Table of contents

1.Introduction: political and legal context9

2.Problem definition13

2.1.What is the problem?13

2.2.What are the problem drivers?20

2.3.How would the problem evolve23

3.Why should the EU act?23

3.1.Legal basis23

3.2.Subsidiarity: Necessity of EU action24

3.3.Subsidiarity: Added value of EU action25

4.Objectives: What is to be achieved?26

4.1.General objectives26

4.2.Specific objectives26

5.What are the available policy options?28

5.1.What is the baseline from which options are assessed?28

5.2.Description of the policy options32

5.3.Options discarded at an early stage39

6.What are the impacts of the policy options?40

6.1.Social impacts40

6.1.1.Impacts on health40

6.1.2.Impacts on health inequalities44

6.2.Economic impacts45

6.2.1.Impacts on direct costs for businesses and public authorities45

6.2.2.Impacts on consumers47

6.2.3.Internal Market impacts49

6.2.4.Competitiveness and trade impacts50

6.2.5.Impacts on SMEs51

6.3.Environmental impacts53

6.4.Impacts of combined options55

7.How do the options compare?55

7.1.General objective 1: Ensuring a high level of health protection for EU and Specific objective 1: Reduce intake of industrial trans fats in the entire EU for all population groups56

7.1.1.Direct health impacts56

7.1.2.Direct and indirect economic impacts of changes in health status56

7.2.General objective 2: Contribute to the effective functioning of the Internal Market for foods that could contain industrial trans fats and Specific objective 2: Ensure that the same rules/conditions apply in the EU to the manufacturing and placing on the market of foods that could contain industrial trans fats, so as to ensure legal certainty of EU food business operators within and outside the EU58

7.3.General objective 3: Contribution to reducing health inequalities, one of the objectives of Europe 202060

7.4.Effectiveness60

7.5.Efficiency (balance of costs and benefits)62

7.6.Coherence with other EU policy objectives65

7.7.Proportionality65

7.8.Specific tests: SME test66

8.Preferred option68

9.How will actual impacts be monitored and evaluated?71

ANNEX 1: Procedural information72

1.Lead dg, decide planning72

2.Organisation and timing73

3.Consultation of the RSB74

4.Evidence, sources and quality77

ANNEX 2: Stakeholder consultation78

1.Introduction78

2.Stakeholder groups covered by the consultation activities78

3.Consultation activities already carried out before the launch of the IA79

4.Outline of the Consultation strategy for the IA on an initiative to limit industrial trans fats intakes in the EU80

5.Results of the Consultation activities for the IA on an initiative to limit industrial trans fats intakes in the EU81

6.Conclusion87

ANNEX 3: Who is affected and how?89

1.Practical implications of the initiative89

2.Summary of costs and benefits91

ANNEX 4: Analytical methods93

1.Study methodology development93

2.Data collection and review93

3.Screening of impacts and assessment of significance98

4.Analysis of impacts98

5.Validation consultation103

6.Strengths and limitations of the method105

7.Discussion of information gaps and uncertainties106

ANNEX 5: Trans fats – a general presentation108

ANNEX 6: Trans fats consumption and its negative impact on health and intake recommendations111

ANNEX 7: Health effects of ruminant versus industrial trans fats and the potential to limit the associated health problem by addressing their intake116

ANNEX 8: Current status of EU and national measures addressing the trans fats problem and consumer knowledge regarding trans fats118

ANNEX 9: Additional information on trans fats intakes in the population and presence in foods123

ANNEX 10: Discussion of the baseline scenario139

ANNEX 11: Intervention logic for the different options143

ANNEX 12: Impacts screening152

ANNEX 13: Assumptions for the health impacts assessment162

ANNEX 14: Additional information on the Sensitivity Analysis168

1.Impact on health care costs (direct and indirect)168

2.Impact on disability-adjusted life years168

ANNEX 15: Impacts on health inequalities and details on appraisal of general objective 3: contribution to reducing health inequalities, one of the objectives of Europe 2020169

ANNEX 16: Impacts on administrative costs for businesses, understanding the requirements and verify compliance175

ANNEX 17: Impacts on compliance costs for businesses179

1.Compliance costs – product testing179

2.Costs of reformulating products181

3.Costs of ingredients187

4.Costs of labelling189

ANNEX 18: Administrative cost for public authorities192

1.Costs of establishing the policy192

2.Costs of consumer information campaigns193

3.Costs of monitoring and enforcement194

ANNEX 19: Assumptions for the impact assessment on consumer prices197

ANNEX 20: Evidence collected by the external contractor concerning the assumptions for the impact assessment on product attributes200

ANNEX 21: Expected impact of each option on the Internal Market201

ANNEX 22: Details on the expected impact of each option on competitiveness and international trade205

ANNEX 23: Evidence on the impacts on SMEs and expected impact of each option on SMEs206

ANNEX 24: Evidence on substitutes for partly hydrogenated oils, environmental impacts of palm oil as well as environmental impacts of alternatives; expected impact of each option on the environment207

ANNEX 25: Impacts of combined options212

1.Combining mandatory labelling with legislation (2 + 1b or 2 + 3b)212

2.Combining mandatory labelling with voluntary agreement (2 + 1a or 2 + 3a)213

ANNEX 26: Further details for appraisal of General objective 1 specific objective 1215

1.Additional details for section 7.1.1 on direct health impacts215

2.Additional details for section 7.1.2 direct and indirect economic impacts of changes in health status216

3.Further details for appraisal of specific objective 1: Reduce intake of industrial trans fats in the entire EU for all population groups216

ANNEX 27: Further details for appraisal of specific objective 2: Ensure that the same rules/conditions apply in the EU to the manufacturing and placing on the market of foods that could contain industrial trans fats, so as to ensure legal certainty of EU food business operators within and outside the EU219

ANNEX 28: Ex ante analyses in the US and Canada on Evidence on legislation to ban partly hydrogenated oils221

ANNEX 29: Consolidated information collected through interviews with EU level business associations by223

ANNEX 30: Aggregated evidence for each type of impact: a list of indicators; the description of the evidence obtained, either quantitative or qualitative; and sources for that evidence304

ANNEX 31: Validation consultation by ICF, survey instrument439

ANNEX 32: ICF Country profiles461

ANNEX 33: Questionnaire for the Open Public Consultation466

List of abbreviations

CAOBISCO        Association of Chocolate, Biscuit and Confectionery Industries of
           the European Union

CI            Confidence Interval

EFSA            European Food Safety Authority

FEDIOL        EU vegetable oil and protein meal industry association

HOTREC        Association of hotels, restaurants and cafés in Europe

IA            Impact Assessment

IIA             Inception IA

IMACE        European Margarine Association

ISG            Inter-services Steering Group

JRC            the Joint Research Centre of the European Commission

NGO            Non-governmental Organisation

OPC             (On-line) Open Public Consultation (carried out for this IA)

RR            Relative Risk

RSPO            Roundtable on Sustainable Palm Oil

SKU            Stock Keeping Unit

SMEs             Micro, Small and Medium-sized Enterprises

SWD            Commission Staff Working Document

TFEU            Treaty on the Functioning of the European Union

WHO            World Health Organisation

Glossary

1.1.

1.Introduction: political and legal context

Trans fats (also called 'trans fatty acids' and sometimes abbreviated as TFAs) are a particular type of unsaturated fatty acids that are present in some foods 1 as natural trans fats in ruminant (dairy and meat) products 2 or as industrially manufactured trans fats. Industrial trans fats, in the form of partial hydrogenated oils, are added to improve stability or texture or for other technological reasons, in a variety of products including pastries and chocolates. One of the common substitution fats with similar technological and cost advantages is palm oil. Trans fats are not synthesised by the human body and are not required in the human diet.

There is scientific consensus that trans fats intake has a negative effect on human health: more specifically, consumption of trans fats has a negative impact on blood cholesterol levels and increases the risk of coronary heart disease 3 more than any other macronutrient compared on a per-calorie basis; the risk of dying from heart disease is 20-32 % higher when consuming 2 % of the daily energy intake from trans fats instead of consuming the same energy amount from carbohydrates, saturated fatty acids, cis monounsaturated fatty acids and cis polyunsaturated fatty acids. 4  

The European Food Safety Authority recommends that trans fats intakes should be 'as low as is possible within the context of a nutritionally adequate diet'. 5 The World Health Organisation recommends that less than 1 % of dietary energy intake should come from consuming trans fats 6 (which equates to maximum 2,2 g of trans fats per day for a person requiring 2000 kilocalories). 7 Currently, in total 7 Member States have introduced legislation regarding intakes of industrial trans fats. In particular, 6 Member States (Denmark, Austria, Hungary, Latvia, Lithuania and Slovenia) have set legal limits and one (Romania) has recently notified a draft legal measure. The legal limit of maximum 2 % of industrially produced trans fats in foods introduced in several Member States is in line both with the intake recommendations of the European Food Safety Authority and of WHO: typical intakes of total fat in European countries are reported to be at a maximum of 48 % of the daily energy intake (95th percentile). 8 Provided that all foods contain trans fats at 2 % in a very unlikely, extreme scenario, intake levels would be at 0.96 % of energy intake, below the WHO recommendation. Assuming a 2000 kilocalorie diet, 0.96 % of daily energy intake equates to a maximum of 2.1 g of industrial trans fats intake per day. Empirical evidence from Denmark, where a 2 % legal limit per 100 g fat content applies, suggests that (in 2014) the average industrial trans fats intake was 0.009 % of energy intake. 9 This very low level could be considered to be in line with the recommendation of the European Food Safety Authority ('as low as possible'). In this context it is noteworthy that some small amounts of trans fats are generated during the normal processing and production of foods. Ruminant trans fats sources typically contribute between 0.3 and 0.8 % of the daily energy intake depending on dietary habits across Europe. 10 Thus, even the combination of ruminant and industrial trans fats typically amount to 0.309 to 0.809 % of daily energy intake.

The issue of trans fats was intensively debated during the negotiations that preceded the adoption of Regulation (EU) No 1169/2011 on food information to consumers 11 . This Regulation does not include trans fats in the list of mandatory nutrition declaration since the co-legislator was not convinced that the introduction of trans fats amounts on food labels would consistently enable consumers to identify the healthier choice. In addition, the efficiency of such measure was questioned since it would not apply to non-pre-packed foods, all of which may contain high levels of industrial trans fats. Finally, trans fats labelling would not distinguish between ruminant and industrial trans fats. Regulation (EU) No 1169/2011 also prohibits operators from declaring the trans fats content of foods on nutrition labels on a voluntary basis. It was considered that this possibility would be used as a marketing tool by some operators only and could lead to consumers' confusion. Therefore, the co-legislator agreed that instead of looking only into the labelling aspect, the Commission should assess the impacts of all means to enable consumers to make healthier choices, including restrictions on the use of trans fats. A report was requested by Article 30(7) of Regulation (EU) No 1169/2011 of the European Parliament and the Council on the provision of food information to consumers.

In its 2015 report 12 , the Commission noted that average trans fats intake in the EU is below nationally and internationally recommended levels, however, this conclusion is not valid for all population groups. Food products with high industrial trans fats content remain available on the EU market, thus, reducing industrial trans fats intakes entails public health gains. The report concluded that a legal limit for industrial trans fats would be the most effective measure in terms of public health, consumer protection and compatibility with the Internal Market but that further investigation is required.

Numerous calls for a reduction of trans fats intakes in the EU have emerged from the agenda of the European Parliament and the Council, individual Member States, and stakeholders. Member States' concerns on industrial trans fats had been voiced in the context of the High Level Group on nutrition and physical activity 13 where 22 Member States indicated industrial trans fats as one of the priorities with respect to reformulation or nutrient policy. 14 Health EU Ministers exchanged views on trans fats at two informal Council meetings: in April 2015 in Riga, a large majority of intervening delegations expressed support to the necessity of reducing industrial trans fats levels in food products. 15 In September 2015 in Luxembourg, Member States discussed possible solutions to reduce industrial trans fats levels in foods. Some delegations called for legal limits to industrial trans fats presence in foods at EU level, while others favoured self-regulatory approaches based on product reformulation. 16  

Council Conclusions of 2014 and of 2016 noted with concern 17 'the high intake of …trans fatty acids…' and 18 'The prevalence of overweight, obesity and other diet-related non-communicable diseases in the European population is too high and is still rising. This has a negative impact on life expectancy, reducing Union citizens' quality of life and affecting society, for example by threatening the availability of a healthy and sustainable workforce and inducing high healthcare costs which may affect the sustainability of the healthcare systems. It thus also imposes an economic burden on the Union and its Member States. (…) Nutrition plays an important role in this context, alongside other lifestyle-related matters: (…). In some Member States, people are still exposed to high amounts of trans fatty acids'. 19

The European Parliament adopted on 26 October 2016 a resolution calling on the Commission to propose legislation setting a limit on industrial trans fats within two years and to carry out an impact assessment evaluating impacts on operators and consumers. 20

Following the adoption of the Commission report, a considerable number of external stakeholders, such as associations representing producers and consumer representatives have expressed a keen interest in this issue. 21   22 All stakeholders that intervened in the debate on trans fats so far have welcomed the Commission's report and/or supported an EU initiative to set legal limits to industrial trans fats in foods, both on the consumers' side 23 and on the industry's side. 24

In this context, of particular note is the joint letter21 addressed on 15 October 2015 to the European Commission by four major food manufacturers, together with leading consumers' and health NGOs and the Standing Committee of European Doctors. Also of note are the number of reformulation commitments to lower the content of industrial trans fats in foods made in the past years by food manufacturers in the EU Platform for Diet, Physical Activity and Health. 25 The positions of industry stakeholders (also well summarised in a statement by Food Drink Europe of 19 November 2015) 26 indicate that the industrial trans fats content of foods can effectively be lowered without disproportionate cost 27 , that an EU initiative would benefit consumers and the industry by setting a level playing field in the Internal Market, and that particular support might be needed for SMEs.

Stakeholders 28   29 also broadly supported national initiatives that set limits to the presence of industrial trans fats in foods.

At the global level, calls for reduction of trans fats intakes led to the REPLACE initiative ('trans fat free by 2023') of WHO in May 2018. 30 WHO recommends to ‘legislate or enact regulatory actions to eliminate industrially-produced trans fats’.

The objective of this impact assessment is to enable an informed decision on how to deal with trans fats, taking into account the potential economic, social and environmental impacts of different policy options, including implementing the option of a legal limit for industrial trans fats. In this context, the factual situation, as regards the issue of excessive trans fats intakes in the EU and its underlying causes and the policy implications of available alternative approaches to setting a legal limit, i.e. mandatory labelling of trans fats and voluntary approaches to food reformulation, are examined. Besides the public health dimension and ensuring a sound basis for consumer choice, the impact assessment also examines the consequences of the policy options available for the businesses, including SMEs and the Single Market.

In addition to the report adopted by the Commission in 2015 on trans fats 31 , the impact assessment takes into account various studies on trans fats at the European level  32   33   34 and internationally 35 , investigating the impacts of trans fats and the potential effects of alternative policy options to limit their use. These build on analyses by the Joint Research Centre of the Commission (JRC) 36 , scientific opinions of the European Food Safety Authority5  37 , international reports by the World Health Organization6  38 and academic studies. In 2017, the European Commission commissioned an external study by the contractor ICF to support this IA. 39

2.Problem definition 

2.1.What is the problem? 

Trans fats are an important risk factor for the development of coronary heart disease5 37, the single leading cause of mortality in the EU. 40 Cardio vascular disease comprises a range of diseases that affect the heart, including heart failure (which can be caused by coronary heart disease, among other factors), arrhythmia (abnormal heart beat) and heart valve problems, and imposes substantial health burdens in the EU. It is estimated that 49 million people live with cardio vascular disease and that the condition imposes costs of more than €200 billion each year in the EU. 41 The European Food Safety Authority and the World Health Organization recommend that their consumption is limited or minimised.5 6  42 Industrial trans fats intakes are particularly high among consumers with lower income, who are also the most at risk of coronary heart disease 43 and intakes continue to contribute to the absolute health and economic disease burdens of cardio vascular disease.

The precise contribution of trans fats intake to health risks and associated economic problems are difficult to assess for the entire EU due to limited data available for trans fats intakes in the entire EU. There is empirical evidence that the introduction of a legal limit for industrial trans fats reduced deaths caused by cardiovascular disease. 44   45 Over 3 years following the introduction of the legal limit, mortality attributable to cardiovascular disease decreased on average by about 14.2 deaths per 100,000 people per year relative to a synthetic control group, meaning that the Danish limit on industrial trans fats saves around 700 people a year in Denmark. 46 Further evidence of the effectiveness of legal measures is available from outside the EU: in Argentina, near elimination of industrially produced trans fats from food is estimated to be associated with an annual 1,3 to 6,3 % reduction in coronary heart disease events 47 . In New York, people living in counties in New York State with restrictions on industrial trans fats in food had a 7,8 % greater decrease in hospital admissions for heart attacks between 2007 and 2013 than people in counties without restrictions 48 .

How widespread are trans fats in the EU?

There is limited availability of comparable/EU-level data on the intakes of trans fats in the different population groups or on the presence of trans fats in foods in the different Member States. Evidence from a number of countries indicates that the intake of trans fats in the EU has decreased considerably over recent years 49 but that the situation is not homogeneous for all products consumed by all population groups in all EU Member States. Studies summarised by the JRC in its 2014 report concluded that 50 :

·Average daily trans fats intakes for the overall EU population are below 1 % of daily energy intake 51 . Yet some population groups have (or are at risk of having) higher intakes.

Examples of such sub-populations are low-income citizens (British male and female participants of the Low Income Diet and Nutrition Survey 52 where all age groups had intake levels above 1 % of energy intake, ranging from 1.2 to 1.4 % of energy intake) or male or female university students aged 18 to 30 years (data from Croatia, intake levels ranging from 1.1 to 1.2 % of energy intake 53 ). Also, according to surveys collected by the JRC, Swedish boys aged 8 and 11 years exceeded the WHO recommendation (1 % of energy intake), as well as Spanish males and females aged 18 to 30 years (1.05 % of energy intake) and French females over 55 years of age (1 % of energy intake) and between 3 to 10 years (1.02 % of energy intake). 54 As calculated by the JRC, up to 25 % of surveyed individuals aged 20-30 years have trans fats intakes above 1 % of daily energy intake. Annex 9 provides more details. Latest information collected during the OPC confirm this assessment. 55

·Most of the analysed food products contain trans fats at amounts below 2 % of the total fat content of the food and 77 % of these contain trans fats at amounts below 0.5 % of the total fat content of the food. However, there are still products in the European food market with high levels of industrial trans fats (e.g. biscuits or popcorn with industrial trans fats values in the order of 40-50 % of the total fat content of the food). While most of the analysed products are pre-packed products, there are also several reported cases of non pre-packed foods with trans fats levels above 2% of the total fat content in food. Examples of products found to contain trans fats in considerable amounts in Member States, generally of industrial origin, are frying fat also for industrial use, stick margarines, margarine used to produce pastry products, bakery products, biscuits, wafers, confectionary products including those with cocoa coatings such as covered puffed rice, soups and sauces. 56 Further recent studies about trans fats content in food in the EU were published after the finalisation of the JRC36 work:

oA study 57 focused on the market for pastries, confectionery, and potato products in Poland in the period 2009-2010 and reported a great diversity of trans fats content (0.1 % to 24.8 % of total fat content). Wafers were characterized by the highest average content of trans fats in the group of pastries (1.94 % of total fat content);

oA research in Germany 58 in 2017 quoting data from the Federal Office for Consumer Protection and Food Safety noted that in the period 2014 to 2017 the mean trans fats content in certain sampled fried bakery products was higher than 10 % of total fat content, sometimes even more than 30 %;

oTests carried out by the Czech consumer association 59   60 found that more than half of the tested margarines, 60% of wafers and 20 % of chocolate waffles tested were above the 2 % limit.

·Quantitative comprehensive data of industrial trans fats use for particular food sectors, or particular regions or sorted by company size in the EU is not available. However, available data 61 has shown significant presence of trans fats in different food categories, such as convenience products, cereal products, confectionary, crisps, savoury, biscuits, fast food products, fats and oils, without however a distinction between prepacked and non-prepacked, locally produced produce or not. Given that larger companies were more likely to participate in reformulation campaigns than SMEs, the residual share of products still high in trans fats is considered to be higher among SMEs.

Consultation with Member States 62 confirmed the findings in the JRC report 63 . In some Member States high intake levels prompted activities to reduce intake levels of trans fats, contributing to enhanced reformulation activities and reduced levels.

A study 64 noted that, in different Member States, industrial trans fats levels in some foods were still above 2 % of their total fat content and that, in some EU countries, industrial trans fats levels in pre-packed biscuits, cakes and wafers have not dropped meaningfully since the mid-2000s. The authors of this study continued analysing the evolution of the market in six countries in South East Europe covered by the previous study (including two EU Member States) and noted that availability of popular foods with high amounts of industrial trans fats increased from a high level in 2012 to an even higher level in 2014. 65 Another study 66 specifically focused on the Portuguese market showed that, in 2013, total trans fats content in different foods ranged from 0.06 % to 30.2 % of the total fat content of the food (average 1.9 %), with the highest average values in the 'biscuits, wafers and cookies' group (3.4 % of the total fat content of the food). 50 samples out of 268 (19 %) contained trans fats at amounts higher than 2 % of the total fat content of the food. Replies during the OPC revealed that 78 % and 77 %, respectively, of respondents agreed with the problem description above with regard to intake levels and content in foods, while 9 % and 8 disagreed, all but one disagreeing respondent stated that intake level and contents in food were actually higher that described above. 67 An unpublished study in Hungary 68 confirmed a steadily increasing trend of trans fats content in foods from 2009/2010 until 2012, which was reverted only in 2013 when the decision ona national legal limit of trans fats was notified to the Commission (further details are provided in section 5.1).

Sources of trans fats

Ruminant trans fats in dairy products or meat from cattle, sheep or goat 69 are present in relatively constant, low proportions of the fat part of those foods, at levels most commonly around 3 % (ranging from 2 to 9 %) of the total fat content. 70  

The primary dietary source of industrial trans fats is partly hydrogenated oils which contain various amounts of trans fats (up to more than 50 % of the total fat content). The partial hydrogenation process turns oils into semi-solid and solid fats. Industrial trans fats in the form of partly hydrogenated oils have been used or introduced into manufacturing processes of foods in order to achieve at comparative low prices a particular technological function, such as a solid fat texture at room temperature (e.g. in vegetable fat cocoa coatings). 71 Other than partly hydrogenated oils, industrial trans fats can also be the result of refining of unsaturated oils or of heating and frying of oils at too high temperatures (> 220°C). 72

Reduction of intake levels of industrial trans fats is technologically feasible. However, the fat composition of ruminant fats with regard to their trans fats content is not modifiable to a significant degree, therefore their intake cannot totally be avoided when consuming ruminant derived foods that are important in the EU diet of the EU population as they contribute essential nutrients. Also, ruminant trans fats sources generally contribute in a limited way to high total trans fats intake. 73 National public health policies generally address the problem of intake of ruminant trans fat intake already by initiatives to reduce saturated fat intake. 74 Although different actions were taken in several Member States and intakes have decreased over the past years, industrial trans fats are still present at levels of concern in certain foods and intakes are still excessive in certain cases. The evidence collected by ICF also suggests that gains obtained in recent years through voluntary industry initiatives may have reached their limits. The issue is of particular relevance in certain Member States and for particular population groups. This results in the following problems:

·Protection of consumers' health

Different levels of protection of consumers' health currently exist in the EU, depending on the presence of foods with high industrial trans fats content in the Member State's market (presence influenced by the existence or not of national regulatory or not-regulatory initiatives) and on consumers' consumption patterns. Consumption patterns are influenced by socio-economic factors (e.g. consumers with lower income are more likely to consume products with high industrial trans fats content that are generally sold at a lower price 75 so that this situation contributes to the perpetuation of health inequalities in the EU. 76 In light of the global trend to reduce intakes of trans fats and the WHO's recent REPLACE initiative ('trans fat free by 2023') recommending to ‘legislate or enact regulatory actions to eliminate industrially-produced trans fats’, a number of countries worldwide have acted and others are expected to act. Therefore, not taking any action at EU level could entail a reputational risk for the EU of not adequately addressing a serious health concern of global dimension.

·Functioning of the Internal Market and international trade

Only some Member States have taken action on industrial trans fats, which is problematic for the effective functioning of the Internal Market: food business operators active in countries where no limit on industrial trans fats exists have no related reformulation costs and are therefore at a competitive advantage vis-à-vis operators active in countries where legal limits exist or operators abide by self-regulatory commitments. The current lack of a consistent approach at EU level means that there is not a level playing field between operators that have reformulated their products in order to reduce or fully remove ingredients containing industrial trans fats, due to self-regulation, voluntary agreements with national governments or legal measures, and those that have not. Generally, manufacturers face higher cost if they produce different varieties of a food with different ingredients to meet diverging national legal limits, rather than benefitting from economies of scale regarding one recipe for a food product. Producers that have not taken any steps to reduce industrial trans fats may save costs as they do not invest in reformulation and through use of lower priced ingredients. This may provide a competitive advantage in the market.

This is particularly relevant for operators active in different Member States. At the same time, operators active in countries where no limit on industrial trans fats exists are negatively affected by the legal uncertainty over whether new initiatives to reduce industrial trans fats intakes will be adopted at national level and might have difficulties in planning R&D investments. The above described situation also hampers international trade: operators from third countries exporting their foods to the EU are subject to different conditions depending on where their foods are marketed. Similar considerations also apply to EU exporters to third countries. In countries without legal measures but with industry complying with voluntary agreements, industry may face unfair competition with producers in third countries. This issue in relation to external trade stems from import of products with high industrial trans fats contents into the EU. Eastern European countries may be at heightened risk for such imports due to their geographical position and the price sensitivity of consumers. Empirical evidence supports this description 77 . Of note, all national legal measures apply to all foods sold in the country, including both foods produced nationally and foods imported from other Member States or from third countries.

Types of stakeholders affected

1.EU consumers are directly exposed to trans in foods and would be affected by any EU initiative on the matter through reductions in their trans fats intakes. Consumers will benefit from reduced risk of contracting coronary artery disease when industrial trans fats intakes are reduced, but they may experience an increase in the price and potentially a change in the quality and attributes of certain food products. Consumers in Member States where foods containing high levels of industrial trans fats are still on the market and consumers with high trans fats intakes are particularly affected.

2.Healthcare providers and healthcare systems are affected by the impact the presence of industrial trans fats has on the incidence of coronary heart disease and associated costs of healthcare.

3.Food businesses, including SMEs, would be impacted by action to limit industrial trans fats in food and additional costs. More specifically:

·Manufacturers of pre-packed foods placed on the market in the EU or exported outside the EU operating chiefly in the following sectors: manufacture of margarine and similar edible fats, bread, fresh pastry goods and cakes, rusk and biscuits, preserved pastry goods and cakes, cocoa, chocolate and sugar confectionery, condiments and seasonings, preserving of potatoes;

·Mass caterers providing non pre-packed foods to consumers (e.g. fries) which (might) contain industrial trans fats, restaurants and businesses offering mobile food service (different sizes of business: multinational, national, SMEs);

·Manufacturers of ingredients placed on the market in the EU or exported outside the EU which contain industrial trans fats or are trans fats-free and can, in the latter case, be used as replacement of industrial trans fats-containing ingredients (e.g. frying oils) (mainly large operators);

·Retailers distributing foods which (might) contain industrial trans fats: they will be indirectly affected (different sizes of business).

·Third-country-based food business operators exporting into the EU would be affected by any EU initiative on the matter.

All food business operators have a role in determining the level of industrial trans fats in their products. Many of the large players have reduced industrial trans fats levels through reformulation. In this context, major producers and associations of the food industry have supported the implementation of a recommendation of a legal limit of industrial trans fats. 78   79

Manufacturers of oils and fats have a critical role to play as suppliers of ingredients that may contain industrial trans fats to food manufacturers, particularly to SMEs. A number of manufacturers have already acted on this issue, while others have not (in particular smaller and less organised businesses).

4.Public authorities of EU Member States are directly affected by the problem and by EU action as they will be responsible for implementing, publicising, administering and enforcing the new rules, incurring costs as a result.

5.Populations around the globe are affected, especially given concern about the potential impact on palm oil consumption and its effects on climate change and biodiversity.

2.2.What are the problem drivers?

The drivers of industrial trans fats intake are partly a matter of efficiency of industrial recipe and process and related lower costs, partly one of different national approaches and partly related to consumer behaviour.

Industrial recipe and process

High trans fats intake results from consumption of food products containing high levels of industrial trans fats. Industrial trans fats are used in the manufacturing process and in the recipe of certain foods for technological reasons. Especially, partly hydrogenated oils are solid at room temperature and relative stable, either to rancidity over storage time or when heated repeatedly as frying oils. 80

In addition, they may be chosen due to their competitive price. Alternative ingredients need to be found when replacing ingredients with high trans fats levels, and sometimes developed, so that the product presents similar characteristics of texture, taste, etc. after reformulation.

Reformulation can entail substitution or development of a new product, and sometimes changes to the manufacturing equipment to accommodate new ingredients. This poses various challenges to industry, and chiefly to smaller businesses, which may be dependent on suppliers to provide alternative ingredients.

In order to overcome cost-related barriers to replace ingredients with a high industrial trans fats content with alternatives, a stimulus to change by the market or regulators, may be needed, such as market pressure, legal obligations or other action by public authorities. The level of corporate social responsibility as well as responsiveness of food business operators vary depending on the Member State.

Different national approaches

National authorities have the power to limit industrial trans fats levels in foods through initiatives at national level if they find it necessary to protect public health. However, evidence 81 shows that national authorities have different approaches to industrial trans fats, with some acting and others not.

Among the Member States that have introduced legislation, a limit of 2 % of industrial trans fats of fat was the preferred choice. However, additionally, 4 Member States have complemented this with different limits established for lower fat products 82 . Due to those differences, all foods that contain between 3 and 20 % of fat with industrial trans fats levels between 2 and 4 % of fat would comply in 4 Member States but not comply in 3 Member States and all foods that contain less than 3 % of fat with industrial trans fats levels between 4 and 10 % of fat would comply in 2 Member States but not in 5 Member States. Those differences are in practice significant, as the majority of food products are below 20 % of fat and many are below 3 % of fat per 100 g of food. Tall existing Member States measures have in common the general 2 % limit for all foods with more than 20 % of fat content, while this food category represents generally a minor share of the total food offer. In Member States where voluntary measures have been taken, reductions were achieved, however, not always in line with legal limits mentioned above.

There is evidence collected by ICF about the effectiveness of both legal as well as voluntary measures in Europe. For example, in Denmark, a legal limit led to virtually eliminating industrial trans fats from the Danish food supply 83 . Data collected in Austria before and after the introduction of the legal measure indicate that from bakery products controlled over time, once before the introduction of the legal measure and twice afterwards, 18 out of 30 samples were not compliant while 3 years after the measure came into force 1 out of 68 samples was not compliant and two years later all samples were compliant 84 . Data collected in Hungary before and after the introduction of the legal limit point to a reduction of industrial trans fats intakes per person foods in the order of 40 % to 75 % 85 .

While it could be assumed that more Member States would take action in the absence of EU intervention, there are no precise indications for all Member States, taking into account that incentives for food business operators to act can vary significantly and national authorities have different approaches to industrial trans fats. If parallel action is not undertaken at national level in all EU Member States, operators would remain subject to different conditions for the manufacturing and placing on the market of foods that could contain industrial trans fats and obstacles to the functioning of the Internal Market would persist. At the same time, products with high industrial trans fats levels would remain on the market in some parts of the EU and intakes of trans fats would remain excessive for certain consumer groups. This would negatively affect the protection of consumers' health and would contribute to the perpetuation of health inequalities in the EU.

Even if action was undertaken at national level in all EU Member States, it is very likely that differences would exist in the timing of the interventions (i.e. not all national actions would be launched at the same time) and in their content (i.e. it is possible that different measures would set different legal limits or cover different products). This explains the clear added value of an EU-based, EU-wide action: the possibility to ensure a level playing field in the Internal Market and the same high level of protection of consumers' health by the means of an initiative that would apply simultaneously in the entire EU and would minimise the risk of national regulatory interventions (further) fragmenting the Internal Market.

Consumer behaviour

Low consumer awareness of the risks associated with the consumption of trans fats may also contribute to industrial trans fats intake. The evidence in the EU points to low levels of consumer information and consumer awareness on trans fats 86 , including which ingredient that is declared on the label or which non prepacked foods may contain trans fats. Many foods are potential sources that are difficult to avoid totally as this would lead to very restricted dietary choices.

Not all consumers can relate the information on the use of partly hydrogenated oils required by Regulation (EU) No 1169/2011 to the presence of industrial trans fats in foods and not all consumers can use that information to effectively compare different products taking into account their overall nutritional composition. 87

Finally, other considerations may influence consumer behaviour (e.g. cost, taste, habits) stronger than the intention to reduce trans fats intake.

2.3.How would the problem evolve 

Whether the decline in industrial trans fats levels in food product and industrial trans fats intake observed in the past years will continue at the same speed and achieve a near elimination of industrial trans fats in the EU is not certain. Contrary, there is some evidence of new products that contain high levels of industrial trans fats being introduced to the market in recent years. 88 Consumer health would continue to be at risk in a number of Member States, particularly in the Eastern and Southern part of the EU. The perspectives provided by stakeholders in the consultation conducted by ICF in the context of the study to support this IA suggested that the problem would remain in the absence of EU action but also that many Member States would act unilaterally in the absence of EU action. Based on previous experiences, national legal measures introduced for public health protection, would likely differ to a certain degree in scope and content and could contribute to fragmenting further the Single Market for food products.

3.Why should the EU act?    

3.1.Legal basis    

EU action could be taken within the framework of Article 114 TFEU, in order to ensure the functioning of the Internal Market, whilst ensuring a high level of protection for health and consumers. The adoption of a legal measure to set limits to trans fats presence in food can be considered through the implementation of existing legislation, more specifically, on the basis of Regulation (EC) No 1925/2006 on the addition of vitamins and minerals and of certain other substances to foods. That Regulation aims at providing a high level of consumer protection whilst ensuring the effective functioning of the internal market. The Regulation empowers the Commission to take measures restricting the addition of certain substances to foods or the use of such substances in the manufacture of foods in view of harmful effects on health which have been identified in relation to a particular substance. For the specific case of the presence of trans fats in food, harmful effects have been identified based on scientific advice provided by EFSA, as explained under point 1.

3.2.Subsidiarity: Necessity of EU action    

The existing situation on industrial trans fats negatively affects the protection of consumers' health and contributes to the perpetuation of health inequalities. Excessive intakes of industrial trans fats are associated with avoidable suffering and pose burden on public health care systems. 89

Industrial trans fats are still present at levels of concern in certain foods in many Member States and particularly in Member States where no national action has been undertaken so far (voluntary or regulatory) to reduce such levels. While average daily trans fats intakes for the overall EU population are below 1 % of daily energy intake, some population groups have (or are at risk of having) higher intakes, including low-income groups and younger population groups (18 to 30 years).36 As calculated by the JRC, up to 25 % of surveyed individuals aged 20-30 years have trans fats intakes above 1 % of daily energy intake.36 But even if population average intake levels are around or slightly below 1 % of daily energy intake, this level can be considered as excessive, taking into account the recommendation from the European Food Safety Authority that intakes should be as low as possible. Empirical evidence supports this view, as reducing intake levels of industrial trans fats from below 1 % or daily energy intake to minimal levels in Denmark, mortality attributable to cardiovascular disease decreased on average by about 14.2 deaths per 100,000 people per year relative to a synthetic control group. 90  

According to the ICF research, levels of industrial trans fats are not necessarily declining in the coming years. While data gathered for the ICF study confirm a trend towards industrial trans fats reduction in food products, it shows also that the limits of the current approach with no action taken at EU level have been reached. Levels of industrial trans fats appear to remain high in certain countries, predominantly Eastern and Southern Europe, and certain sub-groups of food businesses, particularly SMEs. Levels were still above 2 % of their total fat content and in some Eastern and South-Eastern EU countries, industrial trans fats levels in pre-packed biscuits, cakes and wafers have not dropped meaningfully since the mid-2000s 91 . The authors of this study continued analysing the evolution of the market in six countries in South-East Europe covered by the previous study (including two EU Member States) and noted that availability of popular foods with high amounts of industrial trans fats increased from a high level in 2012 to an even higher level in 2014 92 . Another study 93 specifically focused on the Portuguese market showed that, in 2013, total trans fats content in different foods ranged from 0.06 % to 30.2 % of the total fat content of the food (average 1.9 %), with the highest average values in the “biscuits, wafers and cookies” group (3.4 % of the total fat content of the food). 50 samples out of 268 (19 %) contained trans fats at amounts higher than 2 % of the total fat content of the food. Several consultations and triangulation of data have confirmed these findings.

Even under the assumption that a decline in industrial trans fats intake would take place over time without EU level action, evidence suggests that from a society benefit/cost point of view, taking EU level legal action is a highly efficient measure 94 . Therefore, opportunity cost for not acting at EU level are high, and they could be reduced the faster action is taken and measures are implemented, with resulting benefits for human health and cost to society.

3.3.Subsidiarity: Added value of EU action    

The existing situation on industrial trans fats hampers the effective functioning of the Internal Market.

Whilst action has been taken by some countries 95 , and others may be expected to act in the absence of an EU initiative 96 , rapid and universal action on industrial trans fats by Member States is currently not envisaged. Products with high industrial trans fats content would therefore remain on the EU market and industrial trans fats would continue to contribute to health impacts and health inequalities.

In addition, legal measures and voluntary initiatives taken by Member States so far differ, as different national views in relation to acceptable levels exist. 97 Additional measures at Member State level could lead to further differences in approach, adding complexity and cost for food business operators.

Furthermore, as a basis for the Internal Market in foods, the EU has a detailed and rather comprehensive system of general and specific food laws, ensuring that products can be freely traded, but also that consumers can be confident that products offered are safe. To address potential health concerns, food safety measures ensure a high level of health protection of consumers. Excessive industrial trans fats in foods pose risks from a food safety angle. In case a food constituent is linked to serious health concerns, confirmed by an opinion by EFSA, their presence should be either prohibited or limited, both for products produced in the EU and for imported products. Recent EFSA opinions 98   99 in relation to the presence of industrial trans fats in food ingredients recommended that the Commission considers revising the specifications for the ingredients, ‘including maximum limits for trans fatty acids’.

Added value at EU level thus derives from the possibility to ensure a level playing field in the Internal Market and the same high level of protection of consumers' health.. In this context, it is of note that 100 in the consultation that preceded the adoption of the Commission's report, several Member States proactively signalled their preference for an EU level initiative on industrial trans fats.

4.Objectives: What is to be achieved?    

4.1.General objectives    

To address the problems that industrial trans fats intake is an important risk factor for the development of coronary heart disease and contributes to the perpetuation of health inequalities within the EU, the identified general objectives of EU action on industrial trans fats to be achieved are:

·To ensure a high level of health protection for EU consumers;

·This will also contribute to reducing health inequalities, one the objectives of Europe 2020 101 ;

To address the problem of obstacles to the functioning of the Internal Market (unfair competition, legal uncertainty), the identified general objective is:

·To contribute to the effective functioning of the Internal Market for foods that could contain industrial trans fats.

4.2.Specific objectives    

The following specific objectives of EU action on industrial trans fats to be achieved are:

·To reduce intake of industrial trans fats in the entire EU for all population groups;

·To ensure that the same rules/conditions apply in the EU to the manufacturing and placing on the market of foods that could contain industrial trans fats, so as to ensure legal certainty of EU food business operators within and outside the EU 102  

Data collected during the Impact Assessment support the view, that trans fats are particularly a problem in Eastern and South-Eastern Europe, a region that generally also suffers from relatively high rated of heart disease and lower life expectancy than Western Europe. The results of a study 103 suggest that industrial trans fats levels in pre-packaged biscuits, cakes and wafers in some Eastern and South-Eastern European countries have not dropped meaningfully since the mid-2000s. This suggests that in certain parts of the EU little progress has been made, while in some Western EU countries reductions were achieved. The European consumer association BEUC highlighted in their position paper on trans fats in 2014 104 , that regional inequalities between Western versus Eastern EU countries persist, citing results from product testing, which showed consumers in Eastern EU countries are more exposed to industrial trans fats than their Western neighbours. A test on margarines and wafers carried out by the Czech consumer association 105   106 in 2013 and 2014 confirmed that reformulation efforts have not been equal in Eastern and Western EU countries. According to a published study 107 , the same product categories would contain minimal amounts of industrial trans fats, while in Eastern Europe, substantial contents of trans fats were found. Figure 1 summarises the problems, drivers and objectives associated with industrial trans fats in the EU.

Figure 1 Illustrative summary of the problems, drivers and objectives associated with industrial trans fats in the EU

5.What are the available policy options? 

5.1.What is the baseline from which options are assessed?    

In the baseline scenario, option 0, no initiative would be taken on trans fats at EU level. The qualitative and quantitative analysis 108 was informed by the baseline scenario of a study completed by the JRC and the qualitative evidence collected in the external study by ICF.

The JRC study highlighted that the assumed baseline of 10 years in their modelling exercise was chosen as a rather conservative approach to show that measures which are cost effective under this very conservative assumption would prove even more cost effective under any further, less conservative baseline scenario 109 .

The assessment methodology was designed to accommodate uncertainty about the future trend in industrial trans fats intake in the absence of EU action (the baseline scenario). The purpose was to reinforce the analysis by referring to three possible future trends (baselines), taking into account uncertainty rather than focusing on one scenario only.

It is suggested that industrial trans fats levels in food have been declining over time under the influence of various factors, while there is also some evidence that the decline has levelled off, according to the ICF study. In its recent public health economic evaluation32, the JRC extrapolated from available evidence and based its modelling on the assumption that industrial trans fats would be completely removed from the EU food supply chain in 10 years. While data gathered for the study by ICF confirm this trend, it shows also that most changes that could be triggered in the absence of EU policy action have already taken place, either as a result of voluntary initiatives or national legislation. Nevertheless, levels of industrial trans fats in foods appear to remain high in certain countries and certain sub-groups of food businesses, particularly SMEs.

A continuous downward trend in the years to come is not certain. 110 Industry in some Member States has not acted voluntarily on industrial trans fats, and the evidence from certain Member States collected by ICF suggests that a voluntary approach may not deliver any progress there. Data on the industrial trans fats content of foods manufactured and sold in some Member States 111 suggests that, in spite of reductions in certain categories of products, levels of industrial trans fats in other food products remain high. The evidence on voluntary industry initiatives collected by ICF strongly suggests that potential action by those sectors willing to act and sufficiently well organised at national and EU level to carry out coordinated reductions in industrial trans fats havs already been carried out. Other sectors and countries that have not acted voluntarily are highly unlikely to do so in the near future. Further evidence collected in six countries (including the EU Member States Croatia and Slovenia) has found that the number of packages of food products (considering the group of biscuits, cakes, wafers) that contained more than 2% of total fat as industrial trans fats had doubled between 2012 and 2014 112 , indicating that food industry operators had expanded their offer of products with high industrial trans fats content, contradicting the notion of a general downward trend. Further evidence for actual increases of industrial trans fats exposure, particularly in Eastern Europe, is provided in a recent, unpublished study in Hungary, the outcome of a collaborative agreement between the World Health Organization Regional Office for Europe which supported the process and technical product, and the Ministry of Human Capacities of Hungary 113 . Hungary introduced its national legal limit in February 2014 with a transition period of 1 year. In the years proceeding to the enforcement of the national legislation, a steady increase in the percentage of products above the legal limit could be observed: 2009/2010 16% of products surveyed, 2011 27 %, 2012 29 %, respectively.

Figure 2: Compliance with the Hungarian national legal limit by year (%)

Substantial improvements were only seen from the period where the national legal limit had been decided and notified to the EU, in 2013 with 11% of the sampled products above the legal limit, with following steady declines, 2014 7%, 2015 6 % and 2016 2%, showing the effectiveness of a legal limit to revert an increasing trend of products with high industrial trans fats levels on the market. This development is illustrated in Figure 2

Likewise, mean trans fats content in products was seen to steadily increase on the Hungarian market until a national legal limit was decided, as shown in Table 1.

Table 1: Mean trans fats content (g/100 g food) in the food samples by year in Hungary

Possible reasons for increased levels of industrial trans fats in foods are, for instance, availability of food ingredients with high industrial trans fats levels at low prices, a high price sensitivity of consumers, low responsiveness of food business operators to respond to calls for voluntary reformulation and a perceived low reputational risk for food business operators linked to the offer of products with high levels of industrial trans fats. For the Hungarian example described above, it was not possible to determine whether products with high industrial trans fats levels were imported as information was only available about the distributor and not about the manufacturer.

Of note, the national legal measures prohibit the sale of non-complying foods on the national territory, while non complying foods may still be legally produced for export.

A number of published evidence, including research articles, were available for citation to provide evidence, apart from data on trans fats levels collected by JRC, showing and confirming higher levels of industrial trans fats, particularly also in Eastern European countries 114   115   116   117   118 . Despite this fact, the validity of assuming a baseline scenario of no change has been confirmed by ICF. ICF conducted an online consultation to maximise their ability to validate the data collected during desk research and expert interviews and triangulate the findings from the impact assessment with a wide range of stakeholders. This enabled ascertaining the validity of key elements of the analysis. The first part of the consultation posed general questions on current and predicted industrial trans fats use under different policy options. Overall the results from the ICF consultation have confirmed the appropriateness of the assumptions and estimates, while they have helped to qualify the baseline scenario. According to the ICF research, levels of industrial trans fats are not necessarily declining in the coming years. While data gathered for the ICF study confirm a trend towards industrial trans fats reduction in food products, it shows also that most changes that could be triggered in the absence of EU policy have already taken place, either as a result of voluntary initiatives or national legislation.

This suggests that obstacles stand in the way of further changes and of further diffusion of initiatives, either private or public, to that part of the EU food industry that has not yet reduced industrial trans fats levels in its products. Whether these obstacles would be removed in the absence of EU activity is not clear from the evidence that has been gathered by ICF. A continuous downward trend in the years to come is therefore not certain.

This uncertainty in the baseline is mitigated by the analytical approach; three variants of the baseline scenario have been adopted to capture that uncertainty, about how trans fat intakes may develop in the future. The policy options are compared against each variant. This approach helps to ensure that the conclusions about the absolute and relative impacts of options are robust in the context of all foreseen reference scenarios, thereby accommodating the uncertainty about future evolution of the problem in the absence of further EU action (cf Figure 2):

·A continuous decrease leading to the complete elimination of industrial trans fats from the food chain over a period of 10 years (B1 – ’10 year elimination’);

·A continuous decrease leading to the complete elimination of industrial trans fats from the food chain over a period of 15 years (B2 – ’15 year elimination’);

·Industrial trans fats intake remains constant at current levels (B3 – ‘no change’).

The three variants of the baseline represent the spectrum of expected possible trajectories – industrial trans fats intake remaining constant at current levels, a linear decline in industrial trans fats intake to zero over 15 years and an accelerated linear decline to zero over 10 years.

From an impact appraisal perspective, the first variant (B1) is conservative: An option that is cost-effective under the first variant (B1) would be even more cost-effective under the other variants.

Figure 3 Dynamic baseline: illustrative representation of how benefits of industrial trans fats control arise compared to the variants of the baseline scenario 119 (source: ICF)

iTFA : industrial trans fats

5.2.Description of the policy options    

Overall, three options were considered, option 1 and 3 were subdivided into two sub options each to consider different instruments. Logic models and theories of change for each option are presented in Annex 11, Figure 4 describes potential dietary sources of trans fats and indicates where the different options affect those sources.

Figure 4 Overview of potential dietary sources of trans fats and where the different options affect the (% trans fats are given as % of fat content)

FBO: food business operator

Option 1 – Establishment of a limit for the industrial trans fats content in foods

In this option, the EU would establish a limit for the presence of industrial trans fats in foods, both pre-packed and non-pre-packed.

Different limits could be considered, one possibility would be to set the limit of industrial trans fats at 2% of the total fat content of the food, in line with the approach followed in seven Member States that have already taken legislative action on the matter. 120 This limit could be set through different instruments:

Option 1a: Voluntary agreement with the relevant food business operators to set a limit for industrial trans fats content in foods

In option 1a, a limit for industrial trans fats content in foods would be established by a voluntary agreement at European level between relevant food business operators. The agreement as a form of self-regulation would be under the auspices of the Commission, and involve EU-level representative organisations from the industry, themselves representing both national federations of companies and large companies operating across many countries of the EU.

Since some industry sectors are not organised and represented at EU level, this would not be fully inclusive. Voluntary agreements would primarily focus on foods sold to the consumer (and not include ingredients that are sold as inputs to final products).

The agreement is assumed to include an annual reporting requirement for participants. Industry associations would collect and report the information on behalf of their members. This information could be commercially sensitive, and business associations would need to operate as a “safe space” 121 , collecting and anonymizing the information from its members so that it may then be publicised. Such arrangements would build upon the examples of voluntary agreements to reduce industrial trans fats content in food which have been implemented in Germany and in the Netherlands.

It is assumed that the agreement would set a target of achieving levels of industrial trans fats in food products below 2% of fat within a period of 3 years. The evidence collected by ICF suggests that such a timespan would enable producers to factor reformulation into their regular cycle of product review and reformulation (whereas legislation might impose a shorter transition period for businesses).

Reporting obligations (and so the associated costs) would continue to apply even after the participating sectors had reduced industrial trans fats content to below the threshold. A review mechanism and ‘sunset clause’ by which reporting requirements lapsed a specified period after objectives had been met would mitigate ongoing costs incurred even after industrial trans fats had been reduced to levels below 2% of fat. There would be a credible incentive for Member States that legislation would be introduced in the absence of progress.

A part of the food business operators that participated in the consultations favour a voluntary approach with regard to a legal limit, as more flexibility to act would be given. Generally, neither consumers nor NGOs favour this approach as it does not guarantee a high level of health protection. Public authorities think this option is somewhat appropriate as it could deliver some results, while most Member States support a harmonised, legal European approach (option 1b) ensuring the Internal Market and fair competition between food business operators in all EU Member States.

Option 1b: Legally-binding measure to set a limit for industrial trans fats content in foods

In Option 1b, EU legislation would set a limit industrial trans fats content of 2% of the total fat content of final food products sold to the consumer, following the example of 2% limits to final food products in some Member States' legislation. 122   123 The 2% limit assuming it applies to all products consumed in a very conservative scenario means in practical terms an intake of between 0.6 and 0.7 % of energy intake from industrial trans fats for a large number of average consumers (between 33 and 60 %) in the EU. A 2 % limit applies to the content in the particular food or product and it would still enable minimal use of partly hydrogenated oils as raw ingredients containing industrial trans fats by the industry, e.g. for the manufacture of additives. Such additives could continue to be used, provided that the total industrial trans fats content of the final food sold to the consumer meets the 2% limit on fat basis.

In order to implement option 1b, it is assumed that the majority of food ingredients in the EU will comply with the legal limit, so that food manufacturers are sure to comply with the legal limit and that most food manufacturers that buy ingredients will ask for a industrial trans fats specification of not more than 2% of their supplier. In specific cases, ingredients with higher industrial trans fats levels could be used, as explained above. The enforcement of option 1b includes testing of final food products in the market for their industrial trans fats level. The JRC has proposed assessment methods for industrial trans fats and developed a standardised calculation method to estimate the industrial trans fats level in a food that contains industrial and ruminant trans fats.

Alternatively, a more differentiated approach could be chosen, with higher limits (above 2% of total fat) for products with low fat content, and 2% of total fat for food categories with high fat content. Such differentiated limits have been adopted in some Member States. 124 Consistently with the modelling study by the JRC, a transition period of 2 years is assumed.

A large part of the food business operators that contributed during the various consultations favour this option that is achievable and provides a level playing field and avoids any further fragmenting of the EU Internal Market. Also, most public authorities and Member States, as well as consumers, and NGOs favour this approach as it guarantees a high level of health protection, is in line with certain national legal measures already in force in the EU, as it is ensuring the Internal Market and fair competition between food business operators in all EU Member States.

Limits below 2 % of fat content were not considered in detail in this IA. During the normal refining steps (deodorisation) of oils that contain high levels of polyunsaturated fats industrial trans fats can be formed, even if the oil is not undergoing partial hydrogenation. Oils with a high content of polyunsaturated fats, providing essential nutrients, are generally recommended as a part of a healthy diet. Also, in food service establishments, during normal frying processes trans fats are formed to a certain degree. It would not be proportionate to ask small food business operators active in food service to frequently control the level of trans fats produced in the frying oils to ensure that a low threshold limit is not exceeded. The 2 % limit on fat basis has been found to be in line with the need to accommodate trans fats levels generated during normal oil and food processing. However, empirical evidence shows, that with this threshold, very low average intake levels of industrial trans fats, in the order of 0.009 %, were achieved with the legal limit of 2 % per 100 g fat content. Therefore, the 2 % limit was assumed to achieve a high level of health protection while being technologically feasible for food business operators.

Option 2 – Introduction of the obligation to indicate the trans fats content of foods in the nutrition declaration

Option 2 involves the introduction of an obligation to indicate the trans fats content as part of the (mandatory) nutrition declaration for pre-packed foods. This would provide incentives to the industry to reformulate and reduce trans fats from food products and enable consumers to make informed food choices. 125

The labelling obligation would be required for all foods that carry a nutrition declaration, with resulting costs even for foods free of trans fats, while non pre-packed foods e.g. in restaurants, are out of scope. Where applicable, the nutrition declaration would describe total trans fats content, both ruminant and industrial trans fats.

A two-year transition period, would allow a majority of businesses to process label changes into their normal cycle of label updating. 126

A large part of the food business operators that contributed during the various consultations do not favour this option due to the high administrative burden and linked costs. Generally, food business operators active in the vegetable oils sector have been more favourable in principle, given that this measure also covers ruminant trans fats, while food business operators providing ruminant fat sources are not favourable to this option as they are unable to reformulate the basic fat composition of ruminant fats and fear negative impacts on the overall diet of consumers as a potential consequence. Consumers often point to their desire for transparency in relation to the foods they eat and prefer to be provided with comprehensive information, so they supported this option. Public authorities generally 'do not favour' this measure or only 'favour it somewhat' according to the results of the OPC; also NGOs do not largely support this option, one of the reason here being that labelling covers only part of the food offer, pre-packed foods, and therefore only part of the trans fats problem.

Option 3 – Prohibition of the use of partly hydrogenated oils in foods

In this option, the EU would follow a similar approach as adopted in the US and would prohibit the use of partly hydrogenated oils in foods, as primary dietary source of industrial trans fats. This could be achieved through a voluntary agreement with the relevant food business operators (sub-option 3a), or a legally-binding measure (sub-option 3b).

Option 3a – Voluntary measure to eliminate the use of partly hydrogenated oils

In Option 3a, partly hydrogenated oils would be removed from foods through a voluntary agreement negotiated and managed at European level. Food business operators would commit to the ban individually or through their representative associations.

The arrangements for the voluntary agreement would be similar to that for option 1a. There is currently no definition of partly hydrogenated oils in EU law or in the Codex Alimentarius. For the implementation of Option 3a, a definition of partly hydrogenated oils would need to be established at EU level, linked to a measurable indicator, which could then be relied on for monitoring purposes. The US Food & Drug Administration 127   defined partly hydrogenated oils in terms of their 'Iodine Value', which is measurable.

Consumers and public authorities consider this option as somewhat appropriate, while industry and NGOs generally are not supportive about what they see as less effective voluntary action.

Option 3b – Legal measure to prohibit the use of partly hydrogenated oils

This option would mirror action taken in the USA. In June 2015 the US Food and Drug Administration concluded that partly hydrogenated oils are not “generally recognized as safe” for use in human food, and introduced a prohibition on their use, with a compliance period of three years. This allows food companies to either reformulate products without partly hydrogenated oils and/or petition the FDA to permit specific uses of partly hydrogenated oils. A similar measure has been introduced in Canada. 128

This option would introduce via EU law a prohibition on the use of partly hydrogenated oils as food ingredients. Provision could be made for limited derogations applicable to certain categories of products and for technical uses of partly hydrogenated oils in limited quantities. Partly hydrogenated oils are the primary dietary source of industrial trans fats in the diet. Although all refined edible oils and oils heated up under high temperatures during cooking processes contain some industrial trans fats as an unintentional by-product of their manufacturing- or the cooking process, industrial trans fats are an integral component of partly hydrogenated oils and are purposely produced in these oils to affect the properties of the oil and the characteristics of the food to which they are added.

As for Option 3a, the matter of the definition and a suitable test for enforcement purposes would need to be agreed. Neither at EU level, nor at international level (Codex Alimentarius) a definition of ‘partly hydrogenated oil’ exists so far. The definition of 'partly hydrogenated oil’ would need to be decided, and a suitable test would need to be agreed for monitoring and enforcement purposes. In the US, the definition of partial hydrogenation is linked to the extent to which a fat or oil reacts with iodine, referred to as the “Iodine Value (also referred to as IV)”. In this context, partially hydrogenated oils and fats are defined as those vegetable oils and fats with an Iodine Value (IV) above 4 129 . The iodine value does not measure directly the level of trans fats and hence does not always imply a reliable trans fats result and can also vary depending on other technical parameters applied during the hydrogenation process, irrespective of the trans fats content. The Iodine Value can also vary depending on the refining process used or depending on the presence of other substances in some vegetable oils and fats (called “unsaponified components”). However, when comparing option 3b to option 1b it is noted, that according to the definition chosen in the US for partly hydrogenated oils, in many cases oils above levels of 2 % trans fats per fat basis would be covered.

It is expected that difficulties may arise for enforcement of option 3b and the linked Iodine Value measurement based definition in final food products, where in most relevant cases, vegetable oils/fats are only one ingredient. The US approach is based on the ban of an ingredient, which means they must not be brought into circulation, which may be controlled. End product controls using the Iodine Value as an analytical measure are not applicable to finished, multicomponent products.

For cases where suspicion about compliance with the legal ban of using partly hydrogenated oils is raised, a document check or a check of the ingredients used for manufacture of a product or checks at the manufacturing plant for the oils used is necessary. This is thought to be particularly demanding for imported products, but also for control authorities in a Member State controlling compliance of products manufactured in another Member State.

Option 3b completely bans any use of partly hydrogenated oils. It means that a production process is banned, where industrial trans fats are produced in high amounts. The definition of partly hydrogenated oils applied in the US equals roughly oils that contain industrial trans fats levels of 2% on fat bases. Most oils that contain approximately more than 2 % trans fats would be considered by the definition applied in the US as being partly hydrogenated oils and would therefore not be allowed any more in the food chain. This is contrary to option 1b, where residual use of industrial trans fats for the manufacture of ingredients for food business operators would be permitted, as long as the legal limit of maximum 2 % of industrial trans fats per 100 g of fats in the food product sold to the final consumer is complied with.

A large part of the food business operators and NGOs that contributed during the various consultations does not favour this option that would ban a production technology rather than limiting the problematic substance itself, industrial trans fats. Furthermore, in the EU such a measure has no precedent. Public authorities perceived this option to be somewhat appropriate, while consumers are very supportive of option 3b as it ensures a high level of health protection.

Combinations of options

In addition to the above options, the following combinations of some of the options were considered:

·Combining mandatory labelling with legislation (2 + 1b or 2 + 3b)

·Combining mandatory labelling with voluntary agreements (2 + 1a or 2 + 3a)

5.3.Options discarded at an early stage

Fiscal measures, for instance introducing taxes, are proposed as effective measures for addressing nutrients of public health concern that are over consumed in a population. Examples are sugar taxes addressing sugars levels in sweetened beverages. However, industrial trans fats are seen from a food safety perspective, where fiscal measures are less appropriate. Furthermore, the Commission report of 2015 identified already the introduction of a legal limit as the most effective measure and announced the present IA to assess further its impact.

Sub-options of Option 1b with specific requirements for low fats have not been considered. Four of the seven Member States with national legal limits apply different limits for lower fat products. In view of an EU level legal limit, a legal limit of 2 % on fat basis is in line with EFSA and WHO recommendation, seems achievable in practice and is generally accepted by both consumer organisations as well as health NGOs on the one hand, and industry on the other hand. Granting additionally higher levels for low fat products, that are forming the major part of the diet in terms of quantity, could potentially lead to intake levels above 1 % of energy intake. For example, 100 g of a food with a fat content of 3 % and a maximum level of 10 % industrial tans fats of the fat content, would lead to products with 0.3g of industrial tans fats per 100 g of the final food product. For a person consuming a 2000 kcal diet per day, the standard used in EU food law and Codex Alimentarius guidelines, the WHO recommendation of less than 1 % of energy intake corresponds to less than 2.2 g TFA per day. Consuming more than 730 g of foods that are at this threshold would lead to industrial tans fats intakes exceeding the WHO recommendation. Generally, adults consume more than 730 g of food per day. The sub-options were not in detail discussed in the IA report. However, during discussions and consultations of a draft measure well justified proposals could be considered.

6.What are the impacts of the policy options?    

Annex 12 explains how options were screened against possible relevant impacts and how relevant impacts were identified. The main impacts of the policy options described in section 5 which were identified during the screening phase and therefore analysed in detail are: social impacts (health benefits, quantified in terms of health care costs - direct and indirect - and disability adjusted life years; effects on health inequalities), economic impacts (direct costs for businesses and public authorities which consist of administrative burdens for business, compliance costs for business, including the role of innovation and technological development and administrative burdens for public authorities; consumer impacts – prices, choice and product quality; Single Market impacts; effects on international trade; impacts on SMEs) and environmental impacts, particularly in relation to deforestation and implications for climate change and biodiversity. The potential indirect effects of the above on competitiveness, growth and social cohesion were also considered in the analysis.

6.1.Social impacts

The impacts of the options on human health are quantified in terms of direct and indirect health care costs and disability adjusted life years.

These impacts are influenced by the level of industrial trans fats intake currently observed in the population, which varies as a result of the different policy options. Furthermore, dietary habits of certain population subgroups, consumption levels of ruminant trans fats and the type of fat used to replace of industrial trans fats in reformulated products will also contribute to potential health impacts. In this analysis all factors are assumed constant. 130

Impacts on health

The health impact assessment used a number of assumptions that, together with the underlying evidence for those assumptions, are described in detail in Annex 13.

Food policies have the potential to reduce non-communicable disease mortality and morbidity, with associated cost savings quantified in Table 1 and associated health gains expressed in disability adjusted life years averted, quantified in Table 3 for the different policy options assuming three variants of the baseline scenario.

Impact on health care costs (direct and indirect)

Both direct and indirect health-related cost estimates are expressed in 2016 prices (in €). The model of the contractor ICF considers two types of costs, both were based on the European Cardiovascular Disease Statistics 2012 131 :

·Direct healthcare costs: costs related to the use of health resources (i.e., primary care costs, outpatient costs, emergency costs, and medication used during the hospitalization).

·Indirect costs of ill health: costs related to the disease, namely loss of productivity and informal care.

In the case of no EU action (Option 0) all health-related costs for the EU over the course of a lifetime (85 y) have a present value 132 of €10,764,979 million under the 10 year elimination variant (B1). Under the 15 year elimination variant (B2) and ‘no change’ variant (B3) the present value of total health-related costs would be €33,753 million higher and €245,009 million higher respectively.

Table 2 shows the cost savings resulting from each policy option as compared to the baseline scenario variants. The figures are calculated by subtracting the costs associated with the disease burden expected under the given policy with that of the relevant baseline variant.

Options 1b and 3b deliver the highest health-related cost savings; the implementation of legal measures (1b or 3b) would lead to savings with a present value of €58,611 million under variant B1 and €94,008 million under variant B2. In variant B3 disease-related costs savings are much greater than under the other two variants. In B3 there is no reduction of industrial trans fats intake without an action at EU level.

Table 2 Cost savings associated with lower disease burden for each policy option compared to the baseline, under each of the baseline scenario variants (M EUR)

Note: Figures represent the reduction of health-related costs over 85 years, in present value terms, in million Euro

A sensitivity analysis 133 has been conducted to show the impacts of alternative specifications of the starting point – i.e. the initial population industrial trans fats intake when the model starts (point 0). This shows that the results are robust, all options deliver cost savings in all cases, and options 1b and 3b provide the largest benefits.

Impact on disability-adjusted life years

The disability-adjusted life years measure overall disease burden as the number of years lost due to ill health, disability or early death. Resulting disability-adjusted life years are then calculated on the basis of the modelled number of coronary artery disease events and deaths.

In the case of no EU action (option 0) the disability adjusted life years for the entire EU population amount to 1,076 million over the course of a lifetime (85 years) under the best case scenario. Under variants B2 and B3 the total EU coronary artery disease burden in disability adjusted life years would be 1,079 million and 1,142 million respectively.

Table 3 illustrates the number of disability adjusted life years avoided thanks to the implementation each option as compared to the baseline scenario variants. They are calculated by subtracting the estimated disability adjusted life years in the baseline from the disability adjusted life years in the given policy.

Options 1b and 3b lead to the highest reduction in morbidity and mortality (as measured in terms of disability adjusted life years). The implementation of legal measures (1b or 3b) would reduce the disease burden by 4 million disability adjusted life years for the EU population under variant B1 and by 6 million disability adjusted life years for the EU population under B2. In the B3 case the reduction in disease burden is much greater. Options 1b and 3b have the greatest positive impact.

Table 3 Health gains in disability adjusted life years averted (EU28, Millions) for each policy option compared to the baseline, under each of the baseline scenario variants

A sensitivity analysis 134 shows that results are robust, all options reduce the disease burden as compared to the baseline.

Options 1b and 3b have identical expected health benefits. The underlying assumptions explain this similarity. With regard to option 1b, evidence from Denmark suggests that the introduction of legislation limiting the industrial trans fats content of foods was very effective in reducing the population intake. Since the introduction of the measure in 2002, the average intake of industrial trans fats decreased in all age groups of the Danish population. The most recent data suggest that in 2014 the average industrial trans fats intake in Denmark was 0.009 % of the energy intake. Based on this evidence, the health model assumes that for options 1b the industrial trans fats intake decreases to 0.009 % of energy intake after two years (assumption of 2 year implementation period) and then evolves as assumed in each of the three baseline scenarios. With regard to option 3b, introduction of a ban on the use of partly hydrogenated oils as a food ingredient through EU legislation, with a transition period of 2 years, the model assumes that industrial trans fats intake will vary as in option 1b, for instance, that the removal of partly hydrogenated oils from the food supply will successfully eliminate the presence of food with high industrial trans fats content from the market and lead to trans fats intake decreases to 0.009 % of energy intake. Residual small industrial trans fats intake from deodorised oils and trans fats generated during the heating of oils during cooking will remain.

Replacement of industrial trans fats with other ingredients as fat sources could potentially have unintended consequences for health. There is a range of approaches to reformulate foods and reduce industrial trans fats content for example replacing partly hydrogenated oils with alternative oils and fats and/or mixing of various non-hydrogenated oils. The ‘toolkit’ of oils and fats is vast and includes for example food technological approaches to ‘design’ fats of desired composition and properties, in particular by applying fat interesterification and fractionation processes Some stakeholders highlighted during the OPC the need to ascertain that the full health profile of the reformulated product has to be considered; for example there are concerns that reformulation may lead to increased saturated fat content. However, several studies have shown that for a number of food products, industrial trans fats have not simply been replaced by saturated fats, but the reformulated products have increased the content of cis-unsaturated fats, thus leading to an overall healthier profile of the product. Even if industrial trans fats were replaced completely with saturated fat, a net health benefit would result.

The recently launched REPLACE package of WHO 135 asks for encouragement of manufacturers to replace industrial trans fats with the most healthy available alternative fat. Many Member States already work on voluntary reformulation campaigns with industry to replace saturated fat intake. The EU is supporting such efforts, both via research projects or by support to exchange best practice models of reformulation.

One of the potential replacement fats for partly hydrogenated oils is palm oil and potential health implications of such a replacement need to be considered. Palm oil contains various fatty acids that could be considered in relation to their health profile. Around 49 % stem from saturated fat, 37 % from monounsaturated fat and 9 % from polyunsaturated fat. Saturated fat intakes are increasing the risk for developing heart disease and their intake should be limited. However, even if industrial trans fats are completely replaced by saturated fat, a net health gain is achieved. Replacing trans fats with either mono- or polyunsaturated fats yield higher health gains. Palm oil is particularly rich in palmitic acid, approximately 44 % of the fat. This may be seen as problematic as palmitic acid has been reported to be linked to bigger effects on increasing the undesirable LDL blood cholesterol levels and therefore with higher risks in relation to heart disease than other saturates fatty acids. EFSA has stated with regards to individual saturated fatty acids, that the 4 major saturated fatty acids (lauric, myristic, palmitic, stearic acid) may have different effects; however, that data is not sufficient for setting intake recommendations for individual saturated fats.

Impacts on health inequalities

Inequalities in health remain an important issue in the EU and across the globe. While the model of the JRC does not produce quantitative estimates of the potential effects of options on health inequalities, evidence collected by ICF from the implementation of trans fats policies and other dietary policies across the world suggest that the legal limit would be the most effective in reducing health inequalities, followed by the voluntary reformulation. The labelling policy is likely to have a minimal effect upon reducing health inequalities, and could in some populations actually worsen health inequalities.

Consumers with lower income are more likely to consume products with high industrial trans fats content, products that are generally sold at a lower price. As such the current situation can contribute to health inequalities. Another population group at risk off high industrial trans fats intakes are younger population groups. Examples were young males in Germany and Austria, identified as population groups at risk if high industrial trans fats intakes as they consume a high proportion of processed foods and fried fast foods that was found to contain more likely to contain high industrial trans fats levels. For instance, in Austria, young apprentices were identified as a group at-risk population to exceed recommended intake levels of industrial trans fats due to their high consumption of fast foods. Before Austria introduced their national legal measure, in a study covering 2989 young apprentices 136 , 75 % were consuming levels below the national recommended level (1 % of energy intake), 25 % were above this intake level. In Germany, a report by the Federal Institute for Risk Assessment stated in 2013 137 that young adults were found to consume approximately 2,5 times the amounts of industrial trans fats as compared to older adults. While most consumers achieve trans fats intakes below 1 % of energy intake, the average intake is 0,66 % of energy intake. However, even with this low average level, 10 % of the population is above the recommended level of % of the energy intake. Particularly voluntary reductions of ready meals, here deep frozen pizzas, have contributed to reducing high intakes of an identified at-risk population of young adult males.

Detailed considerations and expected impacts for all policy options are provided in Annex 15.

6.2.Economic impacts

Each of the policy options considered has the potential to have a number of economic impacts, most importantly benefits and costs: administrative costs are incurred by businesses in understanding the rules, determining responses and providing information, and by the public authorities in implementing and enforcing the rules, monitoring and reporting. Compliance costs are incurred by businesses in meeting the legal obligations or voluntary commitments. These may include the costs of reformulating products, purchasing alternative ingredients, and product labelling. Further economic impacts were considered as well.

Economic impacts have been assessed by ICF with a cost model developed in MS Excel in parallel to the JRC model. The analysis provides a quantitative assessment of administrative and compliance costs for business, and administrative costs for public authorities. Quantitative estimates of the costs borne by SMEs were also made

Additional evidence collected from the consultations, informed a more qualitative assessment of related impacts on consumers, the Internal Market, competitiveness and international trade. Evidence and data on price impacts, competitiveness, the Internal Market and international trade was not available to enable a quantitative analysis. Available data and empirical evidence, while valid in specific cases in a certain context informed the evaluation, however, it was not possible to extrapolate such data for a quantitative assessment with a sufficient degree of confidence in the robustness of the results.

In order to assess costs for food business operators, the market structure needs to be analysed. For all businesses in relevant food industry subsectors that are potentially affected by the measures, Annex 16 provides an detailed analysis on the number of businesses affected and how the different measures that create costs are concerned. Based on EUROSTAT data, policy option 1a affects according to the ICF estimates 117,918 businesses, option 1b 1,019,240 businesses, option 2260,397 businesses, option 3a 124,403 businesses and option 3b 1,081,514 businesses. With regard to the different sectors, around 85 % are in food service (such as restaurants or caterers) and 15 % in the food manufacturing sector. The food sector, in terms of number of businesses, is dominated by SMEs. In food manufacturing, approximately 99.1 % of businesses are SMEs, of food service approximately 99.9 %, respectively.

Impacts on direct costs for businesses and public authorities

Table 4 presents estimates of the total costs to business and the public authorities of implementing the five options, as compared to the baseline scenario. The detailed costs that are summarised in this table and the underlying assumptions and methods for establishing them are described in Annexes 16 to 18. The figures present the sum of the present value of costs over 10 years, using a discount rate of 4%. 138

Costs are assumed to be zero after 10 years for each option. Many are one-off costs such as reformulation or relabelling costs. It is assumed that monitoring and enforcement will cease to generate costs after 10 years (by which time industrial trans fats will have disappeared from the food chain). By that time, monitoring of foods for the presence of industrial trans fats would likely become part of the routine operations carried out by National Competent Authorities regarding food composition. The development of cost-effective alternative ingredients should be followed by a decrease in costs for the substitute ingredients over time

The present values are calculated by summing the different estimated costs incurred each year over the 10 year period, and calculating the present value of these using the 4% discount rate. These costs are then summed up over the 10 year period to give a total present value.

Table 4 Present value of total costs of implementing options over 10 years (M EUR)

Option 2 is estimated to be linked by far to the largest costs, especially as a result of the costs of relabelling of food products, whether or not they currently contain or are likely to contain trans fats.

Options 1b and 3b are estimated to have significantly larger costs than 1a and 3a, because a greater level of business action is anticipated in response to legislation than voluntary initiatives.

The estimated costs represent a small proportion of the annual value of EU output of the business sectors affected (Table 5).

Table 5 Estimated costs as a proportion of the value of output of affected food business subsectors (%)

Note: Figures are expressed as a % of output of the main sub-sectors affected by action for industrial trans fats. 139 *Costs of option 2 include costs for all pre-packaged food producers.

While the cost estimates are based on broad averages and assumptions, it is likely that the costs for the majority of food businesses will be minor, but that a small proportion of businesses will face greater challenges and costs. Examples of businesses that may face greater challenges and costs are those suppliers of oils, fats and margarines that have not yet reformulated their products, as well as a number of smaller bakeries across the EU that are currently users of partly hydrogenated oils.

Impacts on consumers

The main impacts on consumers (besides health-related impacts discussed in section 6.1.) are expected to be:

·Possible increases in the price of food products; and

·Possible changes in the attributes of food products, including their taste and texture.

Consumer prices

Increases in costs to food businesses could (partly) be absorbed within the food chain, (resulting in lower business profits), but would be expected to be reflected, at least partly, in increases in the price of food products to the consumer.

The expected impact of each option on consumer prices is summarised in Table 6, details about the underlying assumptions to establish the expected impact on consumer prices are provided in Annex 19. Only qualitative data are available as it was not possible to quantify the increase of consumer prices.

Potential price increases on items consumed primarily by low-income citizens could lead to less available budget for food purchase (fruit and vegetables) and therefore have the unintended impact of leading to less healthy diets. However, the price increases for most food items is assumed to be moderate. Furthermore, most substantial price increases are expected to occur in foods with cocoa, vegetable fat coatings. Those foods are generally recommended to be consumed in smaller quantities and price increases could also lead to smaller consumption quantities or to less frequent consumption. The total net impact on health is uncertain.

Table 6 Expected impact of each option on consumer prices

Product attributes

One of the challenges in reducing industrial trans fats is the difficulty of finding alternative ingredients and formulations that allow products to offer a similar experience to consumers in terms of their taste, texture, appearance and shelf-life. If these challenges cannot be adequately addressed, there is a danger that the satisfaction that consumers derive from affected food products will be adversely affected. Also, consumer choice could be affected when products would be taken from the market as reformulation is not possible.

Overall, evidence presented in the ICF study suggests that these issues do present challenges for some sectors of the food industry, but that these challenges are not insurmountable, also considering that products were produced before the wide introduction of trans fats in the middle of the 20th century. Some evidence collected by the external contractor ICF is provided in Annex 20. Options 1b and 3b – by mandating changes in product content – can be expected to have greatest potential impacts (Table 7). In a view of evidence where reformulation may prove difficult and result in the possible disappearance of some food items and/or a loss of product variety, tit would be premature to conclude that product choice available for the consumer may be lost. As an example, the introduction of an industrial trans fats legislative limit in Denmark resulted in a reduction of industrial trans fats shortly after (in one year) its introduction without any obvious side effects for the population. Also in Austria and Hungary, no reduction of product choice available for the consumer was observed. There remain some uncertainties as the proposed measures would cover the entire EU.

Table 7 Expected impact of each option on product attributes

Internal Market impacts

Differences in product standards between Member States can distort the free movement of goods within the EU. National rules may impose higher costs on national operators, affecting competition in the market as a whole. They may also restrict access to domestic markets for producers in countries which do not adhere to the same standards.

In the absence of legal action at EU level, future national actions are likely, leading to further differences in standards across the EU. Further evidence with regard to Internal Market impacts is provided in Annex 21.

Significant differences between the options can be expected, with Options 1b and 3b having a significant harmonising effect. The voluntary options 1a and 3a would seek to raise standards across the EU, without affecting the legal framework. There is a risk that varying rates of progress and uptake of voluntary agreements could have a complicating effect and lead to further differences between countries and sub-sectors. Option 2, relating to labelling, would have no effect in harmonising product standards, but would aim to encourage consumers to make more informed choices. Options 3a and 3b, by focusing on eliminating partly hydrogenated oils rather than placing limits on industrial trans fats, would introduce differences compared to existing legislation in the mentioned seven Member States. This would potentially create confusion in the market and requiring some further action to harmonise standards at national level. Annex 21 provides a summary table and qualitative assessment of expected impact of each option on the Internal Market.

Competitiveness and trade impacts

A number of non-EU countries have introduced legal limits on industrial trans fats in food or banned the use of partly hydrogenated oils in food products (Canada, US). The majority of countries globally have yet to introduce legislation on industrial trans fats.

EU policy on industrial trans fats has the potential to impact on international trade in food products:

·Elimination of industrial trans fats from the EU food chain will help to position EU producers to sell to markets such as Canada and the US, as far as they are accessible, which have taken action to limit partly hydrogenated oils/ industrial trans fats;

·Limiting industrial trans fats use, by increasing costs for food businesses, could potentially hamper competitiveness in price sensitive export markets;

·Legal limits on industrial trans fats /partly hydrogenated oils applied to products sold in the EU would apply to foreign imports as well as domestic production, potentially reducing imports from countries that have not acted to reduce industrial trans fats;

·Voluntary measures could potentially increase costs for EU producers, while exposing them to competition from low cost, high foreign trans fats imports;

·Labelling measures would apply equally to imports and domestic products sold in the EU.

The net effect of these potential impacts is difficult to predict, and will vary between the different options.

Little evidence was found from the literature review of the ICF study to suggest that impacts on trade and competitiveness are likely to be significant, and in general the stakeholders interviewed by ICF did not express this as a concern. This is likely to be because:

·Extra-EU trade represents only a small proportion of the market for most of the industrial trans fats relevant food industry subsectors;

·Most companies active in international markets have already taken action to eliminate industrial trans fats from their products; and

·Any additional costs involved in eliminating industrial trans fats are a small proportion of industry output (as estimated above), such that the presence or absence of limits is unlikely to be a major factor influencing competitiveness.

Where consultees in the ICF study commented on trade issues, a general view was that action to eliminate industrial trans fats from food is taking place internationally, and that taking action on industrial trans fats will tend to enhance rather than reduce competitiveness. Pressure to reduce trans fats levels in foods and related legal measures is expected to increase worldwide in view of the plan to eliminate industrial trans fats from global food supply published by WHO on 14 May 2018. 140 Overall, the expected impact of all the options is small, further details are provided in Annex 22.

Impacts on SMEs

The EU’s food and drink industry is a highly diversified sector with many companies of different sizes. It includes more than 280,000 SMEs which generate almost 50% of the sector’s turnover and value added and provide two thirds its employments. 141 SMEs are particularly prevalent in particular subsectors – such as bakeries and food service – which face greater challenges in reducing industrial trans fats.

Little specific evidence was found in the ICF study through the literature review or stakeholder interviews or the OPC 142 regarding the particular impact on SMEs resulting from action to address the industrial trans fats issue. However, interviewees expressed a general view that SMEs may be impacted by the different policy options on the grounds that:

·SMEs are in general less likely than their larger counterparts to have taken action to eliminate industrial trans fats from their products; and

·SMEs generally have less staff time and fewer resources to devote to product development, and therefore may face greater challenges to reformulate their products

On the other hand, evidence collected by ICF also suggests that many SMEs will benefit from action by their suppliers to reformulate ingredients and this will provide simple routes to compliance with limits on industrial trans fats. For example, many small bakeries will simply use alternative fats and oils developed by larger firms that supply ingredients to the baked goods sector. Substitute frying oils have been developed for use by food service businesses. 143 Micro-businesses, which are prevalent in the food service sector, are likely to make use of these supply chain solutions and may, as a result, incur smaller costs than businesses from the food manufacturing sector. It should be noted, however, that the size of business is not necessarily correlated to the nature and size of the costs borne. Also, evidence from Canada was found by ICF that SMEs were able to follow reformulation activities of large multi-national companies. There was a tendency for SMEs to copy these reformulated products rather than investing in own research and development. As a result, the measures were not as costly to SMEs as may be assumed.

With regard to a considerable part of the SME food business operators in the EU, the hospitality industry, empirical evidence collected by ICF points to the fact that a legal limit on industrial trans fats has a rather limited impact or non-existent impact. The: industrial trans fats contained in meals prepared by hospitality businesses are only the result of the content of such trans fats in supplies bought from the processing industry. If the supplies are already below the limits, food prepared by hospitality businesses will always be below the limits.

Some further evidence on the potential impacts on SMEs is provided in Annex 23.

Overall, the evidence collected in the ICF study suggests that:

·SMEs will bear a significant proportion of the costs identified above, particularly because of their prevalence in the affected sub-sectors, and the tendency for SMEs to have been less active to date in reformulating their products;

·Many SMEs will be able to eliminate industrial trans fats by accepting alternative ingredients developed by their suppliers, and will therefore not face significant costs;

·Those SMEs forced to reformulate their products will face additional costs and may experience greater challenges than larger companies because of their limited resources for R&D. For many small businesses, reformulation may be relatively simple, and require a few hours’ work to test an alternative recipe. The greater impacts will be on those SMEs facing more complex and costlier reformulation.

·The impact of the measures is likely to be greater for SMEs operating in the food manufacturing sector rather than SMEs operating in the food service sector.

The alternative options will have different impacts on SMEs:

·The legal options (Options 1b and 3b) will require all SMEs currently with non-compliant products to take action, potentially imposing significant costs on some;

·The mandatory labelling Option (Option 2) will place similar obligations on SMEs and larger companies. SMEs should be familiar with labelling obligations so should not face particular technical barriers. However, some SMEs may face greater difficulties in absorbing the additional costs involved;

·SMEs which face challenges in reducing industrial trans fats may choose to opt out of a voluntary agreement (Options 1a and 3a). These options are therefore likely to have least impact on SMEs;

Table 8 provides a summary table and qualitative assessment of expected impact of each option on SMEs. A specific SME test is provided in section 7.8, informing about the average cost per SME for the different options. Transition periods will help to mitigate the above mentioned costs. Empirical evidence from a Hungarian SME active in the chocolate confectionary sector demonstrates that adaptation to legal limits (Option 1b) is possible, however, sufficient transition periods, in the specific case between 1.5 to 2 years, are crucial as longer transition periods mitigate cost burden of the necessary adaptations. 144   145 Furthermore, larger food business operators that have removed trans fats from their portfolio, as well as food business associations have committed to guide companies, particularly SMEs, that have not taken action through the process of removing trans fats from all foods in order to meet a legal limit21 22.

Table 8 Expected impact of each option on SMEs

6.3.Environmental impacts

Measures to reduce the use of industrial trans fats have potential impacts on the environment, by altering the use of ingredients and production processes. The primary concern raised in studies to date, and mentioned by interviewees of the ICF study, relates to the substitution of palm oil, a trans fats free, semi-solid fat, for partly hydrogenated oils, and the potential of increased palm oil production to cause deforestation.

The extent of such impacts depends on:

·The degree to which palm oil – as opposed to other possible ingredients – is used as a substitute for partly hydrogenated oils, and hence the extent to which limits on industrial trans fats production result in increased demand for palm oil;

·The degree to which any increase in palm oil demand results in environmental damage, which depends on the sustainability or otherwise of the production systems;

·The relative environmental impacts of palm oil compared to the oils that are partly hydrogenated (typically soy) and alternatives.

Some qualitative evidence in relation to substitutes for partly hydrogenated oils and their environmental impacts collected by the external contractor is provided in Annex 24.

Possible impacts of alternative options

Overall, the situation is complex and the resulting environmental impacts are difficult to predict. It is clear that:

·Palm oil is an attractive substitute for partly hydrogenated oils, particularly in the baked goods sector, on account of its physical properties and cost-effectiveness;

·It is therefore likely that limits on industrial trans fats will lead to increases in use of palm oil in products currently using partly hydrogenated oils. Overall consumption of palm oil in the EU will not necessarily increase, as it is forecasted to decline in the food sector as a whole, although global demand is growing;

·Increased use of palm oil is of concern since it has contributed to deforestation, with adverse impacts on biodiversity and climate;

·The EU is a leading player in the development of markets for sustainable palm oil. There is currently an excess supply of sustainably certified palm oil and any increase at EU level resulting from limits on partly hydrogenated oils could be met from sustainable sources, if consumers were willing to pay a price premium;

·As a result, action on industrial trans fats need not necessarily have an adverse environmental impact. However, there are no guarantees that any palm oil used to replace partly hydrogenated oils would be sustainably sourced; adverse impacts on biodiversity and climate are therefore a risk;

·However, the use of other vegetable oils such as soy also contributes to deforestation, and it is likely that current use of partly hydrogenated oils in food in the EU already impacts adversely on biodiversity and climate. The net effect of any change towards palm oil is difficult to assess. One advantage of palm oil is that it produces a high yield of oil per hectare compared to alternatives;

·Any potential negative impacts on the environment can be mitigated by further action by the EU food industry to ensure that palm and other oils are sustainably sourced.

It is therefore unclear whether or not any net impact on the environment as a result of action to reduce industrial trans fats will be positive or negative. However, it is clear that the magnitude of any environmental impact will be greater for those options leading to greater change in industrial trans fats. On this basis, options 1b and 3b can be expected to lead to greater environmental changes than Options 1a, 2 and 3a. Annex 24 provides a summary table and qualitative assessment 146 of expected impact of each option on the environment.

6.4.Impacts of combined options

Any additional benefit of adding labelling requirements to a legal limit on industrial trans fats or a ban on partly hydrogenated oils is expected to be limited as population industrial trans fats intake will already be reduced to very low levels under Options 1b and 3b.

Combining a voluntary agreement with labelling may be expected to have a higher impact in reducing the population industrial trans fats intake and will lead to greater cost savings and disability-adjusted life years reduction than adopting only one of the two options. However, according to estimates by ICF, details are provided in Annex 25, these benefits are significantly less than those delivered by Options 1b and 3b.

Because all of the combinations of options include Option 2, which has high costs of relabelling, product testing and awareness raising, each combination of options also has high costs. Therefore, even though combining voluntary agreements with mandatory labelling is estimated to lead to additional benefits (while remaining at levels significantly below Options 1b and 3b), the costs are high compared to Options 1b and 3b, as a result of the high relabelling and promotional costs of Option 2. Details are provided in Annex 25.

7.How do the options compare?    

This section considers how the options compare in the expected performance against the stated general and specific objectives and how the options compare in effectiveness, efficiency, coherence and with reference to the proportionality principle.

Of note, with regard to the validity and reliability of modelling results, a number of uncertainties need to be highlighted in order to avoid a false impression of scientific accuracy. Overall, there are limitations of the ICF modelling exercise due to the assumptions needed, data scarcity linked to intakes and future projections, paucity of evidence related to other trans fats health effects, possibilities to model more complex dietary changes making strong simplification necessary. The main purpose of the model was to support with modelling the relative comparison of the viable policy options against a reference of no policy; this outcome of a legal limit performing better under this specific framing of a public health economic evaluation in terms of health benefits and cost-effectiveness has been shown to be robust. Nevertheless, the relative findings are based on past experience. There is inevitably uncertainty how the future trans fats intakes might develop under the alternative policy scenarios. Annex 4 provides additional explanations about uncertainties.

7.1.General objective 1: Ensuring a high level of health protection for EU and Specific objective 1: Reduce intake of industrial trans fats in the entire EU for all population groups

Direct health impacts

The direct health impacts for EU citizens are positive under all options relative to all variants of the baseline scenario. The benefits of prompt action are strongly amplified if, in the baseline scenario, industrial trans fats intake does not decline. If, without further EU intervention, industrial trans fats would be phased out 10 years through industry actions then adopting options 1b or 3b could save around 4 million disability-adjusted life years that would otherwise be lost to coronary artery disease. If, however, industrial trans fats levels were to otherwise persist at current levels then legislating to remove them would conserve 66 million disability-adjusted life years.

The legal options (option 1b, 3b) deliver larger benefits than the voluntary agreements (option 1a, 3a) and labelling option (option 2). There is also a much higher degree of confidence that the legislation will deliver positive results – there is significant uncertainty about whether food business operators that are still placing products high in industrial trans fats on the market will participate in voluntary agreements, and how far consumers will respond to a modification of the nutrient declaration that adds reference to products’ trans fats content. In that context, the figures for options 1a, 3a and 2 in the Table 56 and Figure 13 in Annex 26, 1. may be regarded as upper estimates of potential impact. 147

Health benefits are expected to follow close behind the action taken by food business operators to reduce industrial trans fats. Experience from countries that have acted suggests that signalling that action is going to be taken can result in benefits starting before the legislation comes into force as some producers take proactive action in advance of the deadline.

The health impacts of derogations providing for authorised use of industrial trans fats for technical applications in low fat products under option 1b or partly hydrogenated oils under option 3b are uncertain.

Direct and indirect economic impacts of changes in health status

All options deliver savings in direct and indirect economic costs of industrial trans fats-related disease. These comprise changes in:

·Healthcare expenditure: This is a benefit that accrues principally to healthcare service providers and hence governments (where healthcare is publicly funded) or health insurers. Some of the benefits would accrue indirectly to citizens, whether as taxpayers or purchasers of health insurance.

·The wider economic impact of the changes to health status and coronary artery disease incidence triggered by EU industrial trans fats policies, focusing specifically on changes in productivity and in changes in demand for informal care. Productivity changes will accrue initially to employers and then to the economy as a whole. Changes in demand for informal care will impact directly on carers and may have a wider impact on economic output (e.g. where someone is able to continue in work because the incapacity of a family member due to coronary artery disease is avoided).

The analysis, using the model of the JRC, has calculated the present value of benefits over an 85 year horizon. 148 In baseline variants B1 and B2 industrial trans fats would be phased out after 10 and 15 years respectively so industrial trans fats would not be causing new and additional health impacts after those dates. In variant B3 industrial trans fats intake continues to cause negative health impacts in the baseline scenario in perpetuity so the options that reduce intake avoid a long stream of health impacts. The monetary benefits under B3 are therefore substantially larger than under the other two variants (Annex 26, Table 57). 149  

The analysis shows that the uncertainty in the baseline is not grounds for inaction – the slower the phase-out of industrial trans fats in the baseline, the greater the health impacts of effective EU action increase. The model is constructed to work at EU level, with reference to the EU population and EU-level cost factors taken from third party sources.

The legal options (1b, 3b) deliver larger benefits (cost savings) than either the voluntary agreements (1a, 3a) or the labelling option (2). The assumptions in the model (whereby the residual industrial trans fats intake under a partly hydrogenated oils ban is the same as the intake under a 2% limit) mean that 1b and 3b are equivalent in the healthcare savings delivered and deliver much larger savings than the alternatives. If option 1b was applied to ingredients as well as final products it would have the effect of implementing a partly hydrogenated oils ban of the kind specified in option 3b. It seems likely that this would deliver additional health benefits, but the information required to estimate those effects are not available.

As with the human health benefits, there is a much higher level of confidence that the legal options will deliver the scale of benefits indicated – there are significant uncertainties attached to the estimate of benefits of the voluntary agreements and labelling, and the values indicated are likely to be upper limits. This assumes compliance by food business operators with the legislation, which should be complemented by effective communication, by monitoring and enforcement by regulators.

Combined options are also considered:

·Combining mandatory labelling with legislation is not expected to yield significant additional health benefits over and above those delivered by Option 1b or 3b. There are theoretical direct and induced effects arising from consumers having a preference for industrial trans fats content closer to zero than the 2% legislated threshold, however the labelling option may also lead to adverse effects and heightened social inequalities.

·Combining mandatory labelling with a voluntary agreement is expected to yield additional benefits in terms of further avoided health-related costs, through synergistic effects, estimated at EUR 19,248 million for the combined option as compared to EUR 11,078 million for Option 1a and 3a and EUR 15,353 million for Option 2.

The performance of options against the specific objective 1 mirrors that for General Objective 1 described above, as well as for General Objective 3 on health inequalities. The performance of each option is summarised in Table 58 presented in Annex 26.

7.2.General objective 2: Contribute to the effective functioning of the Internal Market for foods that could contain industrial trans fats and Specific objective 2: Ensure that the same rules/conditions apply in the EU to the manufacturing and placing on the market of foods that could contain industrial trans fats, so as to ensure legal certainty of EU food business operators within and outside the EU

The legal options (Options 1b and 3b) impose a uniform approach across all entities that place food on the market across the EU.

Option 2 would provide a consistent level of visibility for consumers of industrial trans fats content in products but not provide consistent protection against the health impacts of high industrial trans fats products for those not aware of the risks. As it does not set limits for industrial trans fats content, it would also not fully address legislatively-driven cost differentials between producers in national markets where limits on trans fats content apply and producers from other countries.

With full participation and if fully effective the voluntary agreements (option 1a, 3a) approximate to the effects of legislation in their consequences for the Internal Market, but the evidence collected by ICF suggests that participation will be at best partial.

Options 3a and 3b, which aim to eliminate partly hydrogenated oils rather than place limits on industrial trans fats, would introduce differences compared to existing legislation in the seven Member States, potentially creating some confusion in the market and requiring some further action to harmonise standards at national level.

There are also potential indirect effects of non-legislative action in so far as, in the absence of EU legislation, there are some indications that certain Member States may adopt national legislation that varies in specification from those already in place and adds to the emerging legal complexity in this aspect of the market. Table 9 summarises the options’ performance against this general objective.

Table 9 Appraisal of options’ performance under general objective 3: Contribute to the effective functioning of the Internal Market for foods that could contain industrial trans fats

Note: scale of - - to + + indicates a range of strongly negative (- -) to strongly positive (+ +) impacts, with ‘0’ being neutral.

The results for the specific objective 2 mirror those for General Objective 3 described in the table above. The options vary in the number of food business operators directly affected. These differences are determined by the sectors engaged (e.g. non-packaged goods are excluded from Option 2) and the level of participation expected. An important qualifying comment is that most of those subject to legislation will not need to act to reformulate products because their products do either not or not anymore contain industrial trans fats. There is uncertainty about the number of firms that will engage in the voluntary agreements.

Option 1b provides full and immediate legal certainty. Option 3b provides general legal certainty but creates challenges for those Member States that have already legislated and adopted the 2% limit model rather than a partly hydrogenated oils ban. These countries would need to adjust their domestic legislation to fit the EU model.

The other options provide less certainty in that there is the potential for unilateral Member State legislative action in countries that want to go further than Options 2 or 1a/3a provide for. 150

7.3.General objective 3: Contribution to reducing health inequalities, one of the objectives of Europe 2020

The legal options (1b, 3b) could potentially remove all present industrial trans fats-related health inequalities, which is not the case for the alternative options. All food consumers would benefit irrespective of social-economic, demographic status or consumption patterns. The impact of the alternatives is constrained by the limits to engagement by food business operators that have not already acted, and limits to responsiveness of consumers to trans fats-related additions to the back-of-pack nutrient declaration.

The performance of each option is summarised in Table 39 in Annex 15.

7.4.Effectiveness

Effectiveness is measured by the extent to which options are expected to achieve the target objectives, the three general objectives.

The main findings relevant for assessing the effectiveness of each option in achieving these objectives are specified in Table 9.

In relation to the health impact, the used model considers only coronary artery disease. Other potential benefits of lowering trans fats intake, which have been referred to in the literature such as impacts on insulin sensitivity, obesity, diabetes, cancer, or early growth and development, are excluded because of inconsistent evidence 151 and lack of data. As such the impact assessment can be considered to be conservative with respect to achievable health benefits resulting from (fast) industrial trans fats removal from the food supply.

Table 10 Effectiveness of all options and combinations of options under variant 2 of the baseline scenario

Note: scale of - - to + + indicates a range of strongly negative (- -) to strongly positive (+ +) impacts, with ‘0’ being neutral.

Options 1b and 3b would be the most effective, in that they would achieve the greatest improvement in terms of health protection, reduction of health inequalities and contribution to the functioning of the Internal Market.

Option 2 would also prove effective in improving the level of health protection for EU consumers; however the assessment does not suggest that it would be effective in addressing health inequalities nor the current imbalances and fragmentation of the Internal Market in this area.

Options 1a and 3a would be less effective than other options in achieving a high level of health protection for EU consumers, and would contribute less than Options 1b and 3b to reducing health inequalities. Since voluntary agreements would be heavily dependent on the level of organisation of the food industry, they are unlikely to achieve any significant results in terms of addressing the fragmentation of the Internal Market on the matter of industrial trans fats.

The combination of Options 1a and 3a with Option 2 offers potential to provide greater health benefits and reductions in inequalities than these options alone, but does not offer added benefits with respect to the Internal Market. Combining Option 2 with Options 1b and 3b does not enhance effectiveness compared to Options 1b or 3b alone.

7.5.Efficiency (balance of costs and benefits)

The analysis has provided quantitative estimates of the administrative and compliance costs for businesses and public authorities, as well as the social benefits in terms of reduced costs of healthcare. Other relevant costs and benefits, including those relating to health inequalities, the Internal Market, consumers, international trade and the environment, have been assessed qualitatively.

Because some effects have been assessed in qualitative terms only, a comprehensive cost-benefit analysis is not possible. However, it is possible to compare those costs and benefits which have been quantified in money terms. In doing so, it is helpful to consider the likely significance of those costs and benefits that have not been quantified. Furthermore, the degree of uncertainty surrounding the quantified estimates is important.

The cost analysis has attempted to estimate a wide range of administrative and compliance costs, albeit with some uncertainty and the application of a range of assumptions. There is uncertainty about the environmental impacts, which could be positive or negative. The costs of agreeing a shared definition of partly hydrogenated oils and defining a common test for detecting partly hydrogenated oils (under options 3a and 3b) are undetermined but expected to be small relative to the overall costs (and benefits) of the proposed options.

It could be argued that a greater proportion of the costs of the proposed options are likely to have been captured than the benefits since:

·The health benefits are valued only in terms of savings in healthcare expenditure, and gains in productivity. Other health benefits – particularly in relation to human welfare – have not been estimated;

·The estimated savings in healthcare costs relate only to the reduced incidence of coronary heart disease. Other adverse health effects linked to trans fats are excluded.

Monetisation of these ancillary health benefits would increase the overall scale of the benefits. The understatement of benefits is expected to be much larger than any understatement of costs.

Table 11 summarises the monetised estimates of costs and benefits of the different options. In all cases the value of estimated savings in health-related costs exceeds those of estimated administrative and compliance costs. Options 1b and 3b are estimated to deliver the largest net benefits and Option 2 the smallest net benefits.

Table 11 Comparison between the monetised costs (administrative and compliance costs) and benefits (health-related savings) for the 5 options under variant B2 of the baseline scenario (NPV, EUR)

Based on this evidence, action to limit industrial trans fats in food sold direct to consumers appears to be a very efficient use of resources. Legislation to limit industrial trans fats offers the largest potential net gains, followed by legislation to ban partly hydrogenated oils. A legal limit on industrial trans fats content avoids the need to agree a partly hydrogenated oils definition and to establish the capacity across the EU to test oils for compliance.

The finding that legislation to limit industrial trans fats or ban partly hydrogenated oils are the most efficient of all options is supported by ex-ante analyses in the US and Canada, both of which found large benefit: cost ratios for legal limits on trans fats/ partly hydrogenated oils, details are summarised in Annex 28.

The same result emerges when looking at cost-effectiveness as measured by the cost of the average disability-adjusted life years saved, as shown in Table 12. Option 1b delivers disability-adjusted life years at the lowest cost under all variants of the baseline scenario. The cost-effectiveness of the policies by this measure improves significantly in the transition from variant B1 to B2 to B3 (as the costs are assumed to be fixed but the health benefits increase substantially in B3 as compared to B1). The legal options emerge as a highly cost-effective mechanism for ‘purchasing’ health improvements.

Option 2 imposes significant ‘deadweight costs’ on the food manufacturing sector – it imposes additional labelling costs on food business operators for products that contain no industrial trans fats and where there is therefore no direct benefit. Firms that have already removed industrial trans fats from their products and firms whose products will never contain industrial trans fats by virtue of their composition will still need to change the nutrient declaration.

Voluntary agreements also have the potential for deadweight costs if there is substantial participation by firms that already meet the agreements’ objectives.

Of note, the period over which benefits and costs are assessed is in principle the same, but the costs of implementation are zero after 10 years irrespective of the option implemented. An important difference in the profile of costs and benefits is that costs are incurred only while the options are being implemented, while benefits extend over a longer time period as actions to eliminate industrial trans fats from the food chain now will affect the health of the population long into the future.

Benefits under the different baselines are expected to start to materialise well before 10 years. Generally, benefits are assessed against baseline impacts over a 85 year period. While the changes in intake of industrial trans fats would take place only during the 10 year period, these would have ongoing health impacts which are measured over a longer 85 year period.

Table 12 Cost-effectiveness measure of options by variant of the baseline scenario

Note: ‘>’ indicates that the figures show the lowest expected cost per disability-adjusted life year saved given the greater uncertainty about the efficacy of labelling and voluntary agreements in changing intake.

7.6.Coherence with other EU policy objectives

There were concerns from industry that the legal obligation to label the presence of partially or fully hydrogenated oils in a product might interact negatively with a legal limit on industrial trans fats. Industry stakeholders consider that consumers who have been monitoring the mention of “hydrogenated oil” on labels to avoid industrial trans fats may not understand the difference between “partial” and “fully” hydrogenated oil. Products compliant with the legal limit on industrial trans fats content but containing fully hydrogenated oil could be penalised, according to industry. However, studies on consumer awareness in the EU point to very low levels of consumer knowledge about industrial trans fats and the link to partly or fully or hydrogenated oils. 152 Option 3 b would potentially be a measure in coherence with measures adopted in the US and Canada, facilitating external trade with those regions as similar product requirements are established, in line with EU policy objectives to facilitate external trade,.

7.7.Proportionality

Based on the appraisal summarised above the legal options appear to be the most proportionate solution to the problem of the health consequences of industrial trans fats consumption and the Internal Market effects of uncoordinated approaches to tackling them. The legal options are broad in scope as they in principle concern all food business operators. However, more significant costs are imposed only on those food business operators still using ingredients with industrial trans fats levels above the legal limit and that need to use alternative ingredients to comply. This is in contrast to the labelling option which in many cases will impose costs without generating a corresponding benefit. The scale of the direct health benefits on offer, and the associated reductions in burdens on healthcare services and expenditure are substantial.

7.8.Specific tests: SME test

Based on the screening appraisal, the Competition Test and the Fundamental Rights tests specified by the Better Regulation toolbox do not apply. Specific consideration is needed of the impacts on SMEs, which form a large share of the population of food business operators affected. The ICF study collected evidence to document the perspective from SMEs. This has included direct interviews with a small number of SME representatives (see Table 20 in Annex 4). Due to the challenges of reaching out to SMEs directly, the study team has aimed to clarify the SME perspective by engaging with business organisations that represent a large proportion of SMEs within the sector impacted by the policy options. The majority of members were SMEs for nine of the 16 business organisations who responded to the validation consultation in the ICF study.

The assessment of the impacts on SMEs is summarised below. Eurostat data indicate that SMEs account for:

·99% of enterprises and 50% of value added in the food manufacturing sector; and

·99.9% of enterprises and 75% of value added in the food service sector.

The number of SMEs falling within the scope of each option is estimated in Table 1 3. The number is larger for Options 1a, 1b, 3a and 3b, which cover the food service sector, than Option 2, which relates to pre-packaged foods only. In practice, many SMEs will not be affected by Options 1a and 3a as they will choose not to participate in the voluntary agreement.

Table 13 Cost-effectiveness measure of options by baseline variant

The number of SMEs in scope is largest for the combined options, as (like Option 2) they affect all pre-packed food businesses (whether or not their products are likely to contain industrial trans fats), and, like Options 1 and 3, they affect food service as well as manufacturing businesses.

The estimated costs of the options for SMEs are given in Table 14. These costs have been estimated by estimating the share of the overall business cost estimates above that are borne by SMEs. It is assumed that the share of administrative costs borne by SMEs is proportionate to the number of SMEs in the relevant sectors, and that the share of compliance costs is proportionate to the share of output accounted for by SMEs. These costs are then divided by the overall number of SMEs to estimate the average cost per business.

The estimated average cost per business (expressed in present value terms) ranges from €32 for Option 1a to €18,569 for Option 2. This includes both one-off and recurring costs.

Table 14 Present value of expected costs incurred by SMEs

The country research looked specifically for evidence of impacts on SMEs but little was identified beyond reference to:

·The opportunity provided by supply chain innovation for SMEs to achieve compliance through switching to alternative oils or fats from their ingredient suppliers;

·The challenges some producers, including some small firms, had experienced in reformulation due to particular performance requirements of fats or oils in their production.

The average cost per SME for option 1b of 182 Euro seems to be not too excessive, however, as this is an average value individual SMEs may have to bear a larger cost burden. In order to mitigate the cost and therefore the risk for SMEs of being forced out of business, sufficient transition time would need to be considered. During such period, SMEs have to search for alternative ingredients and test them. Empirical evidence from Hungary (the confectionary industry) suggests that 1.5 to 2 years transition periods (rather than the 1 year given in Hungary) would have helped the sector significantly 153 . In the same vein, Slovenia provides for 1 year, also to help small businesses such as bakeries. A transition period of up to 2 years could be considered, which should enable SMEs to factor in reformulation costs and other costs in their planning to accommodate changes when it best suits their situation.

Apart from sufficient transition time, (technical) support from associations and larger food business operators could help SMEs to adapt. Both a numbers of multinationals as well as FoodDrinkEurope have committed to provide technical support to SMEs to eliminate partial hydrogenated oils from foods. Finally, technical (and economic) solutions provided by suppliers are expected to help SMEs to comply with the new regulatory requirements. Examples from Canada and from the Netherlands (both provided in detail in Annex 17.2) show that ingredient suppliers developed formulations to allow bakeries, as well as other producers such as margarine companies, the food service sector, and virtually all food companies to provide products with no trans fats and, in most cases, lower saturated fat. Ingredients to the bakery sector such as bread and pastry mixes were developed more than 10 years ago to replace partly hydrogenated oil with high levels of industrial trans fats content to fully hydrogenated oil with a industrial trans fats content below 2 %. The initiative of suppliers responded to regulatory requirements (including the legislation in Denmark) and customer demands (demands from large customers, supermarkets and producers of bakery products).

The evidence collected by ICF indicates that SMEs are likely to incur significant costs in order to comply with the measures. The views of stakeholders are that most SMEs will address the requirements by switching ingredients, relying on suppliers of oils and fats. This applies notably to food service SMEs: in some countries such as Austria or Denmark alternative oils have been purchased for frying that effectively enable compliance with the 2% limit on industrial trans fats content. However, the evidence collected by ICF also indicates that challenges will be greater in the food manufacturing industry, where SMEs are likely to encounter difficulties when reformulating their products. According to ICF, while business associations, mainly informed by the experience of very large manufacturers, may provide supporting information to SMEs, it is not certain that SMEs will be able to profit from the solutions developed by larger players in order to achieve compliance.

8.Preferred option    

The legal policy options (1b and 3b) perform better than the alternatives in relation to health benefits (measured in disability-adjusted life years), reduction in health inequalities, improvements in the functioning of the Internal Market, efficiency and proportionality. Details are provided in Table50 in Annex 21, Table 11 and section 7.7.

The savings in health-related costs to society are very much greater than the incremental costs for all options except the labelling. The benefit: cost ratio is largest for options 1b and 3b. Details are provided in Table 11. Furthermore, legislation imposing a maximum limit to industrial trans fats content of products sold direct to consumers (option 1b) performs better in terms of efficiency and coherence than a legal ban on partly hydrogenated oils (option 3b) in that:

·Equivalent social benefits are delivered at a lower cost to the industry;

·Its approach is consistent with the measures already adopted by a number of Member States (and actions planned in others);

·Compared to option 3b, option 1b avoids the need to agree a partly hydrogenated oils definition and establish the capacity across the EU to test oils for compliance with it (both for enforcement purposes and for assurance within the supply chain).

A combination of either of the two options 1b and 3b with mandatory labelling of trans fats levels on pre-packed products (option 2) would raise overall costs significantly. Such a combination is unlikely to deliver added social benefits.

The expected benefits of the voluntary options (1a or 3a), while positive, are smaller and much less certain, generating smaller overall costs, and providing much smaller expected benefits than options 1a or 3a. The members of the food business organisations that are likely to participate in EU voluntary agreements have already reformulated their products to reduce industrial trans fats levels or have eliminated industrial trans fats from their products completely. Research collected by ICF suggests that the businesses responsible for much of the residual industrial trans fats in the food chain are unlikely to participate in an EU agreement, either directly or through representative organisations. The voluntary options do not provide the assured protection that is delivered by the legal alternatives.

In summary, legal policy options (1b and 3b) are the preferred options. Legal action at EU level to reduce industrial trans fats in food would generate positive impacts on health that are substantial as compared to the costs. These measures would substantially remove industrial trans fats-related health inequalities, provide assured protection to consumers across the EU, and support the integrity of the Internal Market. They would also help to ensure a consistent standard of food quality across the EU. The results are robust across all foreseen variants of the baseline scenario. The options that perform best in the appraisal are a legal limit of 2% on industrial trans fats content on food products sold directly to consumers and a legal ban on partly hydrogenated oils. A legal limit of 2% on industrial trans fats content performs marginally better than a legal ban on partly hydrogenated oils in terms of efficiency and of coherence with existing Member State legislation. Therefore, selecting between 1b and 3b, option 1b is the preferred option.

The preferred option is 1b rather than 3b, even though both achieve the same health benefits for the following reasons:

Efficiency and coherence

Option 1b) performs better in terms of efficiency and coherence with existing Member State laws on industrial trans fats than a legal ban on partly hydrogenated oils (option 3b) in that equivalent social benefits are delivered at a lower cost to the industry; Its approach is consistent with the measures already adopted by a number of Member States (and actions planned in others); Compared to option 3b, option 1b is not linked to enforcement challenges: for option 3b a definition of partly hydrogenated oils and a test would need to be established. However, available tools, such as the IV value would pose the following challenges: (i) there is no health or consumer benefit rationale as a basis for the use of iodine value as an enforcement tool; (ii) an iodine value cannot be directly related to TFA content; (iii) there are technological problems to assess the iodine value on a composite food product, which also is not a robust indictor for the presence of industrial trans fats.

Acceptance

Particularly industry in the EU has clearly expressed preference for option 1b as compared to option 3b. Also consumer organisations and health NGOs have expressed agreement with and support for option 1b. The European Parliament and the Council have called for legal limits; particularly the 7 Member States that have already implemented legal limits are in favour of such an EU wide measure, option 3b would need to be introduced as a new legal measure in all Member States, option 1b only in 21 Member States. The same would apply to food business operators where with option 3b they would need to adjust throughout the EU, while for option 1 b only FBOs not active in the 7 Member States with existing legal measures would need to adapt, FBO active in 4 of the 7 Member States would need to slightly adapt to the harmonised legal limit of 2 % on fat basis. Choosing option 3b is expected to meet some opposition, particularly from industry side, but potentially also from the MS that already have a legal limit in place.

In relation to option 1b following the model applied in Member States already, derogations for low fat products could be considered. However, the health impact of such derogations needs to be taken into account. Four of the 7 Member States with national legal limits apply different limits for lower fat products. In view of an EU level legal limit, a legal limit of 2 % on fat basis is in line with EFSA and WHO recommendation, seems achievable in practice and is generally accepted by both consumer organisations as well as health NGOs on the one hand, and industry on the other hand.

Small and micro enterprises constitute the majority of food business operators in the EU. Furthermore, particularly those enterprises are assumed to be contributing to a high degree to still high intakes of industrial trans fats as they have not yet followed past reformulation trends. Further derogations for those SMEs, including for micro enterprises, would jeopardise the effectiveness of the measure.

In order to address the cost burden of the legal measure for SMEs, sufficient transition periods could be granted to ease the burden on them and reduce the risk that due to the measure, small and micro enterprises would be forced out of the market. Empirical evidence from Hungary (for the confectionary industry – a sector estimated to face major technical challenges) suggests that 1.5 to 2 years transition periods (rather than the 1 year given in Hungary) would have helped the sector significantly 154 . In the same vein, Slovenia provides for 1 year, also to help small businesses such as bakeries. A transition period of up to 2 years could be considered, which should enable SMEs to factor in reformulation costs and other costs in their planning to accommodate changes when it best suits their situation.

9.How will actual impacts be monitored and evaluated? 

Present knowledge on trans fats intakes in most EU countries is not robust because it is often obtained from pragmatic dietary assessment surveys that do not rely on nutrient composition databases with complete trans fats data. Therefore, current levels in foods in the EU as well as intake levels cannot be determined with a very high degree of confidence, uncertainties remain. Collecting comprehensive data about industrial trans fats levels in foods before and after the measure enters into force is estimated to be costly. Dietary intake is measured in Member States not very frequently and collection methods may have to be considerably adjusted and refined in order to capture differences in industrial trans fats intake, which would be also linked with considerable costs.

Using assessment methods and instruments already in place could generate valuable indications and estimates about development of industrial trans fats levels in foods after the measure is implemented, and could be considered an alternative, more cost-efficient way to measure success of the initiative.

A number of instruments are available at EU level to monitoring health impacts. However, to assess whether those health impacts are linked with the proposed initiative could only be determined in a dedicated research project. Methods are already in place to collect health data in the EU with regard to cardio vascular diseases in the years following the implementation of an EU level policy measure Such data are regularly collected, such as for a two-year initiative undertaken by the European Commission in 'The State of Health in the EU initiative'.

The evolution of levels of industrial trans fats in foods will be assessed regularly by Member States checking compliance. DG Health and Food Safety audits and related non-audit activities ensures that EU legislation on food safety is properly implemented and enforced and could integrate the issue of trans fats levels in foods in the multi-annual programme. Costs of analysis would be borne by Member States, costs for the auditing by the Commission.

With regard to enforcement issues, in 2016, the JRC of the Commission provided support in developing a reliable methodology to determine levels of industrial trans fats. The JRC delivered their final report 'Analytical approach for checking the compliance of fats and oils' that describes a way of measurement of trans fats and estimating the respective content of industrial trans fats by a proposed calculation method.

ANNEX 1: Procedural information

1.Lead dg, decide planning

Lead DG: European Commission Directorate-General Health and Food Safety, DG SANTE

The Inception Impact Assessment on an Initiative to limit industrial trans fats intakes in the EU 155 was published on 11 October 2016 and the corresponding consultation strategy 156 on 23 June 2017. The inception impact assessment set out the context, scope and aim of the exercise.

The Inter-service Steering Group (ISG) on the initiative to limit industrial trans fats intakes in the EU that had been set up held its first meeting on 8 November 2016 and supported DG SANTE for this Impact Assessment. In addition to the Secretariat General and Legal Service, 6 Directorates-General were invited and designated their representatives to the ISG: MARE, AGRI, RTD, GROW, JRC-ISPRA, ENV and TRADE. The ISG was consulted on the consultation strategy, draft documents and questionnaires of the Study to support the Impact Assessment on the initiative to limit industrial trans fats intakes in the EU, the draft questionnaire of the public consultation and the drafts of this IA report. The ISG met six times to discuss preparatory documents and the draft IA report.

Political validation by Commissioner Andriukaitis, Vice President Katainen and first Vice President Timmermans was received for the Agenda Planning Fiche (2016/SANTE/143) on 6 September 2016.

2.Organisation and timing 

3.Consultation of the RSB

The meeting of the Regulatory Scrutiny Board (RSB) took place on 13 June 2018. The Regulatory Scrutiny Board gave its positive opinion on 18 June 2018 together with a recommendation to further improve the IA report with respect to some aspects, which are reported below. All the Board's recommendations were taken into account by adding explanations in the IA report, except for the preference of the RSB to select the most likely scenario as baseline and to include the others in the impacts section as sensitivity analysis; since there is evidence underpinning each of the baseline scenarios, a decision on the most likely scenario could face the risk of being seen as arbitrary. By taking into account the RSB's recommendation to justify the use of alternative baseline scenarios and including clear explanations with respect to the rationale for the alternative scenarios (which are included to take into account uncertainty about future developments) the various baseline scenarios are presented in clearer terms.

Indications on how the RSB's comments, including the paragraphs/pages which have been added/modified to address the RSB comments, can be found under each element of the recommendations below.

Considerations and recommendations for improvement by the Regulatory Scrutiny Board:

(1) The report should further justify the use of alternative baseline scenarios.

This recommendation was taken into account by adding explanatory text and justification on page 28 paragraph 2 and 3 and pages 29 to 31.

The report should elaborate on the reasons for questioning the validity of the JRC projections and provide additional evidence of the levelling-off of the downward trend in TFA intake across Europe.

This recommendation was taken into account by adding explanatory text and additional evidence on pages 15/16 in the 2nd bullet point, page 17 paragraph 1, page 24 paragraph 3, page 27 paragraph 1 and pages 29 to 31.

It should also further acknowledge uncertainties surrounding this trend, and stress the reasons for increase of TFA intake in some regions.

This recommendation was taken into account by adding explanatory text on page 28 paragraph 2 to 3 and pages 29 to 31.

It could better indicate the distribution and size of population subgroups at risk of excessive TFA intake.

This recommendation was taken into account by information and detailed background data on page 15 1st bullet point, page 44 paragraph 3 and Annex 9.

The report could better explain that existing instruments (such as voluntary industry initiatives) have reached their limits.

This recommendation was taken into account by adding explanatory text on page 19 regarding the risk that imported products that would not be covered by voluntary industry initiatives in a Member State is high, particular in certain Member States, and pages 29 to 31.

The introduction could also better reflect the ongoing global trend in terms of adopting legal measures to limit TFA intakes.

This recommendation was taken into account by adding explanatory text on page 12 paragraph 4 and page 18 1st bullet point.

Building on these elements, the description of the need to act could better reflect potential reputational risks for the EU in case of inaction.

This recommendation was taken into account by adding text about the potential reputational risks on page 18 1st bullet point.

(2) The report should better explain the differences between the option to set an upper limit on TFA content and the option to ban partly hydrogenated oils.

This recommendation was taken into account by adding a figure and explanatory text on page 33, page 35, page 38, page 39 paragraph 1 to 2 and page 70.

The impacts section should more clearly explain why the two options have identical expected health benefits.

This recommendation was taken into account by adding explanatory text on page 38 paragraphs 2and 5, page 39 paragraph 1, highlighting that both options would introduce a comparable cut-off point at 2% trans fats of the fat content, and page 43 paragraph 2.

The report could explain why a more ambitious option in terms of health benefits was not envisaged. The report should also better justify the proposed threshold and explain why it does not consider alternative options such as a limit below 2% of total fat content as feasible. It should more clearly compare them in terms of scope, approach, potential implementation issues and impacts on health as well as for businesses.

These recommendations were taken into account by adding explanatory text and justifications on pages 9/10 paragraph 3, page 35 paragraphs 2, 3 and 5, page 36 paragraphs 1 and 2, page 39 last paragraph.

(3) The report could refine its analysis of the impacts of the proposed measures on the food sector, including the costs for SMEs.

This recommendation was taken into account by adding more evidence on the impact on SMEs from Canada and the hospitality sector on page 51 paragraphs 4 to 5.

It could better describe the market structure of the relevant food sector(s) and describe how the measures might impact different actors in the food value chain. A revised intervention logic could support such an explanation by illustrating the channels through which trans fats enter the food chain and the stages at which different measures propose to intervene.

These recommendations were taken into account by adding information and adding an illustration of the channels through which trans fats may enter the food chain and the stages at which different measures would intervene therefore showing how the measures impact on different actors in the food chain on page 33 and page 45 paragraph 4.

The report could also expand on planned mitigation measures, e.g. in terms of transition periods.

This recommendation was taken into account by expanding on mitigation measures on page 67 last paragraph, page 68 paragraphs 1 to 3, page 70 last paragraph and page 71 paragraph 1.

(4) The report should better explain how future monitoring and evaluation would work.

This recommendation was taken into account by adding explanatory text on future monitoring and evaluation on page 71 paragraphs 2 to 4 and paragraph 6.

4.Evidence, sources and quality

The main source of evidence was the study performed by ICF. Robustness of the results of the study was ensured thanks to sensitivity analysis and comprehensive triangulation of data and evidence collected in the first phase of work via input on a validation questionnaire. Annex 4 provides a detailed explanation about the methodology used, as well as under Annex 4, 6. information about the strength and limitations of the method and under Annex 4, 7. a discussion of information gaps and uncertainties.

ANNEX 2: Stakeholder consultation

1.Introduction

Trans fats 157 are present in foods and increase the risk of coronary heart disease more than any other macronutrient. Industrial trans fats are still present at levels of concern in certain foods and intakes are still excessive in certain cases. This lack of homogeneity in the EU hampers the effective functioning of the Internal Market, negatively affects the protection of consumers' health and contributes to the perpetuation of health inequalities.

In this context, the European Commission is carrying out an Impact Assessment (IA) on a possible EU-based initiative to limit industrial trans fats intakes in the diet of EU consumers. The Inception Impact Assessment (IAA) on the trans fats initiative was published on 11 October 2016 for stakeholders' feedback 158 . It included a preliminary reflection on all the key elements of the IA with a listing of the policy options considered 159 . The Consultation Strategy 160 provided a more detailed outline of the consultation activities planned by the Commission in the context of its trans fats initiative.

2.Stakeholder groups covered by the consultation activities

The Consultation Strategy listed the stakeholders expected to have an interest in the trans fats initiative:

·EU consumers;

·EU food business operators, an effort was made in the consultations to try to obtain specific feedback on mass caterers providing non pre-packed foods ready for consumption and SMEs, taking into account the sometimes more limited resources at their disposal;

·Third-country-based food business operators exporting into the EU;

·Public authorities of EU Member States;

·International organisations and associations, academia and think tanks;

·Public authorities of third countries which already took action on trans fats;

·Individual citizens.

3.Consultation activities already carried out before the launch of the IA

The Commission services had already carried out different consultations on trans fats in preparation of the Commission's report of 3 December 2015 161 , and in relation to the feedback mechanism for the IIA, which constitute an important data source for the IA and the related consultations. Two surveys (one with experts of Member States, Iceland and Norway), and one with stakeholders in the context of the Advisory Group on the Food Chain and Animal and Plant Health, that also was consulted via a written consultation, collected factual information and stakeholders' views on trans fats in foodstuffs and diets in Europe and impacts of strategies to reduce population exposure.

The feedback gathered through these surveys contributed to constitute the evidence-base for the Commission's report on trans fats and helped developing different elements of the IA.

The results of the abovementioned consultations are analysed in detail in the Staff Working Document accompanying the Commission's report 162 .

Nine Stakeholders provided feedback during the feedback mechanism period for the IIA. All of them except for one being an organisation representing national business interests, were EU level organisations and identified themselves as representing company/industry (3 replies) or NGOs (4 replies). Eight contributors expressed preferences for the identified policy options, the majority for legal measures, such as option legal limits on trans fats. Voluntary measures were preferred by 2 business interest contributors, a national contributor pointed to positive experiences. Mandatory labelling of trans fats, was only preferred by one business interest contributor.

Furthermore, NGOs considered that options 1a and 3a as well as the labelling options would not address the problem. Also, NGOs emphasised the urgency of the matter, calling for swift implementations and short transition periods in order to save lives and costs. Replacement fats would need to be considered as well and their impacts on environment and health.

Business interest contributors representing food categories that are sources of ruminant trans fats were satisfied with the focus on industrial trans fats, while the 2 contributors representing business interests of the vegetable oils and fats sector highlighted that scientific evidence was pointing to similar health effects of industrial versus ruminant trans fats and that measures focussing on industrial trans fats only could lead to unfair competition.

4.Outline of the Consultation strategy for the IA on an initiative to limit industrial trans fats intakes in the EU 

The objective of the consultations for the IA on an initiative to limit industrial trans fats intakes in the EU was three-fold:

·to fill in data/information gaps with respect to the baseline scenario and the potential impact of the different policy options retained in the IA;

·to corroborate the findings on the expected economic, social and environmental impacts of the different policy options;

·to give an additional possibility to all stakeholders and individual citizens to provide their views on key elements of the IA.

The planned consultations should also allow the Commission to identify whether anything has been left out in its assessment and to foster transparency and accountability and ensure broadest public validation for the EU initiative.

The following consultation activities were foreseen as part of the Consultation Strategy.

1.Targeted consultations: these were carried out by the contractor ICF which prepared the external study to support the IA and included:

a.Interviews with national competent authorities in the areas of health and food safety and relevant food business representative organisations. The interviews were aimed at collecting primary data to fill in information gaps. The interviews were carried out in the official language of each country selected or in English.

b.Targeted follow-up contacts with sector associations and/or individual businesses to gather additional data. Getting insights into impacts on SMEs was a key objective of these follow-ups.

c.An online survey of a variety of stakeholder groups based on a questionnaire that allowed the contractor to corroborate its findings on the economic, social and environmental impacts of the different policy options. The online survey was distributed to stakeholders at EU and national level in English, responses were accepted in other languages.

2. Open Public consultation (OPC).

A questionnaire translated in all EU official languages was published on the "Your Voice in Europe" website for 12 weeks with the possibility to reply in all EU official languages. The questionnaire built on the progress in the IA process and feedback received.

The methodology used to process the data of the OPC was done via counting from excel tables and clustering of open text field replies in order to qualitatively assess major themes.

5.Results of the Consultation activities for the IA on an initiative to limit industrial trans fats intakes in the EU 

The results of the targeted consultations 163 fed into the IA report on an initiative to limit industrial trans fats intakes in the EU and are in detail reported there. Generally, a number of replies there were in line with replies received from stakeholders during the OPC. Legal limits received support by a number of (also industry) stakeholders, while particular views relating to the labelling of ruminant trans fats or a link with requirements to indicate hydrogenation of oils in the ingredient lists were fully in agreement with the replies received in the OPC.

Regarding the OPC 164 , 118 replies were received, 54 % from individuals 165  and 46 % from stakeholders, experts or participants replying on behalf of an organisation. NGOs represented 20 % or all replies. 15 % of all replies were from business. Of those, 3 SMEs replied, as well as 11 national or EU level business associations that represent a membership with more than 30 % of SMEs. Therefore 12 % of all replies represented SMEs' views. Business was active predominantly in the following sectors: margarines and spreads; dairy products; oils and fats. 7 % or all replies were from public authorities. Two respondents identified themselves as "other", and one international organisations as well as a think tank/research institute participated. A campaign could not be identified in the replies. Not all respondents provided replies for all questions.

With regard to geographic representation, respondents from 23 Member States and 1 respondent from a non-EU country were registered. Over 10 respondents replied from Spain (22) Belgium (19), Germany (15) and the United Kingdom (11).

The first set of questions asked whether the problem description with regard to the trans fats intakes and the trans fats level in foods 166 and the conclusions of the 2015 Commission report on trans fats, summarised again in the IIA 167 were supported. With regard to the trans fats intake, among the respondents to this question, 72 % of consumers, 71 % of industry and 88 % of both public authorities and NGOs respondents agreed, 11 % of consumers and one of the public authorities replying to this question as well as one of the 21 NGOs replying to this question disagreed. 168 Number of responses by stakeholder category is given in Table 15. All but one (individual) respondent that disagreed indicated that actual trans fats intakes and trans fats levels in foods were higher than described in the IIA.

Table 15: Number of responses by stakeholder category in relation to trans fats intake levels and levels in foods as described in Inception Impact Assessment

Asked about their level of concern, most of respondents that answered this question were very concerned or concerned about the impacts of industrial trans fats consumption on the health of the population as a whole (46 % and 27 %, respectively), on the health of particular social groups (61 % and 29 %, respectively) and about current differences in rules and standards regarding industrial trans fats content in the EU market and impacts on consumer protection levels (48 % and 35 %, respectively).

Table 16 Number of respondents rating their level of concern of different issues related to trans fats by stakeholder category

a) consumers    b) industry    c) Member States        d) NGOs    e) others

Most respondents were very concerned or concerned about current differences in rules and standards regarding industrial trans fats content in the EU market and impacts on the functioning of the Internal Market (33 % and 27 %, respectively) and about legal uncertainty on future developments on industrial trans fats and impacts on the functioning of the Internal Market (30 % and 31 %, respectively).

Most respondents totally agreed or agreed that food business operators tend to engage into reformulation only if there is an adequate incentive, which vary depending on the Member State (67 % and 19 %, respectively), that consumers could reduce industrial trans fats intakes by reducing consumption of products that contain them while in the EU, there are different levels of nutritional literacy/consumer awareness of the negative effects of trans fats on health so that not all consumers are actively seeking to avoid trans fats from their diet (63 % and 22 %, respectively).

Table 17 Number of respondents rating their level of agreement with different issues related to trans fats by stakeholder category

a) consumers    b) industry    c) Member States        d) NGOs    e) others

Asked about their agreement with the approach to focus the EU trans fats initiative on industrial trans fats, 86 % of respondents agreed, whereas 8 % disagreed. 169

Concerning subsidiarity, most respondents that answered this question agreed (91 %) with the statement in the IIA that an EU level trans fats initiative was in line with subsidiarity considerations 170 , 9 % disagreed. 171  

Table 18: Number of responses by stakeholder category in relation to the Inception Impact Assessment proposed focus on industrial trans fats and the analysis in relation to subsidiarity

Respondents were asked to rate the different policy options to address industrial trans fats intakes in the EU.

Table 19 Number of respondents rating the different options to the OPC by stakeholder category

a) consumers    b) industry    c) Member States        d) NGOs    e) others

Option 1b was considered by the highest number of respondents as very appropriate, followed by the other legal measure, option 3b and option 2, mandatory labelling. Few respondents considered self-regulation, options 1a and 3a as very appropriate. Likewise, self-regulation, options 1a and 3a were considered as somewhat appropriate.

With regard to the open text replies, the most frequently made comments were that the protection of the health of consumers at EU level should be the central focus for the Commission. Furthermore, it was frequently suggested that the Commission should choose the legal option that best protects health and is already implemented successfully at Member State level. Many respondents urged the Commission to speed up the process and act swiftly.

Also frequently comments called for EU regulation, establishing legal limits as the measure to best protect health, ensuring the effective functioning of the Internal Market and contributing to reducing health inequalities. Legislation should be clear, practical and not include exceptions. National legislation should be avoided. A number of comments highlighted that self-regulation would not be effective.

Concerning mandatory labelling, views were varied. Most frequently, particularly from individuals, the provision of clear information on labels was requested. However, labelling ruminant trans fats was supported by some (predominantly from stakeholders active in the vegetable oils and margarine and spreads sectors). Also, particularly those stakeholders called for the abolition of the requirement to label partly and fully hydrogenated fats. However, a high number of respondents, mostly active in the dairy sector, requested that ruminant trans fats should be excluded from mandatory labelling. A high number of respondents viewed labelling as a not effective.

Further comments, particularly from individuals, called for effective sanctions and enforcement. A number of individuals called for citizen education campaigns. Many comments highlighted the need to protect vulnerable groups. There were calls for further research, consideration of availability and health effects of substitution fats as well as calls to consider SMEs that could face particular problems.

SMEs and associations representing them, active in the dairy or the margarines and spreads sectors voiced the views with regard to mandatory labelling of ruminant trans fats as described above. Furthermore, individual SMEs preferred legal limits, considered national legal measures to be problematic, and that consumer health should be considered by the Commission. Association preferred EU wide regulation or commented that future reductions of trans fats levels were likely due to further national legal measures and voluntary reformulation efforts.

6.Conclusion

A number of consultations have fed into the work on trans fats in the past years in a stepwise approach, results of the different consultations were taken into account for the documents developed during the stages leading to this IA report (Commission report of 2015, IIA, consultation strategy, study by the external contractor). Generally, results of the consultations provided additional information, which were taken into account; views by stakeholders remained rather stable over the years and were considered for drafting the final IA report. The feedback received from the last consultation, the OPC, generally the feedback received confirmed the conclusions of the Commission's report of 3 December 2015, as well as the content of the IIA. Overall, there is a widespread support for introducing a legal limit of trans fats content in the EU, voluntary agreements are less supported and mandatory labelling of trans fats is supported by a number of individuals, while other stakeholders consider labelling to be not effective. No particular, important issues were raised during the OPC that have not been captured during previous consultations and considered in the IA report.

ANNEX 3: Who is affected and how?

1.Practical implications of the initiative

According to the preferred policy option, EU legislation would set a limit industrial trans fats content of 2% of the total fat content of final food products sold to the consumer, following the example of 2% limits to final food products in some Member States' legislation (Denmark (2003), Romania (2017) and Slovenia (2017) .

Alternatively, a more differentiated approach could be chosen, with higher limits (above 2% of total fat) for products with low fat content, and 2% of total fat for food categories with high fat content. Such differentiated limits have been adopted in Austria (2009), Hungary (2013), Latvia (2015) and Lithuania (2017). Austrian/ Hungarian legislation established a maximum content of trans fats at 10% of the total fat content where the total fat content is less than 3% of the product, and at 4% where the total fat content is between 3% and 20% of the product.

A transition period of 2 years is assumed, however, this could be modified during the negotiating and drafting phase of the legal measure.

With regard to food business operators and public administrations, financial and human resources are required to develop and implement the new legislation, develop and implement new products and processes, source alternative ingredients and monitor and enforce implementation.

Administrative costs are incurred by businesses in understanding the rules, determining responses and providing information, and by the public authorities in implementing and enforcing the rules, monitoring and reporting. Compliance costs are incurred by businesses in meeting the legal obligations. These may include the costs of reformulating products and purchasing alternative ingredients. Affected food business operators include those active in the pre-packed and non-prepacked food businesses, and food service companies. Only subsectors whose products are likely to contain industrial trans fats will be affected and businesses in countries with existing legislation are not affected.

With regard to the potentially significant impact on SMEs that is expected to be negative, such impacts are supported by the situation that all SMEs producing foods above the legal limit will be forced to take action and that SMEs may face relatively greater costs and challenges compared to larger firms. However, many SMEs will adopt solutions developed by suppliers, limiting costs. They are likely to be followers of ingredient substitution strategies developed by suppliers or larger firms already. Furthermore, there is a commitment by a large food business association to further encourage and support particularly SMEs, who still face technological difficulties in achieving the elimination of trans fats from partial hydrogenated oils from their products. 172

In this context, FoodDrinkEurope supports the implementation of a recommendation of maximum 2% industrial trans fats of the total fat content of the product sold to the final consumer.

From the business perspective, after the introduction of new legal rules some further activities would be needed, such as the provision of information, new product development, sourcing of alternative ingredients (substitution of ingredients with high industrial trans fats content with polyunsaturated, monounsaturated and saturated fats), implementation of new products and processes. Public administrations would need to provide guidance and advice, while taking care of monitoring and enforcement.

The resulting output consists of a decrease of industrial trans fats content in food below 2% of fat, and the linked output consists of a reduction of industrial trans fats consumption for all population subgroups, ongoing product development and innovation by food business operators, achievement of a level playing field within Internal Market, including imports, and a shift in alignment with practice in export markets. The long-term impacts lead to a decrease in cardio vascular diseases prevalence and mortality, improved productivity in EU economy from healthier consumers, reduced health inequalities amongst consumers, reduced economic burden on healthcare systems, enhanced image, competitiveness and innovation of food industry and increased trade across EU Member States (and third countries).

2.Summary of costs and benefits

ANNEX 4: Analytical methods

1.Study methodology development

This section provides an outline of the methodology developed by ICF for their study that was used as the principal source of data for this IA report.

Firstly, a methodology refinement was performed by ICF, including adjustments to the data collection strategy and impact assessment approach, reflecting discussions held with the ISG. Here, the baseline and policy options specifications as well as associated theories of change were developed. Theories of change make explicit the mechanism by which each intervention is expected to lead to the intended outcomes, and the key assumptions that need to be satisfied for it to do so.

The theory of change provides a narrative description of cause and effect, and the principal assumptions made about behaviour, context, etc. This framework also supports identification and analysis of factors that contribute to uncertainty about benefits (the level of assurance one has that the intervention will achieve its intended results) and costs (the likelihood that the costs will be higher or lower than the central estimate). This includes uncertainty relating to estimation of benefits and costs, and uncertainty about whether the benefits or costs will be realised (e.g. due to lack of compliance).

The analysis of the options through the development of theories of change helped to identify their respective expected impacts. The analytical framework included to outline for the different questions to be answered for the IA judgment criteria, indicators, sources of evidence, and methods of triangulation and validation.

2.Data collection and review

Information and data gaps left after the analysis of available information by ICF were identified and closed. As many data had already been collected previously to the ICF study and some analysis had been undertaken for a number of the impacts to be assessed, targeted efforts by ICF were carried out to complement those data with additional information that would enhance the analysis. It was also focused on closing information gaps in relation to:

·The baseline scenario and basic data required to support option appraisal;

·Studies which could help to inform the analysis of the impact of agreed potential policy options, and especially environmental impacts, for which comparatively few data are available.

Given the tight timetable set for the ICF study, the research was concentrated over a short period of time and was entirely aimed at informing the tools for the impact assessment models. It involved two sub-tasks:

·An in-depth review of existing data; and

·The collection of primary data from stakeholders in countries that have implemented similar measures to tackle trans fats intake via:

oA programme of interviews with competent authorities and food business representative organisations in the target countries;

oFollow-up research with selected sectors in those target countries to gather supplementary information.

ICF also consulted a number of representative organisations at EU level. These additional consultations were conducted to map better at the EU level those elements of the food supply chain that are relevant to the trans fats problem. The results informed extrapolation from existing data on how different policy options may impact the whole EU industry.

Review of existing literature and data

The desk research of ICF focused on sources identified earlier in the project, and was completed with additional literature search in the language of the countries selected for further investigation. Data were collected according to a common framework and a list of keywords defined for use in the search of publications and data. All publications were reviewed in order to extract relevant information, which was then inserted into a common template.

Interviews

The ICF team carried out 24 interviews with competent authorities and food business representatives in EU Member States and third countries. These interviews were carried out following a common approach to fill out gaps identified during the desk research. This included also some interviews with EU-level representative organisations in order to obtain additional inputs on impacts. The full list of interviews is provided in Table 20.

Targeted follow-ups

A number of targeted follow-up actions by ICF followed the interviews and literature review. These solicited a number of email submissions, particularly from industry. A number of additional phone conversations were held with various actors from the industry and researchers with expert knowledge of the topic in the individual countries.

The full list of interviews and targeted follow-ups is provided in Table 20.

Table 20 List of interviews and targeted follow-ups carried out

Step/task 2.1: Review of existing literature and data

Step/task 2.2: Interviews

Synthesis

The evidence collected in the country research by ICF was consolidated into a single document for each country. These country case studies are provided in a separate document (Annex 32). They summarize the data collected from the desk research, interviews and targeted follow-ups. The information collected through interviews with EU level business associations is consolidated in Annex 29.

The evidence was also aggregated in a single MS Excel file document that includes, for each type of impact: a list of indicators; the description of the evidence obtained, either quantitative or qualitative; and sources for that evidence. This information has been replicated in Annex 30.

3.Screening of impacts and assessment of significance

The ICF team carried out a screening of impacts and assessment of their significance, in line with the guidance on impact assessment set out in the EC Better Regulation guidelines. All potentially significant impacts were retained for more detailed analysis, while those which are insignificant were discarded. This screening was based on a thorough analysis of the evidence. The outputs of this task in this report appear in Annex 12.

4.Analysis of impacts

Baseline assessment

This task involved qualitative and quantitative analysis to inform specification of the baseline scenario that describes the production and consumption of trans fats in the EU in a context of no additional EU intervention. The work was informed by the baseline scenario of a study completed by the JRC 173 , and the qualitative evidence collected before by ICF.

Analysis of impacts of each option

The assessment of impacts has been carried out by ICF on the basis of a detailed specification of the policy options, developed by ICF in conjunction with the Commission at the start of the study. The options that are compared to the baseline are defined in the main text of this IA report. The impacts of each option were then assessed by ICF.

The estimation of health costs was based on a model developed by the JRC and published in 2016. 174 A number of the assumptions have been modified. To assess impacts on health inequalities, the team used outputs information emerging from the JRC model to then produce a qualitative assessment of impact on health inequalities, informed by the scientific literature and available data.

The original specification of the JRC model is described here, together with a list of the assumptions that were modified and added for this assignment. These assumptions are explained in more detail below.

The model can be used to estimate the impact of EU-level policies that lead to changes in population industrial trans fats intake. It expresses the results in terms of changes in health treatment costs and overall health benefits (measured in disability-adjusted life years). The model considers only coronary artery disease. Other potential benefits of lowering trans fats intake, such as impacts on insulin sensitivity, obesity, diabetes, cancer, or early growth and development, are excluded because of inconsistent evidence and lack of data. As such the impact assessment can be considered to be conservative with respect to achievable health benefits resulting from (fast) industrial trans fats removal from the food supply.

It is a state-transition model (Markov model) built in Excel. The Markov model is used to simulate how people move in yearly cycles through four health states in each of the policy options. The four health states are as follows:

·Well: the state for each individual with no history of coronary heart disease; a person can remain here until death or move to “coronary heart disease”;

·Coronary heart disease: state for individuals who have coronary heart disease move to this state for a maximum of 1 year; from this state, individuals can move either to “History of coronary heart disease” or “Death” but not back to the “Well” state;

·History of coronary heart disease: state for post–acute coronary heart disease individuals; survivors from a “coronary heart disease” state move to this state until death or until they suffer a new coronary heart disease event, in which case they move to the “coronary heart disease” state;

·Death: any individual can move to this state at any time.

The model is applied to the EU population and accounts for all costs and effects applicable or resulting from the policy options over the course of a lifetime (85 years). The current industrial trans fats intake, defined as percentage of total energy intake, used as starting point for the model (“today”) is calculated as a weighted average of data at Member State level collected through existing evidence and a survey.

The JRC chose the 85 years 'life-time horizon' following relevant methodological guidance, such as NICE (UK) or ISPOR (international). The 85 years are slightly above the average life expectancy in the EU. The NICE guidelines (for assessment of alternative health technologies) notes on the appropriate time horizon: "Long enough to reflect all important differences in costs or outcomes between the technologies being compared". Downs et al. 175 reflect on modelling studies' limitations: 'There are several limitations to the assumptions used in many of the modelling studies included in the review that need to be considered. The time horizons used for the models were short, with the exception of the article by Martin-Saborido et al..174 In one study, the deaths averted were only examined for 1 year. Longer time horizons would be more appropriate, because the implementation of a trans fats policy would not result in instantaneous effects on cardio vascular disease, the use of a lifetime approach as was used in the model examining the impacts of trans fats policies in the European Union would likely be more appropriate.'

The model calculates, for each option, coronary heart disease events and mortality in yearly cycles over a period of 85 years. The relative risks for coronary heart disease associated with the different industrial trans fats intakes are based on the calculations in Mozaffarian et al (in which the “pooled multivariable-adjusted relative risks for 2% of total energy intake of trans fats, as an isocaloric replacement for carbohydrate, was 1.23 (95% CI = 1.11–1.37).” This is then applied to the different industrial trans fats intakes to calculate the probability of a coronary heart disease event.

Costs (of policy implementation and healthcare related) and outcomes (expressed in disability-adjusted life years, which measures overall disease burden) are estimated as the population circulates through the model. These are calculated for each policy option and then compared with the baseline. The model applied some simple assumptions to assess the broad scale of costs of public sector interventions, but excluded costs for business. Because of the limited scope and detail of the cost assessment, the model’s capacity to estimate costs of policy implementation was not used in this appraisal and as such this aspect is not discussed further.

Concepts of industrial trans fats -related diseases used in this report, coronary artery disease, coronary heart disease and cardio vascular disease are explained in Annex 6.

For the starting point of the model (“today”) the risk of coronary heart disease is calculated on the basis of hospital discharges and already includes the risks from current industrial trans fats intakes, which are specific according to country, age, and gender. The reduction in coronary heart disease risk linked to industrial trans fats reductions in the following years from “today” is then calculated by using the relative risks above. Subsequently, the resulting disability-adjusted life year are then calculated on the basis of the modelled number of coronary heart disease events and deaths.

Given the uncertainty related to trans fats intake data, the JRC model tests three scenarios for intake in addition to the baseline.

Table 21 industrial trans fats intakes across the baseline and alternative scenarios as considered in the JRC model

The reference case built into the model assumes the highest population trans fats intake over the modelled horizon. JRC assumed that in the absence of EU action industrial trans fats consumption decreases over time and would reach zero in 10 years' time.

The JRC used the model to test scenarios based on a voluntary agreement, mandatory labelling and a legal limit on industrial trans fats content. The details of these scenarios are provided below for comparison to the scenarios tested for the current study (which are explained in the main text of the IA report):

·JRC - Voluntary agreement: This option assumes the creation of a voluntary agreement between the food industry and policy makers across the EU. The model assumes a decrease in industrial trans fats intake which would reach zero in 5 years' time. Costs of the option are related to food inspections to monitor and evaluate the agreement as well as the healthcare costs;

·JRC - Mandatory labelling: This option assumes that the current European legislation on the nutrition declaration on foods (Regulation EU (No) 1169/2011) would be changed to include also the declaration of trans fats content. The measure would apply only to pre-packaged food. The resulting decrease in industrial trans fats intake is slower than in the voluntary agreement case because it would lead to reformulation only in pre-packaged foods. industrial trans fats intake related to pre-packaged food (it is assumed to be 50% of the total population intake) decreases to zero in 3 years' time. Costs of the option are related to information campaigns to increase consumers' understanding of harmful effects of trans fats, as well as the healthcare costs;

·JRC - Legal measure: This option assumes the introduction of legislation at European level that limits the content of industrial trans fats in the food supply. The model assumes that the industrial trans fats intake is completely eliminated after 2 years. Costs of the option are related to food inspections to enforce the legislation as well as the healthcare costs.

For this assignment the JRC model was adapted in the following ways:

·The baseline scenario was developed further to accommodate known uncertainty about the future trend in industrial trans fats intake in the absence of EU action. Three variants of the baseline were specified to represent the spectrum of expected possible trajectories – industrial trans fats intake remaining constant at current levels, a linear decline in industrial trans fats levels to zero over 15 years and an accelerated linear decline to zero over 10 years;

·More conservative assumptions were defined for the impacts of voluntary agreements;

·The assumed impact of a legal limit on industrial trans fats content on industrial trans fats intake was revised from zero in the JRC model to 0.009 % of total energy intake, which corresponds to the average intake in Denmark as of 2014;