards for the amount of saturated fat in specific food products, to determine the tax basis for the constituents subject to the coverage tax which were used in the production of the food.

(25)

Where the taxable business was unable to present the specified producer declaration, constituents other than those specified in Article 1(1) of the Act would be subject to a coverage tax based on the total content of both saturated and unsaturated fat in the food product or, where such information could not be provided, on the basis of the net weight of the food product.

(26)

In case the Danish Tax and Customs Administration found that the coverage tax calculated in accordance with Article 12(2) of the Act would constitute a higher amount than the coverage tax calculated on the basis of the total fat content or the net weight, the Danish Tax and Customs Administration could determine the tax basis on the basis of an estimate.

(27)

As described in section 2.1.8 of the opening decision, the Act contained 6 possible aid measures. These were:

(28)

In the opening decision, based on the available information, the Commission took the preliminary view that the conditions of Article 107(1) TFEU were fulfilled for measures 1-3 and 5-6. The measures were deemed not to be justified on the basis of the nature or general scheme of the system of reference as Denmark failed to demonstrate that the measures were necessary for the functioning and effectiveness of the general system of the tax on saturated fat and that there would be a link between the alleged specific features of the measures and the tax relief granted for certain groups of producers only.

(29)

The Commission initiated the procedure under Article 108(2) TFEU, as the Commission had doubts as to whether any of the measures were compatible with the internal market, in the light of the information available at that time.

(30)

The Danish authorities transmitted their comments to the Commission on 26 May 2015. Following a meeting between the Commission and the Danish authorities on 18 December 2015 and as a response to the comments submitted by interested parties, Denmark submitted further comments on 21 December 2015 and 29 January 2016. The comments can be divided into three parts — those concerning the system of reference, those concerning the thresholds for taxation and those concerning the differential treatments.

(31)

Denmark reiterates that the objectives of the Act were to ‘promote better diets and therefore improve the health of the population’ (9). It was not to impose a pecuniary burden on (all) products containing saturated fat in general, but only on those products which were the primary sources of saturated fat in Denmark (10), while taking into account official dietary guidelines (11). Based on this objective, the tax was designed to encourage a reduction of the intake of saturated fat and to shift the intake from products with a higher content of saturated fat to products with a lower content of saturated fat, while taking into account official dietary recommendations and economic and administrative efficiency.

(32)

It was not the purpose of the tax to exclude saturated fat from the diets of the Danish population. Fat should be a part of a normal diet as it contributes to the intake of the vital fatty acid and to absorb fat-soluble vitamins. However, according to the official dietary guidelines, fat from meat and dairy products should be limited and to a wide extent substituted with fat from vegetable oil and fish (12).

(33)

Further, the Danish authorities submit that Union State aid rules do not require the Member States to be able to present positive scientific proof that a tax regulation does indeed have the presupposed behavioural or other effects.

(34)

The Danish authorities state that they always seek to either reduce existing administrative burdens on business or, in case of new legislation, not to strain it more than absolutely necessary, particularly for small and medium-sized enterprises (SMEs). This has been a particular aim of successive governments for years and is apparent in the application of the Danish VAT law for example.

(35)

The proposal for the Act, section 2, described these concerns in the following way: ‘There are many aspects that need to be taken into consideration in the design of a completely new tax. On the one hand, a tax on saturated fat should include a significant proportion of ordinary food products, to ensure this has the greatest effect on consumption. It should also be as simple as possible and ideally be as far back in the supply chain as possible, so the fewest number of entities as possible are subject to the administrative burdens associated with being registered, declaring, and settling tax.’

(36)

To levy the tax at the beginning of the supply chain and not at the end (i.e. retail sale to consumers) minimised administrative burdens as the number of retailers is significantly higher than the number of producers/importers. This approach was also the most concordant with the purpose of the Act.

(37)

From the objectives of the Act followed the necessity to establish a positive list of food products to be taxed as being the primary sources for the intake of saturated fat without counteracting the official dietary guidelines. The Danish authorities submit that — having regard to the objectives and the actual consumption habits in the Danish population — they had a wide margin of discretion as regards the decision on which products to include in this positive list in order to pursue the health and dietary objectives of the Act (13).

(38)

The Danish authorities argue that a primary source for the intake of saturated fat in the Danish population is determined by two factors: (i) the consumed quantities of the product; and (ii) the actual share of saturated fat in the product. A product is not necessarily a primary source of the intake of saturated fat in Denmark just because it has a high fat content. Moreover, a product does not necessarily raise health concerns even if it contains saturated fat, if that product is a low-fat product and/or if it has other good qualities according to official dietary recommendations.

(39)

Denmark refers to a report (14) which identifies all kinds of meat products, all kinds of dairy products and all kinds of fats/oils as the primary sources for the intake of saturated fat in Denmark. These groups of products together contribute with 81 percent to the intake of saturated fat (meat — 19 percent, dairy — 30 percent, all kinds of fat — 32 percent). Other product groups were not considered as primary sources of saturated fat as they did not contribute significantly (6 percent or less) to the consumption of saturated fat. Based on these facts Denmark argues that there was no need for a clear cut-off point for defining what was and was not a primary source.

(40)

The Incentive Report (15) estimates that 88 percent of the saturated fat consumed in Denmark while the tax was in force was subject to the tax (dairy — 84 percent, meat — 98 percent, other taxed products — 90 percent). The non-taxation of part of the saturated fat in dairy products (16 percent) and meat (2 percent) was a consequence of the 2,3 percent threshold set out Article 1 of the Act.

(41)

Denmark further states that saturated fat in for example bread, chips, crisps, and candy was taxed by the Act, as the saturated fat in dairy products, fats/oils, and meat used for the production of such processed food products was taxed.

(42)

Denmark refers to Directive 2000/13/EC of the European Parliament and of the Council (16) which defines ‘meat’ as ‘skeletal muscles of mammalian and bird species recognised as fit for human consumption […].’

(43)

Denmark concludes that poultry is by definition a type of meat in line with pork and beef and that, consequently, it would have been arbitrary to exclude poultry when other types of meat were taxed, just as it would have been arbitrary to exclude a certain type of cheese (e.g. goat cheese) or oil (e.g. sesame oil) when all other cheeses and oils were taxed.

(44)

In Denmark’s opinion, excluding poultry would also have gone against the basic consideration of the Act to ensure that products that were close substitutes to taxable food products were also taxed (17).

(45)

Logically, all sources of saturated fat, which were not classified as primary sources on the basis of the two criteria defined in recital 38, were considered as non-primary sources. Examples of products with no content of saturated fat are limited to a number of fruit and vegetables (e.g. lettuce and melon) and a few fish. Examples of products with a content of saturated fat between 0,1 and 0,5 percent are almost every other type of fruit and vegetables.

(46)

Food products that contributed to the total intake of saturated fat with 1 percent or less (such as fish (18), eggs, nuts, and seeds (19)) were not considered to be primary sources for the intake of saturated fat. This was the main reason for not including these product categories in Section 1 of the Act.

(47)

Denmark submits that the changes in consumption due to the tax would be in line with the dietary guidelines from the Danish Veterinary and Food Administration, which are the official Danish recommendations for a healthy lifestyle. These official recommendations include the Diet Compass — the path to a healthy balance (20) , which consists of 8 dietary guidelines — eat fruit and vegetables (6 each day); eat fish and fish products (several times a week); eat potatoes, rice or pasta and wholegrain bread (every day); cut down on sugar (especially from soft drinks, confectionery and cake); cut down on fat (especially from dairy products and meat) and select low-fat products; vary your intake of food and keep a standard weight; quench your thirst in water; be physically active (at least 30 minutes every day). In addition, the dietary guidelines recommended a daily intake of 500 ml low-fat dairy products, such as skimmed milk or buttermilk as part of a healthy diet (21). The recommendations are based on the nutritional and health benefits of consuming dairy products (22).

(48)

Denmark explains that the consumption of fruit, vegetables, and fish cannot be compared to the consumption of cheese, meat, and butter even if both kinds of products contain saturated fat. Fruit and vegetables contribute considerably to the intake of carbohydrates, dietary fibres, and several vitamins (23). According to the study of the Danish population’s dietary habits 2003-2008, both adults and children had an intake of fruit and, especially, vegetables that was below the recommended quantities (24).

(49)

Denmark submits that the National Food Institute of the Technical University of Denmark (DTU) concluded in a report on fruit, vegetables and health that almost every study shows a reverse connection between the intake of fruit and vegetables and the risk of cardiovascular diseases. Further, a number of studies have shown that the higher the intake of fruit and vegetables, the greater the health benefits (25).

(50)

Potatoes are an important source for the intake of carbohydrates and dietary fibres and they include many dietary fibres, vitamin C, B vitamins, thiamine, and niacin. Moreover, the content of saturated fat in potatoes is only 0,1 percent, and it is one of the most filling food products. The official dietary guidelines recommend a daily consumption of potatoes (26). As an alternative, rice and pasta are also recommended as especially the wholegrain versions of rice and pasta contain dietary fibres (27).

(51)

Cereals contain many important nutritional components and wholegrain products are rich on dietary fibres and B vitamins. The official dietary guidelines recommend a daily intake of bread and cereals (28).

(52)

Fish contains essential fatty acid, vitamin D, iodine, and selenium which are hard to get from other food products. Different kinds of fish contain different amounts of these healthy substances. Thus, the intake of fish should be varied among the different species and include both fatty and lean fish (29).

(53)

Denmark points out that the study of the Danish population’s dietary habits 2003-2008 shows that the intake of fish by both adults and children was less than half of the recommended amounts. Moreover, the low intake of fish contributed to a lack of vitamin D in the Danish population (30). Especially fatty fish contain important amounts of vitamin D, as concluded by the Danish Veterinary and Food Administration in a report on the consumption of fish (31). In addition, a number of studies have demonstrated a connection between the intake of fish and a reduced risk of ischaemic heart disease (32).

(54)

Denmark submits that, according to the official guidelines on milk to infants applicable at the time of the adoption of the Act, whole milk could gradually replace breast milk in the infant’s diet from the age of 6 months and as of the age of one until the age of three years, the intake of dairy products should be based on semi-skimmed milk (33).

(55)

Denmark emphasises that it lies within the discretionary power of Member States to decide on the economic policy which they consider most appropriate and to pursue for example health objectives in their tax policy.

(56)

Denmark argues that the system of reference in the case at hand is a tax on the saturated fat in the primary sources for the intake of saturated fat in Denmark, taking into account the official dietary recommendations. In other words, the ‘normal’ taxation under the Act is the taxation of saturated fat in those products identified as the primary sources for the intake of saturated fat in the Danish population taking into account the official dietary recommendations. These products and their direct substitutes are the products taxed under the Act and Denmark argues that no food products are as such exempt from the tax and that due account of the official dietary guidelines was a guiding principle throughout the preparation of the Act.

(57)

The system of reference suggested by the Commission in the opening decision would in practice imply a tax on almost every kind of food product. Such a system of reference would, however, be conflicting with the very objectives of the Act. A tax on the saturated fat in all food products would have diluted the measure and would have been counterproductive to the objective of encouraging better dietary habits.

(58)

In addition, the fact that almost 90 percent of all saturated fat consumed in Denmark is taxed under the Act (34) substantiates that the determination of the scope of the Act was in no way arbitrary, but carefully designed in order to cover all major sources.

(59)

In the Danish authorities’ view, it lies within the Member States’ discretionary powers to decide how to pursue health objectives through their tax policy, including deciding on such objectively justified thresholds as they consider appropriate for the purpose of pursuing these health objectives (35). The specific threshold is a matter for the Member State to decide as long as the threshold chosen lies within the limits of what can be justified in light of the purpose of the tax.

(60)

In support of this argument, Denmark refers to the case law of the Union Courts, according to which a measure that consists in criteria that are objective and applicable to any potentially interested operator which fulfils those criteria is not of a selective nature (36).

(61)

Denmark further argues that the question of whether the 2,3 percent threshold was a measure that derogated from the normal tax regime should be determined by an assessment of whether the producers of products below the threshold were in a factual and legal situation that was comparable to the situation of the producers of products above the threshold (37). To that end, the effects of the threshold must be assessed in the light of the objectives of the tax regime (38).

(62)

Thus, the crucial question is whether the effects of the measure entailed a derogation from the system of reference as defined by its objectives.

(63)

Denmark refers in that regard to the General Court’s judgment in the British Aggregates case, where the Court held that some of the untaxed materials were at least equally, if not more, harmful to the environment than the extraction of other, taxed, materials, and that it had not been demonstrated that it was the environmentally harmful nature of the extraction of the untaxed materials that distinguished their situation from that of the taxed materials (39).If the effects of a measure contribute to the objective of the system the measure is not a derogation from the system of reference (40).

(64)

According to the Danish authorities, the 2,3 percent threshold was introduced in light of the intrinsic purpose of the Act. The explanatory note to the draft act reads: ‘It is proposed that a de minimis threshold is set for food that is included in the tax, to minimise the number of low-fat foods which are included in the tax and at the same time to take into account that nutritional recommendations state that a part of the recommended daily intake originates from fat and ideally unsaturated fat’ (41).

(65)

In recital 107 of the opening decision, the Commission highlights that the Prevention Commission recommended that the tax on saturated fat should apply to all dairy products including milk. Denmark acknowledges that this recommendation may be true when focusing narrowly on reducing the intake of saturated fat in the population in general, but underlines that the tax did not have such a narrow focus. On the contrary, the objective of the tax was to improve the dietary habits of the Danish population. Since the Prevention Commission did not consider the negative impact that a tax on all dairy products would have contrary to the official dietary recommendations regarding consumption of dairy, the legislator did not follow the Prevention Commission’s proposal on this particular point.

(66)

The Danish authorities state that the level of the 2,3 percent threshold ensured that the tax did not counteract the official dietary guidelines which explicitly recommended a daily intake of 500 ml low-fat dairy products (cf. recital 47) and that low-fat meat should be preferred over fattier meat. Setting a threshold amplified the behavioural incentive to choose low-fat products and Denmark argues that a tax without a minimum threshold could be considered to have effects which would run counter to the overall objectives of the tax to promote better diets and improve the health of the Danish population. This threshold is therefore intrinsic in the purpose of the tax and not prima facie selective. In addition, a threshold reduced unnecessary administrative burdens.

(67)

Denmark argues that the threshold could have been set at 2,0, 2,5, or even 3,0 percent and still would have had positive effects on the health of the Danish population and would have amplified the incentive to choose low fat products over those with a higher fat content. However, setting the threshold at 2,3 percent specifically also ensured that the tax did not counteract the official guidelines on milk to infants as set out in recital 54.

(68)

The consumption of whole milk is highly sensitive to price changes, where estimates show that a price increase of 1 percent will lead to a decrease in consumption of 1,14 percent (42). Denmark therefore argues that a tax with a threshold below 2,3 percent could have created an incentive for families with infants to either limit their consumption of milk or to buy skimmed milk instead of whole milk because of the lower tax on skimmed milk. Accordingly, it was assumed that the tax could have caused a reduction in the consumption of whole milk by Danish infants contrary to the dietary recommendations. In the Danish authorities’ view, this is sufficient justification for setting the threshold at 2,3 percent and showing that the effects of the threshold contributed to the objective of the tax.

(69)

Denmark points to the fact that the threshold did not create an incentive to buy whole milk instead of skimmed milk. As such the threshold merely ensured that the consumption of any kind of milk was not discouraged by the tax in accordance with its objective.

(70)

In response to the Commission’s view that a large intake of food products with a content of saturated fat below the 2,3 percent threshold could significantly have contributed to the total intake of saturated fat (recital 107 of the opening decision), Denmark argues that, even though this is true, a large intake of the food products below the 2,3 percent threshold necessarily entails a reduced intake of food products above the threshold. This is in fact a central purpose of the tax.

(71)

Denmark argues that the fact that some undertakings (for example, but not only, producers of milk) would benefit more from the threshold is simply a necessary consequence of setting a threshold. It is a consequence of the basic and guiding principles of the Act.

(72)

Denmark also argues that it is not relevant for the assessment of whether the 2,3 percent threshold should be considered a derogation from the system of reference whether it can be shown that some undertakings (for example producers of milk) have benefitted the most from the threshold or whether the threshold was set in order to exempt milk from the tax. Denmark argues that the only relevant question is whether or not the effects of the 2,3 percent threshold contribute to achieving the objectives of the tax. If so, the threshold is not a derogation from the system of reference, since products below and above the threshold are not in comparable situations.

(73)

In response to the questions posed in the opening decision, Denmark has submitted information about the tax revenues under the Act, showing the impact of the 2,3 percent threshold on the companies in the Danish food industry.

Table 1

Tax revenue by company, area of business and main industry

Table 2

Tax value of non-taxed products: Additional tax payment had the tax been implemented as a general tax on saturated fat in all food products

Table 3

The estimated tax value of the non-taxed dairy products and […] (*2) share hereof

(74)

From Table 1 it can be read that by far the largest tax payer under the Saturated Fat Tax Act was […]. As can be concluded from Table 3, […]was also the main beneficiary from the 2,3 percent threshold, with an estimated tax value of DKK […]. However, Table 2 makes it clear that even though the dairy sector (DKK 140 million) and hence […] was the main beneficiary of the threshold, the meat sector (DKK 9 million) also benefited as did other sectors (DKK 74 million).

(75)

The Danish authorities consider a threshold to be an efficient way to ensure that the tax did not counteract the official dietary guidelines and the best way to obtain the purpose of the Act.

(76)

One theoretical alternative could have been to introduce a subsidy (in the form of food stamps or a general transfer) to compensate families for the increased cost of milk due to the tax to encourage them to live by the dietary guidelines. Denmark argues that it would have been difficult and costly to ensure that such a subsidy was actually spent on whole milk to infants. Besides, the general concern that a tax with no threshold would have decreased the consumption of low-fat versions of the taxable food products could not be addressed adequately by a targeted subsidy or food stamps without insuperable administrative difficulties.

(77)

Denmark submits that the alternative that day-care institutions could have been required to buy and serve a certain amount of milk would neither target the concern of the intake of whole milk by infants, as the vast majority of infants below 1 year are not in day-care institutions in Denmark, nor be sufficient to ensure the entire recommended daily intake. Neither information nor education schemes, to ensure the intake of milk despite the tax, were valid alternatives. Information schemes on milk for infants and young children were already widely used (45).

(78)

In sum, Denmark reiterates that alternative measures to ensure that the tax did not result in economic incentives contrary to the official dietary guidelines would have been far less efficient and most likely more costly than the exemption of whole milk.

(79)

Denmark asserts that the threshold of DKK 50 000 in annual sales of taxable food products was a general measure which applied to domestic producers, importers, intermediaries and distance sellers alike and was effectively open to all economic operators on an equal access basis and is hence not selective in the meaning of Article 107(1) TFEU.

(80)

To enhance the administrative manageability of the tax the threshold was set at a level below which the administrative burdens for both the producers/importers and the Danish authorities would exceed the tax revenue.

(81)

Denmark estimates that an importer or producer of e.g. chicken with annual sales up to DKK 50 000 would have been liable to pay an annual tax of approximately DKK 500 to DKK 1 700 (all depending on the actual quantity and content of saturated fat). As the tax was payable on a monthly basis this would imply levying the amount of DKK 44 to DKK 142 per month.

(82)

Further, Denmark estimates that the costs for small undertakings for calculating and declaring the tax would amount to DKK 100 to DKK 200 per month per taxable food product.

(83)

The estimates given in recitals 81 and 82 assume the highest possible annual sale below the threshold, and only one type of taxable product. Denmark considers it may be safely assumed that a great number of e.g. farm sales have annual sales well below the DKK 50 000 threshold and more than one type of taxable food product.

(84)

Denmark considers that from an administrative point of view it would have been disproportionately bureaucratic and onerous to collect these small amounts.

(85)

Denmark holds that the threshold of DKK 50 000 was an appropriate balance between the revenues generated by the tax and the administrative burdens imposed on both the authorities and businesses.

(86)

Denmark argues that a national tax measure cannot be deemed to be selective solely by virtue of the fact that it distinguishes between domestic and imported goods. A measure that distinguishes between domestic goods and imported goods may be relevant in an assessment of possible breach of Articles 34, 30 and/or 110 TFEU but not for an assessment of selectivity within the meaning of Article 107(1) TFEU, as was made clear in the Banco Santander judgement (46).

(87)

Further, Denmark asserts that the measure does not involve a transfer of State resources. Increased administrative burdens and resulting additional costs for importers do not entail a transfer of State resources within the meaning of Article 107(1) TFEU. Thus, with the coverage tax, Denmark did not forego any tax revenues to which it would have otherwise been legally entitled by this measure. It follows from the case law of the Union Courts that such a measure does not amount to State aid within the meaning of Article 107(1) TFEU (47).

(88)

In recital 121 of the opening decision, the Commission questions the rationale of imposing the tax as far back in the supply chain as possible. In this connection, the Commission notes that it would have been more consistent with the objective of the tax to base the tax on the amount of saturated fat in the final product.

(89)

The Danish authorities disagree with this view. The taxation as early as possible in the supply chain ensures a tax that is targeted as carefully as possible to the primary sources for the intake of saturated fat and to shift consumer habits towards more healthy alternatives. This aim would not have been achieved to the same extent by imposing a tax on the saturated fat in every type of food product.

(90)

Imposing the tax as early as possible and only on primary sources of saturated fat creates an incentive for producers of processed food products to use the healthiest constituents. Contrary to the preliminary view of the Commission, it would go against the objective to encourage better dietary habits if the tax had been imposed also on for example the amount of saturated fat in the vegetables in a vegetable pie. Only the saturated fat in the butter and the cheese in the pie should be taxed in order to encourage the consumption of pies with less butter and cheese and for example more vegetables and low-fat dairy products.

(91)

It was inherent in the objectives of the Act that only the saturated fat from the taxable food products would be levied the tax and consequently, an inherent mechanism necessary for the functioning and effectiveness of the tax system that the importer had to present a producer declaration stating the amount of saturated fat in the taxable food products used in the production. The amount of saturated fat from taxable food products in the final processed food product could not have been easily stated.

(92)

Denmark submits that the Act is based on a system that distinguishes between stockists and consignees. The same distinction between stockists and consignees applies in Council Directive 2008/118/EC (48) that sets out common provisions which apply to all products subject to excise duties under Union law. This implies that the warehouse of a stockist is untaxed and the warehouse of a consignee is taxed. This distinction has derived effects on the structure of the system, which are logical consequences of the basic and guiding principles of the system and they neither constitute discrimination of consignees nor selective advantages to stockists.

(93)

Directive 2008/118/EC defines a stockist as an ‘authorised warehousekeeper’, meaning ‘a natural or legal person authorised by the competent authorities of a Member State, in the course of his business, to produce, process, hold, receive or dispatch excise goods under a duty suspension arrangement in a tax warehouse’; cf. Article 4(1) of that directive. This is similar to Section 4(1), cf. Section 3(1), No 1, of the Act.

(94)

Directive 2008/118/EC further defines a consignee as a ‘registered consignee’, meaning ‘a natural or legal person authorised by the competent authorities of the Member State of destination, in the course of his business and under the conditions fixed by those authorities, to receive excise goods moving under a duty suspension arrangement from another Member State’; cf. Article 4(9) of that directive. This is similar to Section 4(3), cf. Section 3(1), Nos 2 and 3, of the Act, and Section 10(1), cf. Section 9(1), Nos 1 and 2, of the Act.

(95)

Under the Act, as a stockist an undertaking could produce, process, store, receive and import from other countries, and deliver to other stockists taxable food products without the tax being paid (Section 5(1) of the Act).

(96)

Further, importers who only conducted intermediate trading of taxable food products to undertakings registered under the Act (i.e. stockists or consignees) could register as stockists (Section 4(3) of the Act). These intermediaries could receive taxable food products from other stockists and receive and import taxable food products from other countries without the tax being paid (Section 5(2) of the Act). All other importers were to be registered as consignees (Section 4(1) of the Act).

(97)

An undertaking registered as a stockist was to pay tax on the basis of delivered taxable food products (Section 7(1) of the Act). An undertaking registered as consignee was to pay tax on the basis of received/imported taxable food products (Section 7(3) of the Act).

(98)

According to Article 7(1) of Directive 2008/118/EC, the excise duty must be paid once the goods are released for consumption. In general, goods are released for consumption as soon as they leave a duty suspension arrangement or are imported without being placed under a duty suspension arrangement; cf. Article 7(2)(a) and (d) of that directive.

(99)

Goods leave a duty suspension arrangement at the time of receipt by a registered consignee; cf. Article 7(3)(a) and Article 17(1)(a)(ii) of Directive 2008/118/EC. This is similar to Section 7(3) of the Act.

(100)

Goods also leave a duty suspension arrangement at the time of receipt at a place of direct delivery; cf. Article 7(3)(c) and Article 17(2) of Directive 2008/118/EC.

(101)

As it appears, both the system under the Act and the system under Directive 2008/118/EC imply a fundamental difference between stockists and consignees: The warehouse of a stockist is untaxed and the warehouse of a consignee is taxed. This fundamental difference justifies the difference in treatment of stockists and consignees.

(102)

Denmark argues, on that basis, that Measures 5 and 6 must be considered to be justified in light of the tax system’s internal logic, character and management (49), and that those measures are a direct consequence of the basic and guiding principles of the Danish tax system (50).

(103)

The Commission received a total of three submissions from interested parties with comments on the opening decision and the possible aid elements in the Danish Fat Tax.

(104)

In its letter registered on 20 May 2015, Bech-Bruun Law Firm, representing Landbrug & Fødevarer F.M.B.A (hereinafter: L&F), submitted their comments on the opening decision. L&F, or the Danish Agriculture & Food Council as they are called in English, represents the farming and food industry of Denmark including businesses, trade and farmers’ associations.

(105)

L&F argues that the non-taxation of certain products under the Act does not lead to a ‘manifestly arbitrary or biased favouring of certain companies’. The taxed products are the primary sources to saturated fat in Denmark which makes these products different from other products.

(106)

Moreover, according to L&F there is no competition between taxed products (meat, dairy products, all kinds of fat, spreadable blended products and substitutes for the aforementioned) and non-taxed products (eggs, nuts, seeds, bread, corn, potatoes, vegetables, fruit, sugar, candy, supplements, vitamins, additives, fish, food with less than 2,3 percent fat).

(107)

L&F puts forward that the 2,3 percent threshold reflects the threshold laid down in Annex XIII to Council Regulation (EC) No 1234/2007 (51), which regulates how milk for human consumption may be marketed within the Union. Drinking milk can only consist of the categories laid down in that Annex (i.e. raw milk, whole milk, semi-skimmed milk and skimmed-milk). Whole milk (‘sødmælk’) is defined as milk with a fat content of at least 3,5 percent (m/m), which corresponds to 2,3 percent saturated fat. Thus, the Danish legislature chose a threshold which implied equal treatment (non-taxation) of all the three types of drinking milk available on the Danish market, i.e. whole milk (at least 3,5 percent fat), semi-skimmed milk (1,5-1,8 percent fat) and skimmed-milk (less than 0,5 percent fat).

(108)

According to L&F, the 2,3 percent threshold is a balanced and non-discriminatory way of pursuing the objectives of the scheme, whilst preventing competition from being distorted.

(109)

The non-taxation of drinking milk reflected the nutritional recommendations in Denmark at the time, particularly with regard to infants and children aged between 6 months and 3 years (52).

(110)

In its letter registered on 20 April 2015, […] (*3), representing Margarineforeningen (MIFU) submitted their comments on the opening decision. MIFU, the Association of Danish Margarine Manufacturers, is a trade organisation of producers and importers of margarine in Denmark.

(111)

MIFU is of the opinion that the design of the Act runs counter to the alleged healthcare objective of wanting to reduce the overall Danish consumption of saturated fat. According to MIFU, there is no scientific proof that the Act did in fact reduce the overall Danish consumption of saturated fat, nor is there any scientific proof that the Act otherwise improved the general health of the Danish population. Instead MIFU argues that the single economic and behavioural change caused by the Act was the fact that it led to a significant change in terms of the extent to which certain food products contribute to the overall Danish consumption of saturated fat.

(112)

MIFU states that the Act caused a significant decline in the consumption of the taxed food products, such as margarine and butter, which consequently implied a decline in the extent to which these particular taxed food products contributed to the overall Danish consumption of saturated fat. Nonetheless, there is in MIFU’s view no proof that the decline in the consumption of saturated fat from the taxed food products, such as margarine and butter etc., implied that the overall Danish consumption of saturated fat dropped and that the general health of the Danish population was similarly improved. According to MIFU, the decline might instead have been countered by an equivalent increase of the consumption of saturated fat contained in non-taxed or less taxed food products, resulting in an increase of the overall Danish consumption of saturated fat. As a result, the Act may merely have caused a shift in terms of the extent to which certain sources (food products) contribute to the overall Danish consumption of saturated fat.

(113)

In addition to the views and documentation previously submitted to the Commission, MIFU contends that these findings about the effects of the Act are supported by independent scientific studies of the implications of the Act and of other similar European food tax schemes.

(114)

In that regard, MIFU refers in particular to the report ‘Food taxes and their impact on competiveness in the agri-food sector’ of 17 July 2014 (53) (‘ECORYS’ report’), which was based on a study of the implications of European food tax schemes, including the Danish Act. MIFU emphasises that the report concludes that food taxes can generally be found to lead to reduced consumption of the taxed products and also to product reformulation in terms of reducing sugar, salt and fat levels. Food taxes can also result in product substitution, both through an increase in the consumption of taxed products from cheaper brands and non-taxed or less taxed product substitutes. As to the question of whether the changes in consumption result in public health improvements, the report concludes that this is still widely debated and that evidence from academic literature is inconclusive and sometimes contradictory.

(115)

The ECORYS’ report further concludes that food taxes lead to an increase in administrative burdens and may have a negative impact on profitability and employment. They may also impact on the competitiveness of individual firms.

(116)

As to the particular implications of the Act, the ECORYS’ report points towards a significant increase in the prices of the taxable food products and an equivalent significant drop in the consumption of these same taxable food products. The report also identifies a clear shift from taxed products to the less-taxed products as regards olive oil and vegetable and seed oil, but acknowledges that this may be for other reasons than taxation.

(117)

MIFU further refers to a scientific study of the implications of the Act conducted in 2013 (shortly after the Act was abolished), which showed that the level of consumption of taxable products such as butter, butter-blends, margarine and oils dropped by 10-15 percent as a result of the introduction of the Act (54). According to MIFU, this study indicates that the negative impact of the Act on the taxable food products was 3 to 4 times higher than the envisaged effect by the Danish legislator in the commentary to the draft Act.

(118)

MIFU submits that the ECORYS’ report reflects on another study of the implications of the Act based on data for the entire year 2012. It follows from this study that the Act in 2012 caused, inter alia, an increase of the Danish margarine prices by 12 percent implying at the same time a significant decline in the Danish consumption of margarine and the revenue generated by the Danish margarine industry as the purchase and consumption of margarine dropped by 8,2 percent. Butter prices rose by 13 percent and the demand or consumption of butter dropped consequently by 5,5 percent (55). The actual 2012 decline in the margarine consumption, 8,2 percent, was more than double the decline that was foreseen by the Danish legislator in the commentary to the draft Act.

(119)

MIFU takes the position that the conclusions in the report further corroborate the view that the design of the Act was not in line with its purported healthcare objective of reducing the Danish level of consumption of saturated fat.

(120)

MIFU also argues that the levy has been designed in a clearly arbitrary or biased way and that the contested aid measures can therefore not be justified by the intrinsic objective or nature of the general scheme of reference (the Act).

(121)

MIFU claims that the design of the 2,3 percent threshold is an overt matter of industrial policy as its explicit intent and effect was to favour the highly influential Danish dairy sector. According to MIFU, the threshold was designed deliberately with a view to exempt milk from taxation because any taxation of milk would have led to very serious financial repercussions to the Danish dairy sector in light of the economic and competitive situation facing the sector at the time when the Act was adopted.

(122)

MIFU argues that the design of the 2,3 percent threshold under the Act constitutes a prime example of a de facto selective measure that cannot be justified.

(123)

MIFU further argues that the Danish milk sector was the main beneficiary of the de facto exemption by way of the foregone taxes, which in turn mitigated a cost that these undertakings would otherwise have to bear. The taxable food products under the Act incurred a decline in demand and thus in revenue as a result of the Act (the case studies indicate a decline of 5,5 percent-15 percent (56)). By comparison, milk was not taxed and did not experience any drop in demand during the period when the Act was in force. On the contrary, the Danish consumption of drinking milk (57) rose a bit in 2012 by comparison to its 2011 level while achieving a consumption level in 2012 that was virtually the same as the one achieved in 2010 (58).

(124)

MIFU further claims that no ‘expert’ has ever recommended that milk be exempted from taxation under the Act. Even the Danish Government’s ‘Prevention Commission’ recommended that milk be taxed too. According to MIFU, the Danish Veterinary and Food Administration Agency’s dietary advice concerning small children’s consumption of drinking milk did not in any way support the exemption resulting from the 2,3 percent threshold but was only used as an argument to exclude milk from the tax.

(125)

MIFU further argues that it was in any event inconsistent and disproportionate to exempt milk as such from taxation under the Act merely for the alleged purposes of wanting to cater for the possible milk consumption needs of very small children. According to MIFU, this objective could have been achieved in a better and less distortive manner, such as traditional regulation, market-based instruments, industry self-regulation, information and education or tax refunds to households with small children and kindergartens.

(126)

In addition, MIFU argues that the exemption of milk not only favoured the consumption of drinking milk in competition with other substitutable drinking sources but also favoured the use of milk for industrial purposes as an ingredient in other food products over taxable ingredients.

(127)

MIFU thus concludes that the Danish milk sector was the main benefitting sector from the de facto exemption and that no single entity benefitted as much from that exemption as Arla.

(128)

In its letter registered by the Commission on 7 June 2016, […] (*4), representing Arla, submitted their comments on the opening decision. Arla Foods is Europe’s largest dairy group and is cooperatively owned by Danish and Swedish milk producers. The Arla group processes more than 90 % of the Danish milk pool.

(129)

In its letter, Arla, states that the objective of the Saturated Fat Tax was to promote better diets in order to improve the health and mean life expectancy of the population in Denmark. Further, to achieve this overall objective, Arla argues that the objectives of this indirect tax were not to merely tax all food products containing saturated fat, but also to steer the population towards food products that contained lower amounts of saturated fat without lessening the intake of healthy and necessary food products that contained other important nutrients. A diet free of saturated fat is not the objective of the tax as many products which contain saturated fat are also necessary food products.

(130)

Arla goes on to argue that the 2,3 percent threshold for taxation is not selective as food products with less than 2,3 percent saturated fat are in a different factual and legal situation in light of the objectives of the Saturated Fat Tax when compared with products that have higher saturated fat content. Taxing only food products with more than 2,3 percent contribute to the objectives of the Tax.

(131)

Arla continues to explain how the 2,3 percent threshold was based on the official nutritional recommendations with regard to drinking milk at the time of the introduction of the tax. It further lists the benefits of milk. Lastly, Arla argues that there is no competitive relationship between drinking milk and other products.

(132)

The ‘Food taxes and their impact on competitiveness in the agri-food sector’ Report was the outcome of a study, commissioned by the European Commission, Directorate-General for Enterprise and Industry. The study was carried out in 2013-2014 by the European Competitiveness and Sustainable Industrial Policy Consortium (‘ECSIP Consortium’). The Consortium was led by Ecorys Netherlands and consists of Cambridge Econometrics, Danish Technological Institute, Euromonitor, IDEA Consult, IFO Institute and WiiW, as well as a group of specialised subcontractors and individuals. The report looked at the effect non-harmonised taxes have on the competitiveness of the agri-food industry. The study included an analysis of the Danish Saturated Fat Tax. According to the ECORYS’ report (cf. recital 123, there are indications that the Saturated Fat Tax did achieve its primary aim of influencing purchasing patterns and it also helped finance tax cuts elsewhere, but it did so at great administrative costs for the affected enterprises.

(133)

The ECORYS report included the econometric analysis study ‘The Danish tax on saturated fat — short run effects on consumption, substitution patterns and consumer prices of fats’ (59). According to this analysis the introduction of the tax on saturated fat in food products had an effect on the market for the products in question, as the level of consumption of fats dropped by 10-15 percent. However, due to the relatively short data period with the tax being active (nine months, corrected for seasonality effects), interpretation of these findings from a long-run perspective should be done with considerable care. Economic reasoning suggests behavioural adjustments and reductions in fat consumption in the long run, both for consumers, and for the manufacturers, for example in terms of product reformulation towards products with a lower content of saturated fat, etc. In addition to the reduction in the consumption of fats the study found evidence of product substitution, observing that consumers reduced purchases of butter and increased margarine and blend purchases (60). The aspect of substitution effects with regard to products that contain less saturated fat was not directly analysed. The substitution effects may enhance (if the substituted products are healthier) or undermine (if the substituted products for example contain a lot of sugar) the direct incentive effects of the tax (61).

(134)

At the same time, the administrative costs imposed by the tax were found to be significant for the companies. It was estimated that the tax has cost the companies in the retail and wholesale sector app. DKK 200 million (62) (app. EUR 27 million).

(135)

The Danish Chamber of Commerce asked 99 of their members (primarily retail organisations) whether they see signs that the Saturated Fat Tax has directed consumers towards healthier products. Only 12 percent of their members can see this development (63), which suggests that the substitution effects were undermining the incentive effect of the tax, but no academic analysis exists to support this as of now.

(136)

Article 107(1) TFEU lays down that any aid granted by a Member State or through State resources in any form whatsoever which distorts or threatens to distort competition by favouring certain undertakings or the production of certain goods and affects trade between Member States is incompatible with the internal market.

(137)

The qualification of a measure as aid within the meaning of this provision therefore requires the following cumulative conditions to be met: (i) the measure must be imputable to the State and financed through State resources; (ii) it must confer an economic advantage to an undertaking; (iii) that advantage must be selective; and (iv) the measure must distort or threaten to distort competition and affect trade between Member States.

(138)

In the opening decision, the Commission took the preliminary view that Measures 1-3 and Measures 5-6 had all the characteristics of State aid. With regard to Measure 4, the Commission took the view that the measure did not constitute aid in the meaning of Article 107(1) TFEU.

(139)

During the formal investigation procedure, the Commission received observations and comments from Denmark and several interested parties. As regards measure 4, the formal investigation procedure has not revealed any information that would make it necessary to reach a conclusion that would differ from those expressed in the opening decision. Hence, as regards measure 4, the Commission is of the view that that exemption is not selective as it relates to products, which are not in comparable legal and factual situation to the products subject to the tax. Namely, the objectives of the tax (recital 166) related to improving human health and diets, therefore from this perspective, the products not intended for human consumption are not comparable. Therefore, measure 4 does not constitute State aid.

(140)

As regards measures 1-3, 5 and 6, Denmark and interested parties argued that the measures did not have all the characteristics of State aid. Denmark presented information on the legal, administrative and factual elements of the measures which make it necessary to re-examine the presence of State aid with regard to these measures.

(141)

An advantage, within the meaning of Article 107(1) of the TFEU, is any economic benefit, which an undertaking could not have obtained under normal market conditions, that is to say in the absence of State intervention.

(142)

The precise form of the measure is irrelevant in establishing whether it confers an economic advantage to the undertaking. Not only the granting of positive economic advantages is relevant for the notion of State aid, but relief from economic burdens (such as taxes) can also constitute an advantage. The latter is a broad category, which comprises any mitigation of charges normally included in the budget of an undertaking (64). This covers all situations in which economic operators are relieved of the inherent costs of their economic activities. This includes in particular situations where some operators do not have to bear costs that other comparable operators normally have to bear in a given legal order.

(143)

In that regard, the fact that certain products were not subject to the Saturated Fat Tax may involve an advantage for their producers. This means that these products and their producers, which are undertakings providing goods on the market, may benefit from an economic advantage in the form of a relief from the tax burden potentially arising from the tax.

(144)

The question whether the tax burden arising from the Saturated Fat Tax is a ‘normal cost’ that any product has to bear or a ‘charge normally included in the budget of an undertaking’ amounts to assessing the comparability of the products in the light of the objective of the levy. This will be dealt with in the selectivity part of the present decision (see below).

(145)

Only advantages granted directly or indirectly through State resources and based on a decision imputable to the State can constitute State aid within the meaning of Article 107(1) TFEU (65). Waiving revenue, which would otherwise have been paid to the State constitutes a transfer of State resources (66).

(146)

All the measures, as listed in recital 27 of this decision, were the result of a legislative act adopted by the Danish Parliament (Folketinget) and were as such imputable to the Danish State.

(147)

Moreover, as concerns measures 1-3 and 6, a ‘shortfall’ in tax revenue due to exemptions or reductions in taxes granted by the Member State fulfils the affectation of State resources requirement of Article 107(1) TFEU (67).

(148)

However, as regards measure 5, one of the conditions of Article 107(1) TFEU, the presence of State resources, was not fulfilled. It is not sufficient that a Member State imposes or relieves some producers from a burden compared to other producers for a measure to qualify as State aid (68). The advantage that domestic producers of taxable products might have had due to the increased administrative burdens and resulting additional costs for importers/producers of foreign goods due to the coverage tax is merely due to the fact that the tax is due very early in the production chain but did not cause Denmark to forego any tax revenues to which it would have otherwise been legally entitled (cf. recital 87. It merely might have increased the total cost of imported taxed products. Therefore, the Commission agrees with Denmark that increased administrative burdens and resulting additional costs for importers in Measure 5 do not entail a transfer of State resources within the meaning of Article 107(1) TFEU (see recital 87). There is no need to assess whether the other conditions of State aid are fulfilled for this measure.

(149)

The Commission therefore concludes that Measure 5 did not constitute State aid within the meaning of Article 107(1) TFEU.

(150)

A distortion of competition within the meaning of Article 107(1) of the Treaty is generally found to exist when the State grants a financial advantage to an undertaking in a liberalised sector where there is, or could be, competition (69). The Union Courts have also ruled that ‘where State financial aid strengthens the position of an undertaking as compared with other undertakings competing in intra-[Union] trade, the latter must be regarded as affected by the aid’ (70).

(151)

Production of the products at stake in the present case takes place in a liberalised sector. Cross-border trade is important with regard to these products. Any selective advantage to some of these producers would therefore be liable to affect competition and cross-border trade.

(152)

To fall within the scope of Article 107(1) of the Treaty, a measure must ‘favour certain undertakings or the production of certain goods’. Hence, not all measures that favour economic operators fall under the notion of aid but only those that grant an advantage in a selective way to certain undertakings or categories of undertakings or to certain economic sectors.

(153)

In order to establish the selectivity of a measure, the Court of Justice has frequently applied either a two-step or a three-step analysis. In the two-step method, first it has to be determined whether certain undertakings benefit from an advantage compared to other undertakings in a comparable legal and factual situation; and second it is ascertained if the differentiation could be justified because it flows from the nature or overall structure of the system of which it forms part (71).

(154)

By contrast, in the three-step analysis (72),

(155)

In respect of tax schemes, the Court of Justice has established that the selectivity of the measure should, in principle, be assessed by means of the 3-step analysis (73). In reality, the difference between the two approaches is rather academic. It consists of splitting the first step into two separate steps in the three-step approach. Under both approaches, it is necessary to define the appropriate reference framework (74).

(156)

According to the Gibraltar (75) judgment, determining which undertakings are in a comparable factual and legal situation is key to identifying the appropriate reference framework. It follows from the case law that this assessment follows from the purpose of the measure concerned (76) and, more generally, the objectives pursued by the system of which it forms part (77). Seen in this manner, the selectivity analysis consists not only in determining whether a particular measure formally derogates from a particular reference framework, but mainly about determining whether one or certain undertaking(s) enjoy(s) an advantage with regard to other undertakings that are in a comparable factual and legal situation (78).

(157)

For the purposes of the present decision, the selectivity of the measures will be assessed by means of the three-step analysis, as described in recital 154.

(158)

The reference system constitutes the benchmark against which the selectivity of a measure is assessed. It is composed of a consistent set of rules that generally apply — on the basis of objective criteria — to all undertakings falling within its scope as defined by its objectives.

(159)

In the case of taxes, the reference system is normally based on such elements as the tax base, the taxable persons, the taxable event and the tax rates. The same applies to special-purpose (stand-alone) levies, such as levies on certain products or activities having a negative impact on the environment or health, which do not really form part of a wider taxation system. As a result, and provided the boundaries of the levy have not been designed in a clearly arbitrary or biased way (79) — so as to favour certain products or certain activities which are in a comparable situation with regard to the underlying logic of the levy in question — the reference system is, in principle, the levy itself. However, the purpose of imposing a tax or levy may be related to additional objectives (such as health, environment, urban planning, etc.) going beyond the mere fiscal ones, and these equally need to be assessed when establishing the overall reference system for such a measure.

(160)

In accordance with the case law (80), the intrinsic objectives or internal logic of a measure must be defined and the assessment of selectivity (discrimination) must take place only on the basis of such objectives, not any extraneous considerations. As such, health (or similar environmental (81)) objectives cannot justify the exclusion of otherwise selective measures from the scope of Article 107(1) TFEU. However, if it is established that such an objective or even several objectives (such as health, environment, urban planning, etc.) are part of the internal logic (intrinsic objectives) of the tax and the system of which it forms part, then it becomes permissible for a Member State to balance them and ensure their respect at the same time (82). The result of such balancing and simultaneous application of objectives may be that, despite seeming prima facie discrimination from the perspective of one objective, such a differentiation is justified (or the undertakings are not in a comparable position) from the perspective of another, equally applicable objective. However, the invocation of several objectives should not lead to arbitrary discrimination, and the justification of any exclusions from the tax (or lower taxation) must relate to the pre-defined intrinsic objectives of the measure and must be consistently applied and apparent from the method of imposing the tax.

(161)

Whether (i) one finds prima facie discrimination from the perspective of one objective and later finds this justifiable from the perspective of all the objectives applied together; or (ii) simply finds that there is no discrimination because the undertakings are not in a comparable legal and factual situation due to the combined application of several objectives; is ultimately equivalent and comes to the same result, namely lack of selectivity. For the sake of clarity, in the present case the Commission will follow the first of these two ways to conduct the assessment. Accordingly it will take as a starting point the objective of reducing saturated fat intake, as this objective results from the technical features of the tax, namely the subject matter or the chargeable event covered by that tax (83). Any deviations from the logic pertaining to this objective will be recorded as prima facie discrimination (recitals 168-171), which may be justifiable on the basis of the nature and logic of the overall system, including all applicable objectives. But this method of presentation is without prejudice to the existence of all intrinsic objectives followed by the measure and to the order of priority which the Member State has defined for them.

(162)

As regards the measures at stake, in the opening decision, recital 80, the system of reference was defined as ‘In the present case, the system of reference is the Danish tax system, and in particular the rules on taxation of saturated fat in foodstuffs. As described above (recital 17 et seq.), products containing saturated fat are in principle subject to the tax on saturated fat. In accordance with the applicable provisions of the Act, a tax of DKK 16 per kg of saturated fat in the respective foodstuff applies.’ Based on the information provided by the Danish authorities and third party submissions, the Commission needs to adapt this system of reference to include all the intrinsic objectives followed by the measure, see recital 166.

(163)

The Danish authorities have in their comments to the opening decision (cf. recital 56 et seq. of this Decision), objected to the Commission’s definition of the system of reference. Denmark reasons that the system of reference should be defined by the intrinsic objectives pursued by the tax. According to the Danish authorities, the objectives pursued by the Saturated Fat Tax was promoting better diets and improving the health of the Danish population by taxing the primary sources of intake of saturated fat in Denmark, while at the same time taking into account official dietary guidelines.

(164)

The Commission notes that the comments to the Act which were prepared for the first reading of the Act in the Danish parliament only mention that ‘the purpose of the Act is to promote better diets and therefore improve the health of the population’ and that ‘this is achieved by reducing the intake of saturated fat through a tax of DKK 16 per kg of saturated fat in specific products…’.

(165)

While there is no mentioning of primary sources in the comments to the Act, they do refer to dietary recommendations stipulating that fat should be part of a daily diet, and that it is better to consume unsaturated fat than saturated fat. Further, the comments to the Act stated that the tax should cover a significant part of normal foodstuffs to achieve the highest effect on consumption, but at the same time be as simple as possible, and be imposed early in the production chain, in order to reduce the number of undertakings affected by the administrative burden of the tax (registration, declaration, payment).

(166)

In conclusion, the Commission understands the reference framework (determining the subject matter or the chargeable event covered by that tax (84)) as follows: the tax targeted the saturated fat content, being imposed on the weight of saturated fat in the food products listed in Article 1 of the Act, if the saturated fat content in the food product exceeded 2,3 percent of the weight of the product (recital 15). The tax was fixed at the flat rate of DKK 16 per kg of saturated fat in the respective food product (Article 2 of the Act, recital 16). Article 3 of the Act defined operators liable to pay the tax (recital 17). This tax was imposed in order to achieve the following aims (intrinsic objectives):

(167)

The Commission notes that the objectives pursued by the tax are related primarily to public health (including a better diet) and it is permissible in this context to pursue several public health objectives at the same time with the same measure.

(168)

Therefore, the focus of the tax (and its taxable event) is the taxation of the weight of saturated fat in the food products, while preserving other health and dietary objectives (as mentioned at recital 166 above).

(169)

Within the logic of the objective related to the saturated fat, any product containing saturated fat would be in a comparable situation to the products that are subject to the tax (prima facie selectivity). The saturated fat content should therefore be the benchmark for assessing whether there is prima facie discrimination by not taxing certain products. However, as will be examined below, exclusions related to other health and dietary objectives explained above and summarised at recital 166, which were pursued when designing the tax, can be considered justified under the logic of the system, which aims at improving public health. Other bases for a possible justification inherent to the tax system can be, for instance, the need to fight fraud or tax evasion, administrative manageability, the principle of tax neutrality or the need to avoid double taxation (recital 176).

(170)

The tax does not appear to have been designed in a clearly arbitrary or biased way, so as to favour certain products or certain activities which are in a comparable situation with regard to the underlying logic of the tax in question (85). The overall design of the tax is justified, taking into account all of the intrinsic objectives it pursues.

(171)

The Commission considers that the reference system should comprise prima facie all products with saturated fat content — i.e. also those products that are not taxed by the tax — and examines whether the non-taxation of those other products can be justified within the logic of and considering all the objectives forming part of the reference system.

(172)

In a second step, it should be examined whether the non-taxation of certain products constitutes a derogation from the system of reference.

(173)

As mentioned above, within the logic of the objective related to the reduction of saturated fat consumption, any product containing saturated fat would be in a comparable situation to the products that are subject to the tax. Excluding from the scope of the tax products containing saturated fat would therefore be prima facie selective.

(174)

In the light of the objective related to taxing products containing saturated fat, the following products (containing saturated fat) appear to be in comparable factual and legal situation to the products subject to the tax and their non-taxation seems therefore prima facie selective, as established in recital 86 of the opening decision:

(175)

The Commission will re-examine whether the exclusions and exemptions, which were subject to doubts in the opening decision, are justified in the light of the objectives of the Act.

(176)

Where a measure is prima facie selective, it needs to be assessed in the third step if that derogation can be justified by the logic of the system. This is the case where a measure derives directly from the intrinsic basic or guiding principles of the reference system or where it is the result of inherent mechanisms necessary for the functioning and effectiveness of the system (86). In contrast, it is not possible to rely on external policy objectives which are not inherent to the system (87). The basis for a possible justification could, for instance, be the need to fight fraud or tax evasion, the need to take into account specific accounting requirements, administrative manageability, the principle of tax neutrality (88), the progressive nature of income tax and its redistributive purpose, the need to avoid double taxation (89), or the objective of optimising the recovery of fiscal debts. Moreover, the Commission has recognised the pursuance of multiple health objectives as a valid ground for such derogations (90) , (91).

(177)

Measures 1 and 2 concern the exclusion from taxation of any food products not listed in Article 1 of the Act. Measure 2 will be assessed separately below but both measures relate to the application of various criteria (recital 181) to define the taxable products in Article 1 of the Act, called primary sources of saturated fat. The complainant puts forward that certain food products were excluded from the tax whereas others were not, even though they contribute to the consumption of saturated fat to the same extent. The complainant emphasises especially the differentiation between fish, eggs and poultry. Fish and eggs were not taxed but chicken was, even though their contributions were similar to the consumption of saturated fat (cf. recital 52 of the opening decision).

(178)

In recital 30 of the opening decision the Commission listed the main products not taxed because they were not listed in article 1 of the Act. It was the Commission’s preliminary view that taxed and non-taxed products were in a similar factual and legal situation in the light of the intrinsic objective of the system of reference and that the exclusion of products from Article 1 (i)-(vii) of the Act constituted State aid to the producers of all products not covered by those provisions. Denmark was requested to provide additional justification for the exclusions if the Commission was to accept them.

(179)

However, based on the information provided by the Danish authorities and third party submissions, the Commission is of the opinion that there are reasons to depart from its preliminary assessment in the opening decision with regard to Measure 1.

(180)

Although the non-taxation of foodstuffs (intended for human consumption) containing saturated fat is prima facie selective, such exclusions if aiming at improving health and promoting better diets (the intrinsic objectives of the tax), as stated in recital 31, may be justified under the general logic of the system.

(181)

In the present case the Danish authorities established a list of food products to be taxed based on five criteria: (i) following application of the 2,3 percent threshold which is assessed in measure 2; (ii) the consumed quantities of the product; and (iii) the actual share of saturated fat in the product (recital 38); (iv) without counteracting the official dietary guidelines (recital 37); and (v) making sure that the close substitutes of the primary sources of saturated fat are also taxed (recital 44). Therefore, the purpose of the tax was not to eliminate all consumption of saturated fats, but rather to target the primary sources, i.e. products with a high content of saturated fats, consumed relatively frequently, in order to reduce the intake of saturated fats and to induce a shift in consumption towards food products with more health benefits, at the same time without counteracting the official dietary guidelines and making sure that the close substitutes of the primary sources of saturated fat are also taxed.

(182)

As noted in recital 46, food products that contributed to the total intake of saturated fat with 1 percent or less (such as fish, eggs, nuts and seeds) were not considered to be primary sources for the intake of saturated fat. This could be either due to the fact that the consumed quantities of the product in general by the population were low or because the actual share of saturated fat in the product was low. While for fish there were also reasons to exclude from taxation related to the national nutritional guidelines (recital 197), for eggs (92), nuts, and seeds the main reason was their low contribution to the total intake of saturated fat in Denmark.

(183)

As regards fish and eggs, it has to be noted that these products with their meagre one percent contribution (individually) to the consumption of saturated fat, are not primary sources of saturated fat, in accordance with the intention of the Danish legislator. It was therefore logical to exclude fish and eggs from the scope of the tax. Poultry is defined as meat in the nomenclature and according to Directive 2000/13/EC (cf. recital 42). As poultry is considered to be a type of meat, and hence a primary source of saturated fat, it was logical and justified to include it in the scope of Article 1(i) of the Act. On nuts and seeds, recital 46 notes their small contribution to saturated fat intake.

(184)

As described in section 4.1.3 above, the non-taxed products did not qualify as primary sources of the intake of saturated fat in Denmark, nor are they products classified as direct substitutes to taxed products in accordance with Article 1(vii) of the Act.

(185)

It should be recalled that the saturated fat in processed food products — e.g. bread, crisps and candy — was taxed by way of taxing the constituents (cf. recital 41).

(186)

By way of imposing the tax early in the supply chain, an incentive was also created for producers of processed food products to use the healthiest constituents, i.e. those containing less saturated fat (cf. recital 90).

(187)

To achieve the aims of the Act the determination of the taxable products had to take account of the question whether or not the consumption levels of particular products correspond to the national nutritional guidelines for a healthy diet.

(188)

Therefore, a decision to list a product as a taxable product in Article 1 of the Act was based both on the level of saturated fat in each particular product and on considerations whether or not the consumption levels of particular products correspond to the national nutritional guidelines for a healthy diet, as well as the other criteria referred to in recital 181.

(189)

As a general rule, such an approach is consistent with EU policies in relation to the reduction of the intake of saturated fats and the findings of the European Food Safety Authority (EFSA).

(190)

According to EFSA there are public health issues related to diet and some unfavourable trends in food consumption patterns in all Member States (93). In 2005, the Commission asked EFSA to develop a guidance intended for the European population as a whole, on healthy eating advice expressed on food category level that would help to maintain good health through optimal nutrition (food-based dietary guidelines). In its Scientific Opinion of 2009 establishing Food-Based Dietary Guidelines, EFSA concluded that given the differences in health issues, diets and lifestyles of EU Member States, such guidelines ‘have to be based directly upon the diet and disease relationships that are particularly important to the individual country’ (94).

(191)

In that Opinion, EFSA developed general principles and various steps, which can be applied in the development of the food-based dietary guidance at national level. Health aspects are to be taken into account in various steps in developing national guidelines as determined by EFSA, such as ‘Identification of diet-health relationships’ (95), ‘Identification of country-specific diet-related health problems’ (96), identification of ‘Food groups that are sources of nutrients of public health importance’ (97) and ‘Food groups with established relationships to health’ (98).

(192)

In 2010, the EFSA delivered an Opinion on Dietary Reference Values for fats, including saturated fatty acids, polyunsaturated fatty acids, monounsaturated fatty acids, trans fatty acids, and cholesterol and concluded that saturated fat intake ‘should be as low as possible within the context of a nutritionally adequate diet (99). It is explained that ‘[n]utritionally adequate diets are the subject of food-based dietary guidelines and mean a dietary pattern which provides all essential nutrients in adequate amounts as well as energy delivering macronutrients in proportions that are known to maintain health’ (100).

(193)

Independent of their saturated fat content, the intake of certain food categories may reduce the risks of, inter alia, developing cardiovascular diseases. Therefore, reduced intakes of such categories of food, if they are already below recommended intakes, may worsen the health situation, even if the saturated fat intake is reduced.

(194)

Therefore, both issues — health effects of overconsuming certain nutrients, such as saturated fat, and independently, health effects of under-consuming certain food categories — must be taken into account when developing policy measures.

(195)

The official Danish recommendations for a healthy lifestyle issued by the Danish Veterinary and Food Administration which prevailed in Denmark at the time when the Act was in force are summarised in recital 47. They promote the consumption of fruit and vegetables, fish and fish products, potatoes, rice, pasta and wholegrain bread, lean meat and low fat dairy products (even though these products may contain saturated fat). They also advocate eating less sugar and fat (especially from dairy products and meat). It specifies that low fat dairy products and lean meat should be consumed, rather than high fat ones, and it recommends a daily intake of 500 ml of low-fat dairy products as part of a healthy diet (cf. recital 47).

(196)

In its submissions, Denmark has further demonstrated the health benefits of consuming vegetables, fruit, fish and low fat dairy products and has shown that the consumption of these products was below the daily recommended intake (cf. Section 4.1.4). A tax on these products would have further reduced their intake with a harmful effect on the equally applicable objective of the tax related to the dietary guidelines. In the same way, if consumers substitute high saturated fat products with lower saturated fat products such as vegetables and fruit, this is desirable from a health and dietary point of view and therefore in line with the objectives of the tax.

(197)

As regards fish, Denmark has explained (cf. recitals 52 and 53) that the intake of fish by both adults and children was less than half of the recommended amounts. However, fish contains essential fatty acid, vitamin D (in particular fatty fish) iodine, and selenium that are hard to get from other food products. The low intake of fish contributed to a lack of vitamin D in the Danish population which is why it was desirable to increase the intake of fish. Reducing the consumption of fish would have counteracted the Danish official dietary guidelines and accordingly compromise the aims of the Act.

(198)

The complainant argued that there is no proof that a decline in consumption of saturated fat in taxed food products implied that the overall consumption of saturated fat declined and that the general health of the Danish population improved. MIFU argues that there is a risk that substitution from taxed food products to non-taxed food products could lead to an increase in the consumption of saturated fat (cf. recital 112).

(199)

As explained above (recital 196), Denmark took the policy decision to exclude such substitution products from the tax with a view to steering the population to increase the consumption of such products for the very reason that they have a positive effect on health. The evidence provided and set out above supports this approach. The complainant has not submitted sufficient indications or evidence that would show the contrary.

(200)

On the contrary, according to the ECORYS’ report, there are indications that the Saturated Fat Tax did influence purchasing patterns. Further, ECORY’s report puts forward evidence that the level of consumption of fats dropped by 10-15 percent. The report also presents evidence that there was some consumer shifts towards healthier products (cf. recitals 132-135 of this decision). The estimations in economic studies on the effects on consumer behaviour showed that the tax did indeed cause a 10-15 % reduction in the level of fats consumed (101). Further, the analysis presented in the INCENTIVE report finds that the combined effect of excluding food products that were not primary sources for the intake of saturated fat and other food products with a low content of saturated fat (i.e. products covered by Measure 1 and Measure 2) has supported a shift in consumer demand for healthier food products (102).

(201)

In this context it has to be recalled that as set out in Article 168 of the Treaty on the Functioning of the European Union.: ‘Union action shall respect the responsibilities of the Member States for the definition of their health policy […].’ Member States therefore enjoy a wide margin of discretion in the design of their health policy. Unless such policy decisions are evidently wrong or do not take into account clear and unequivocal scientific evidence, the Commission considers that it is the responsibility of the Member States to decide on the appropriateness of setting incentives in one way or the other. Accordingly, Member States may define several intrinsic health and dietary objectives of a tax, they may determine their order of priority by balancing them and they may ensure their respect at the same time (recital 160).

(202)

First, in the present case, the Commission did not detect any arbitrariness or abuse in the definition of the primary sources of saturated fat, which motivated the exclusions under Measure 1. This definition, which follows from the balancing of health and dietary objectives, as crystallised in the criteria described in recital 181 was consistently applied and duly justified by Denmark.

(203)

In this context, the Commission notes that including substitute products for the primary sources of saturated fat in the scope of the tax — although based on the other criteria they would not have qualified as primary sources — was equally justified and did not lead to granting State aid. This widening of the scope of the tax was motivated by the aim of limiting the distortion of competition created by the tax. The justifications related to competition concerns and equal treatment among competing products are in line with the logic of State aid to the extent that it widens the scope of taxable products and reduces the difference between the taxed and the non-taxed products. At the same time, it creates an artificial differentiation between products, which normally should have been considered as non-primary sources and hence not taxed. However, no advantage is granted to the producers of products that were subject to the additional taxation, on the contrary they faced an additional cost. The situation of undertakings which were in a comparable situation (producers of non-primary sources of saturated fat) but which were not taxed even before the application of this fourth criterion, was not altered by the fact that some other products were taxed because they were substitutes. Therefore, this kind of discrimination is not relevant from the perspective of State aid.

(204)

Moreover, this differentiation between otherwise non-primary sources of saturated fat is not an indication that the tax would have been arbitrarily designed as in the Gibraltar case (recital 156). It was not done arbitrarily, but followed a consistent method that could be reconciled with the overall logic of the tax. On the latter aspect, the Commission finds that the taxing of substitutes is once again a measure intended to make the tax more effective and achieve the health and dietary objectives in a more targeted manner, while avoiding an undue interference in the competitive situation of products. The tax normally intended to steer consumers away from primary sources of fat in order to achieve a number of health and dietary objectives. These objectives could be more effectively fulfilled through a targeted tax that is socially accepted and effectively implemented. The widening of the scope to main substitutes of the primary sources of saturated fat was done for that purpose and fits in the overall logic of the system.

(205)

Therefore, the Commission accepts that the criteria for defining the primary sources for the intake of saturated fat were adequate and not arbitrary (recitals 170 and 181).

(206)

Second, the targeting of the tax to only the primary sources was in line with the objectives to encourage a reduction of the intake of saturated fat and to shift the intake from products with a higher content of saturated fat to products with a lower content of saturated fat, while taking into account official dietary guidelines. The Commission agrees that it is a valid objective from a health point of view to focus the tax on primary sources and to exclude from taxation the non-primary sources of saturated fat. While it would have been conceivable to apply the tax also on the saturated fat content of non-primary sources, from the perspective of effectiveness of the tax and in particular the aim of influencing consumer behaviour, it is justifiable to limit the tax to the primary sources. In this way, the price difference between the primary sources and the non-primary ones is increased even more and this enhances the steering of the population away from the consumption of primary sources for the very reason that they have a negative effect on health (cf. recital 31).

(207)

Therefore, the non-taxation of non-primary sources of saturated fat is justified in the light of the overall logic of the system, including all the health and dietary objectives of the tax mentioned in recital 166, in particular the health objective to reduce the consumption of products that are disproportionately harmful for health because of their important saturated fat content and the absence of other health benefits.

(208)

The non-inclusion of certain products in article 1 of the Act, classified as Measure 1 in section 2.1.8.1 of the opening decision and mentioned in section 2.2 of this decision, was therefore justified by the logic of the tax system, i.e. not selective and did not constitute State aid within the meaning of Article 107(1) TFEU.

(209)

Measure 2 exempts all food products, and notably milk, from taxation which have a total saturated fat content of 2,3 percent or less and is one of the five criteria in defining the taxable products in Article 1 of the Act (recital 181). In the opening decision, the Commission took the preliminary view that products containing 2,3 percent or less saturated fat (non-taxed products) are in a similar factual and legal situation in the light of the intrinsic objective of the system of reference as products containing more than 2,3 percent saturated fat (taxed products). The opening decision therefore concluded that the non-taxation of products with a saturated fat content of 2,3 percent or less was prima facie selective.

(210)

In the opening decision, the Commission relied on the recommendations made by the Prevention Commission, i.e. that the tax should apply to all dairy products, including milk. However, as pointed out by the Danish authorities in the course of the investigation procedure, the objective, to reduce the intake of saturated fat, which was the basis for the recommendations of the Prevention Commission, is not entirely the same objective as that of the Act — to promote better diets and improve the health of the Danish population. It is therefore important to reassess Measure 2, taking into account all of the health and dietary objectives of the Act.

(211)

In response to the opening decision, Denmark reiterates that it lies within the Member States’ discretionary powers to decide how to pursue health objectives through tax policy and to decide on minimum thresholds as long as they remain within the limits justifiable by the purpose of the tax.

(212)

As put forward by Denmark (cf. recital 66), a threshold which exempts all products with a saturated fat content below a certain threshold from taxation, amplifies the behavioural incentive to choose low-fat products and is hence in line with the official dietary recommendations as described in recital 47 of this Decision. The amplification effect holds true for all products but even more so for dairy products and meat as it is mainly those products which have alternatives above and below the threshold. In the case of dairy products, a threshold also promoted the explicit dietary recommendation to consume 500 ml of low-fat dairy products per day by way of increasing the price difference between low and high fat products.

(213)

The official dietary recommendations at the time the Act was in force also recommended that young children below the age of one year should drink full fat milk (milk with 3,5 percent total fat content, or 2,3 percent saturated fat content) (103). The choice of the 2,3 percent threshold avoided a price increase on milk and facilitated fulfilling this dietary recommendation. One could have considered a differentiated approach according to which full fat milk for young children on the one hand would have been exempted from the tax whereas the same product intended for the consumption by older children and adults on the other would have been covered. However, such an approach would have been very difficult if not impossible to manage and would therefore have been unjustified. Therefore, a uniform treatment of full fat milk seems justified.

(214)

The complainant has put into question the efficiency of the threshold in ensuring that the act did not counteract the dietary recommendations that children should drink milk, and small children should drink full milk.

(215)

However, in its submissions Denmark has argued that fixing a threshold was the most efficient and the least costly way to ensure that the tax did not counteract the official dietary guidelines. Alternative measures to ensure that the tax did not result in economic incentives contrary to the official dietary guidelines would have been far less efficient and administratively more costly than the exemption of whole milk (cf. recitals 75-78).

(216)

The complainant has put forward that even though the 2,3 percent threshold could be seen as a general measure, it is selective in its effects as the main beneficiaries are dairy producers and processors, and then specifically Arla as being the main processor of dairy products in Denmark (cf. recitals 139 to 146).

(217)

Denmark has argued that the 2,3 percent threshold reflects the basic and guiding principles of the Act. The fact that it benefitted some undertakings more than others was simply a necessary consequence of setting a threshold (cf. recital 72).

(218)

Denmark has submitted figures on the main tax payers and also on the main beneficiaries of the 2,3 percent threshold (cf. recital 74). These figures show that […] (*5) was the main tax payer under the Saturated Fat Tax Act in respect of products which were subject to the tax (DKK […] million), while it also had the main share of the estimated tax value of the non-taxed dairy products under the 2,3 percent threshold (DKK […] million). However, even though the figures show that the threshold concerned to a large extent the dairy sector (DKK 140 million), and hence […], they also show that products from other sectors fell within its scope (DKK 83 million).

(219)

As regards measure 2, which was one of the five criteria (recital 181) for determining the taxable products called the primary sources of saturated fat, the Commission draws the same conclusions as in recitals 202-207 above. First, it did not detect any arbitrariness or abuse in the definition of the 2,3 percent threshold. The definition of this threshold, which follows from the balancing of health and dietary objectives, was consistently applied and duly justified by Denmark.

(220)

Second, the targeting of the tax to only the primary sources also through a threshold was in line with the objectives to encourage a reduction of the intake of saturated fat and to shift the intake from products with a higher content of saturated fat to products with a lower content of saturated fat, while taking into account official dietary guidelines. The Commission agrees that it is a valid objective from a health point of view to focus the tax on primary sources and to exclude from taxation the non-primary sources of saturated fat. While it would have been conceivable to apply the tax also on the saturated fat content of non-primary sources (below the threshold), from the perspective of effectiveness of the tax and in particular the aim of influencing consumer behaviour, it is justifiable to limit the tax to the primary sources. In this way, the price difference between the primary sources and the non-primary ones is increased even more and this enhances the steering of the population away from the consumption of primary sources for the very reason that they have a negative effect on health.

(221)

Therefore, the non-taxation of non-primary sources of saturated fat (as also defined through the threshold) is justified in the light of the overall logic of the system, including all the health and dietary objectives of the tax mentioned in recital 166, in particular the health objective to reduce the consumption of products that are disproportionately harmful for health because of their important saturated fat content and the absence of other health benefits.

(222)

The criterion relating to the non-taxation of products containing 2,3 % percent or less of saturated fat is therefore consistent with the objectives of the tax as mentioned in recitals 31, 32 and 166 and is justified by the logic of the tax system. The non-taxation of these products, classified as Measure 2 in section 2.1.8.2 of the opening decision and mentioned in section 2.2 of this decision, is therefore not selective and does not constitute State aid within the meaning of Article 107(1) TFEU.

(223)

In the opening decision it was left open whether the exemption from taxation in accordance with the Act for businesses with an annual sale of taxable food products of DKK 50 000 or less was selective and to be considered aid in accordance with Article 107(1) TFEU. As explained in the opening decision, the exemption level was set at the same one as that in the Danish VAT Act, which states that businesses with an annual sale of taxable food products of DKK 50 000 or less do not have to register in accordance with the VAT Act and do not have to pay VAT. According to Denmark’s submission in response to the opening decision, if these companies do not have to register for the purpose of paying VAT, they should not need to register for the purpose of paying tax in accordance with the Saturated Fat Tax Act.

(224)

Denmark argues that the threshold of DKK 50 000 in annual sales of taxable food products was a general measure which applied to domestic producers, importers, intermediaries and distance sellers alike and was effectively open to all economic operators on an equal basis.

(225)

Further, in the light of the health objective of the tax and the limited impact on health of the limited quantities that were not taxed due to the DKK 50 000 threshold, the exemption seems justified by the nature of the system of reference.

(226)

The Commission takes note, that the amounts of taxes not levied due to the threshold may be small (cf. recital 81), and accepts that the collection of such small amounts would have implied a disproportionate administrative burden. As explained by Denmark, undertakings with a turnover of DKK 50 000 or less do not have to register for VAT purposes, because of the administrative burden for the undertakings concerned and for the administration. It would be disproportionate in view of the amount of VAT that would otherwise be collected. The same holds true for the case at stake here. The tax to be paid by undertakings with a turnover of taxable food products of DKK 50 000 or less would be too low to justify the administrative burden connected to its collection. Accordingly the Commission did not detect any arbitrariness or abuse in the definition of the DKK 50 000 sales threshold on taxpayers. The definition of this threshold, which follows from the balancing of health and dietary objectives together with effectiveness of the tax and administrative efficiency, was consistently applied and duly justified by Denmark.

(227)

Next, the Commission agrees that it is a valid objective from a health point of view to focus the tax also as regards the scope of taxable persons to larger undertakings and to exclude from taxation the very small undertakings with an annual sale of taxable food products of DKK 50 000 or less. While it would have been conceivable to apply the tax also on taxpayers with DKK 50 000 sales or less (below the threshold), from the perspective of effectiveness of the tax and administrative efficiency, it is justifiable to limit the taxpayers to undertakings above the threshold. In this way, the price difference between the primary sources and the non-primary ones is increased even more and this enhances the steering of the population away from the consumption of primary sources for the very reason that they have a negative effect on health.

(228)

The exemption from paying the tax for businesses whose annual sales of taxable food products was total DKK 50 000 or less, classified as Measure 3 in section 2.1.8.3 of the opening decision, is therefore not selective and does not constitute State aid within the meaning of Article 107(1) TFEU.

(229)

This measure, involving 4 sub-measures, related to the fact that the Act differentiated between domestic producers and importers of the same or similar food producers with regard to certain aspects of the tax deferral and tax deduction regime. In essence, the measure boils down to different treatment between stockists (recital 233) and consignees, because the measure did not apply to the importers registered as consignees. The opening decision recital 135 found that the measure appears to discriminate between imported products and national products, and might therefore be selective and also in contravention of the provisions laid down by Article 110 TFEU.

(230)

Based on the information provided by the Danish authorities in response to the opening decision, the Commission is of the opinion that there is reason to depart from its original assessment in the opening decision with regard to Measure 6.

(231)

In recital 74 of the opening decision, based on the idea that the Act envisaged certain preferential regimes for domestic food producers as compared to importers, the Commission took the view that the described system appeared to constitute an advantage for domestic producers within the meaning of Article 107(1) TFEU as they were put into a more favourable financial position compared to undertakings importing the same type of products.

(232)

That concern was strengthened by the fact that the Danish legislator had conceded in the accompanying preparatory legislative documents (104) that the aim and effect of the preferential regime was to enable the domestic producers to gain a cash flow advantage (105). As confirmed by the Danish authorities, these deferral and tax deduction options were not available to undertakings importing the same type of food products and selling them on the same Danish markets.

(233)

However, as argued by the Danish authorities in their response to the opening decision, the system under the Act, based on the system under Directive 2008/118/EC on excise duties, implies a fundamental difference between an authorised warehousekeeper (106) (stockist in the Act) and a registered consignee (107) (consignee in the Act): The warehouse of a stockist is untaxed and the warehouse of a consignee is taxed (cf. recital 92(ff)). This is compliant with the case law of the Union Courts (108) which stipulates that a tax should be levied at the same stage in the supply chain, i.e. at the time of delivery by domestic producers and at the time of import by importers.

(234)

In accordance with the rules of Directive 2008/118/EC and those of the Act, an undertaking is registered according to its role in the supply chain and not depending on the origin of the products it sells. A domestic producer is usually registered as an authorised warehousekeeper/a stockist and an importer is usually registered as a registered consignee/consignee. However, the Act also made it possible for importers only conducting intermediate trading to register as a stockist and this way keep their warehouses untaxed (recital 42 of the opening decision). Retailers of domestic and/or imported taxable products conducting direct sales all have their warehouse taxed and are in the same position when it comes to the taxation of taxable products. At the same time, as described in recital 41 of the opening decision, Danish producers could register as a stockist for the part of the production, which they planned to export.

(235)

According to Denmark, the fundamental difference between stockists and consignees justifies the difference in treatment under the Act and Measure 6 is not, as claimed by the complainant, a way to discriminate between domestic and foreign producers. Nor does it discriminate in any other way, it merely ensures that taxes are levied in accordance with the case law of the Union Courts (109).

(236)

The Commission agrees that Measure 6 does not discriminate between domestic and foreign producers or undertakings in the same legal and factual situation. Stockists and consignees are in a different position from the perspective of their role in the supply chain, which in turn is the principle followed by the Act and also Directive 2008/118/EC. The way undertakings are taxed under the Act depends on their role in the supply chain. Which undertaking belongs to which role is decided on the basis of objective criteria, over which the undertaking has a certain level of control. An importer can for instance decide on its own if it wishes to only conduct intermediate trading and that way be able to register as a stockist and keep their warehouse untaxed. However, if it wishes to engage in direct sales, it will have to be registered as a consignee and will have its warehouse taxed (just as the domestic producers engaging in direct sales).

(237)

The Commission concludes that in case of Measure 6, the differentiation in time of levying the tax, does not give rise to discrimination between domestic and foreign producers or undertakings in the same legal and factual situation. Therefore, Measure 6 is not selective in the meaning of Article 107(1) TFEU. For the same reason and because the difference between a stockist and a registered consignee under measure 6 corresponds to the system under Directive 2008/118/EC on excise duties, it is not in breach of Article 110 TFEU. This is because it does not impose a higher tax on imports than on domestic goods and is not an indirect protection against products from other Member States.

(238)

Concerning Measure 5, one of the conditions of Article 107(1) TFEU, the presence of State resources, was not fulfilled. The Commission therefore concludes that Measure 5 did not constitute State aid within the meaning of Article 107(1) TFEU.

(239)

Concerning Measures 1, 2, 3 and 6, one of the conditions of Article 107(1) TFEU, the presence of selectivity, was not fulfilled. The Commission therefore concludes that Measures 1, 2, 3 and 6 did not constitute State aid within the meaning of Article 107(1) TFEU,