Here we will see the level of involvement of stakeholders and/or the public in the drafting and monitoring of HC Regulation. We will also understand if its practical implementation can be considered transparent.

In a preparatory step, the EC gave a number of invited stakeholders the possibility to comment on the Nutritional Claims and Functional Claims Discussion Paper (Directorate General Health and Consumer Protection, 2001), drawn up in preparation for the future Health Claim Regulation. According to official data, among a total of more than 90 invited stakeholders, there were nine consumer organizations, as well as a handful of scientific institutions, European agencies and NGOs (like advertisement associations). The remainder of invited stakeholders came from the food industry. Practically all of the discussions centered on issues particularly relevant to industry, rather than consumers. As we will see, the ultimate design of EU health claim regulation, as subsequently adopted by the European institutions, for the most part does not reflect the concerns voiced by the stakeholders in this deliberative process (including those from the food industry).⁸

There was no further stakeholder or public participation nor deliberation with respect to the Regulation, until the publication of the final Common Opinion of the European Parliament and the Council (European Parliament and Council, 2005) that formed the basis of the NHCR. One of the fundamental decisions included in this Opinion was to task the European food authority, EFSA, with the scientific assessments. Following the Regulation’s adoption, the EC tasked EFSA and its NDA Panel with preparing the entire regulatory process foreseen in the NHCR (European Parliament and Council, 2008).

More recently, EFSA organized a number of seminars with stakeholders from industry in order to keep collecting opinions (Valtueña & Siani, 2017). Another potential avenue for external scrutiny is the process for an ex-post assessment of the NHCR that the EC offered to a number of stakeholders in 2015. This process was very narrow, however, focusing exclusively on health claims made on plants.⁹

Therefore, we can conclude that since the enactment of the Regulation the expert system that underpins it has not been opened up to greater public scrutiny nor debate. Nor have there been any relevant changes to its regulatory objectives, or the operationalization of the assessments and authorizations of health claims. And this is still the case, even after the enactment of the ‘Interinstitutional Agreement on Better Law-Making’, which explicitly considers that, not only ex-post evaluation of existing legislation, but also public and stakeholder consultation will help achieve its objectives (European Parliament, Council of the European Union, and European Commission, 2016).

As for transparency: a lot of relevant information has been available on various EC and EFSA websites and in periodic publications, particularly expert opinions, discussion papers, panel reports related to the establishment of the NHCR (including related to the monitoring of the NHCR’s implementation), as well as all relevant guidelines for industry and stakeholders elaborated so far. All of EFSA’s decisions on claim authorization have been publicly available in the EFSA Journal, giving the applicants from the food industry the possibility for appeal in case of rejection of the authorization. Furthermore, with the enactment of Regulation (EU) 2019/1381 on the transparency and sustainability of the EU risk assessment in the food chain (European Parliament and Council, 2019), applicable as of the 27^th of March of 2021, the new ‘OpenEFSA portal’ also allows for public access¹⁰ to the documents and information that are part of the life cycle of any EFSA assessment process for all HC applications (at least as long as there are no property rights conflicts). In other words, any public consultations and comments received,¹¹ as well as any official responses, are publicly available at least until the authorization procedure has come to a close.

In one sense we can thus consider the regulatory process as transparent, even more so since 2021, due to the implementation of EFSA’s new ‘transparency and stakeholders engagement’ approach.¹² It is, however, not transparent in the broader sense as understood by the ‘new governance’ approach, as explained above.¹³ That is because neither the EC nor EFSA have ever given attention in their published documentation to possible implications, impacts or alternative choices, as far as the objectives of the regulation and the operationalization of the assessments are concerned.

In sum, ‘public scrutiny’ and ‘transparency’ (according to the procedures foreseen in the ‘new governance’ approach) have been either absent or only partially implemented in European HC regulation. In fact, public scrutiny and deliberation would appear very difficult to conduct in the current framework of the NHCR. This point applies particularly to any possible deliberations as to the regulation’s objectives and their contribution to social preferences or needs, including questions like, for instance, what kind of knowledge ought to underlie the regulatory process or what role the expert committees should have in the decision processes. However, it is issues like these which underlie, as we have seen, the ideas of a ‘new governance’.

This partial lack of transparency, as well as absence of social debate with respect to regulatory objectives and the role of expert assessments in Europe, is very important to take into account, given that in health claim regulation in other parts of the world there exist well-established alternative approaches to scientific assessments and stakeholder involvement that differ from the EC’s policy choices and that result in social impacts that differ from the European ones (see in Discussion).

Here we will analyze the 3. and 4. objectives: less hierarchical structures and more participation, particularly the type of participation and stakeholder dialogue that is possible under the NHCR.

From its inception, the NHCR has foreseen the possibility for comments from applicants, stakeholders or the general public on EFSA’s scientific opinions in the case of certain types of health claims (those subject to an individual authorization procedure).¹⁴

In spite of this possibility for non-binding comments, as well as the recent creation of the ‘OpenEFSA portal’ as EFSA’s new way of public information with respect to each particular assessment process, there is no institutionalized mechanism for facilitating any kind of particular participation (or in any way less top-down interaction) of operators from the food industry (beyond their submitting proposals for health claims and appealing rejected applications). Nor do their exist any deliberative mechanisms that would facilitate an actual dialogue between experts, regulatory authorities, as well as non-experts with respect to non-scientific or technical issues of the Regulation. In practice, the NHCR basically assigns a passive role to consumers and the food industry: they are supposed to accept regulatory decisions that are ‘good for them’, meaning being protected from uncertain or false information (false health claims).

In any case, the EC received several uninvited letters related to the NHCR from stakeholders and the general public after the first and third series of EFSA opinions on Article 13 claims. The Directorate-General for Health & Consumers responded with ‘Collective Answers’, which have also been made public.¹⁵ In response to stakeholders’ concerns about the effects of the Regulation on the sector of food with health claims, these Answers have focused on the need for maintaining the current strict scientific requirements in order to achieve the Regulation’s objectives. EFSA experts, in fact, have a critical and predominant role in the entire regulatory process. The European Commission itself made explicit that it defers to these experts: the scientific assessments are in the hands of EFSA’s expert committees whose conclusions are not to be questioned by the Commission, nor any other authority, like for instance EU member states:

In sum, not only the assessment procedure, but –in fact– the entire regulation can be considered a top-down policy instrument that does not allow for binding participation. There is no possibility for stakeholder intervention in the mechanism for assessing health claims. Rather, the entire regulatory approach remains highly expert-based. Given the way the regulation has been designed and operationalized, it becomes clear that consumers, market operators and food innovators are understood as mere passive receptors of the benefits provided by scientific knowledge that is considered ‘certain beyond any reasonable doubt’, given that it previously has been assessed by the NDA Panel.

As stated in the NHCR, food market operators and, even more, consumers are the targeted beneficiaries of this Regulation. However, as far as those stakeholders are concerned, the NHCR makes a number of implicit, underlying assumptions. On the one hand, the entire European HC regulatory process presupposes a certain type of consumer of foods with claims, as –homogeneously and narrowly– having specific knowledge and attitudes. It has to be taken into account that, from the standpoint of EU regulation, the ‘consumer’ in this context is a mere theoretical entity, namely, the ‘average consumer’, as established by the Council of the European Communities (Council of the European Communities, 1984/2005). The NHCR incorporates this definition of a consumer: one ‘who is reasonably well-informed and reasonably observant and circumspect, taking into account social, cultural and linguistic factors’ (European Parliament and Council, 2006). This concept carries with it the implicit (and questionable) supposition that consumers in the real world will make the ‘rational choice’ of improving their health (by consuming foods with health claims) if presented with accurate, relevant and ‘certain’ information on the food labels (Sanz Merino, 2020).

On the other hand, EU regulation also presupposes a consumer with particular needs and preferences: EFSA and its NDA Panel pre-define this consumer as (almost exclusively) worried about being misled by false or erroneous health claims (Todt & Luján, 2021). A similar point can be made with respect to industry: the NHCR and EFSA interpret that the food sector’s most important need, in the development of a market for health claims and related innovation, is protecting food companies from fraudulent competitors.

On the basis of such suppositions about stakeholders’ preferences and needs, EFSA has since worked on the assumption that the NHCR’s objectives can best be achieved by providing scientific certainty, in the sense of requiring the highest possible scientific standards (namely, clinical trials) in data generation. Thus, as we have seen above, the entire regulatory process is geared towards restricting authorization of claims to those that can be considered certain beyond any reasonable doubt.

Two further circumstances contribute to this very strict regulatory approach and give it its internal coherence: firstly, the EC’s long-standing separation in food governance of risk assessment from risk management (part of the EC’s response to the late 20^th century EU food crises and the concomitant loss of public trust in institutions and experts (Levidow, Carr & Wield, 2005)); a separation that the EC is appealing to in health claim regulation, too, as we have already seen¹⁶. Secondly, the trust that the EC places in its experts, and in EFSA’s particular approach to HC regulation: providing scientific certainty to consumers by way of assigning most relevance to data from RCTs in the hierarchy of evidence (EFSA, 2017; 2021b). However, bestowing such relevance onto RCTs in this context is in itself a methodological decision that is based on particular epistemic suppositions: namely, that in nutrition –as a general rule– it is feasible to generate useful RCT data on causal relationships between intake and outcome, and that such data (and the certainty that come with them) are indispensable for regulating health claims (Todt & Luján, 2017).

In sum, EFSA’s decisions on scientific standards and preferred methodologies naturally flow from EFSA’s interpretation of the principal aim of HC regulation as protecting food markets and consumers from erroneous health claims (Todt & Luján, 2021). However, while EFSA presents its approach based on the ‘best science’ as obvious and self-evident, this interpretation is not unanimously shared within the relevant scientific community. In fact, there is a controversy in the nutrition sciences about which scientific methods are most appropriate for assessing the health and nutritional benefits of food.¹⁷ The European regulatory process does, however, not show any sign of recognition of the existence of this controversy, as well as its possible implications for choices of regulatory aims and scientific methods.

We can conclude that not only are the new governance’s objectives of public scrutiny and wider participation not reflected in the NHCR. Even more so, neither do the EC nor EFSA consider them necessary or advantageous in HC regulation, at least as long as the scientific assessments make use of the has been stablished as the highest possible scientific standards. The predominant idea is that it is the experts who guarantee the application of the best science; thus, questioning of the experts does not carry any regulatory or social advantage. This stance is, of course, coherent with the current lack of non-expert participation or stakeholder dialogue during the assessments: any non-expert participation is useless if scientific certainty is the ‘best’, and only, way of protecting the market and consumers from false claims (the objective of highest social concern). This of course does nothing but reinforce the consideration of consumers and applicants as mere passive receptors of absolutely reliable information on health outcomes. In other words, the NHCR’s successful practical application, as measured by its own standards, is not dependent upon any of the features of the participatory turn; even more, incorporation of any such features would be considered counterproductive.

The new EU transparency regulation does not change this situation: as already stated, this regulation does nothing to further the objectives of the new governance approach, namely, being transparent in terms of making public any alternative views on science and technology regulations, possible alternative outcomes, the varying uses of expert knowledge in regulation or the mentioned nutrition scientists’ and industries’ controversial opinions.

In the following, we will see that, in other jurisdictions, authorities have found approaches to HC regulation that differ in fundamental ways from the European approach. Those alternative regulatory approaches are based, of course, on alternative assumptions. Assumptions that, under the ‘new governance’, should be the subject of public and stakeholder debate.

The previous discussion has shown that in EU regulatory practice consumer interests other than not being misled, like informed choice or usefulness of health claims for improving consumer health (explicitly mentioned in the NHCR), are subordinated to the primary aim of minimizing false positives (as are other NHCR objectives, like fostering innovation in the European food market). Giving more relevance to these alternative objectives would have a direct impact on how to treat false positives and false negatives, which particular scientific methodologies to require for generating decision-relevant data, or how to rank scientific methods according to their relevance for data generation (see footnote 17). One important point for our discussion is that the underlying assumptions are none other than implicit or explicit choices made by the regulatory and assessment authorities. Some of such choices, in principle, are open to public and stakeholder debate and engagement, because they are not based on scientific facts, or not alone, but rather grounded in social, political and epistemic preferences. Even more so because these possible choices have direct effects on relevant regulatory outcomes. For instance, making obtainment of regulatory approval more difficult, as EU Regulation does, result in fewer approved claims reaching the market (and will tend to depress industry funding for health claim-related food innovation) (Khedkar, Bröring & Ciliberti, 2017; Sanz Merino, 2022).

In sum, it is uncertain or false information (false health claims) that the European regulatory process is designed to weed out, on the basis of the regulators’ unquestionable preference for the ‘best possible science’ (establishing causal relationships between intake and outcome), as well as expert assessment. Therefore, as a result of the epistemic and social assumptions and decisions made in EU regulation, European consumers are being protected against fraud and error, but they also have both less choice of food products with approved health claims, as well as fewer options for participation in food governance. In practice, EU consumers of foods with health claims have one type of (binary) choice only: to buy or not to buy. They do not have, for instance, the possibility of assessing for themselves what ‘level of certainty’ (or lack thereof) they are comfortable with in order to proceed with the decision to purchase a food identified by a health claim.

An alternative regulatory approach with less stringent evidence requirements would lead to more health claims obtaining authorization, thus increasing consumer information and choice. In order to better understand such alternative regulatory outcomes, we will contrast EFSA’s current regulatory choices with the choices made by HC regulators in other relevant jurisdictions. We will consider here the cases of the United States and of Japan.

In the U.S., for instance, health claims can obtain authorization on the basis of varying evidence requirements (FDA, 2003; 2009/2018). One important difference, as compared to the EU, is that in the U.S. claims may be authorized as ‘Qualified Health Claims’ on the basis of not (yet) fully confirmed but still credible scientific information (Domínguez-Díaz, Fernández-Ruiz & Cámara, 2020; Sanz Merino, 2022). Foods with qualified claims come with one of four possible levels of scientific backing, identified on the food item’s label (FDA, 2009/2018). Overall, U.S. health claim regulation is more tolerant towards false positives, which is the inevitable trade-off for aiming at maximizing the number of approved, reasonably substantiated claims. U.S. consumers, as compared to EU consumers, have more choice but less certainty as to the quality (reliability) of each individual claim. In part, consumers assess for themselves the level of available evidence, in order to decide what ‘level of certainty’ (or lack thereof) they are comfortable with in order to proceed with the decision to purchase a food identified by a health claim. Therefore, consumers assume the responsibility for judging if the advertised level of scientific substantiation is good enough for them in order to trust the claim, or not.

Another important example of a different approach to health claim regulation is provided by the Japanese approach (Iwatani & Yamamoto, 2019) which currently is mostly based on industry self-regulation (Kamioka et al., 2019). Apart from foods with fully-fledged Health Claims, as well as Qualified Health Claims, with assessment criteria similar to the U.S., there is a category called ‘Foods with Function Claims’. This type of health claim (that only exists in Japan) is subject to a notification system only. Companies submit information on food safety, efficacy and possible risks to the regulatory agency prior to proceeding with commercialization. All those steps are ‘under food business operators’ own responsibility’ (CAA, 2015). The system has led to a certain liberalization of the regulatory environment, resulting in an important increase in health claims (of this latter kind) on the market (Iwatani & Yamamoto, 2019).

Therefore, as these examples of alternative approaches to health claim regulation show, it is the regulatory authorities’ choices in interpreting the requirements set by regulation that drives the overall regulatory approach, as well as the final (non-epistemic) outcomes (Sanz Merino, 2021, 2022). In other words, health claims can be (and are) regulated (and even assessed) in a wide variety of manners and the decision as to which of the alternatives to prefer is a fundamental regulatory choice, non-scientific one. (See Table 1)

Considering the participatory turn’s objectives, it would be reasonable to expect that regulators, vis-a-vis consumers, stakeholders and the public, could make of this regulation a more socially robust one in such senses as much as possible. It seemed viable in this case, since the unveiled regulatory assumptions and options are typical subjects of public scrutiny and decision-making according to the (still valid) EU’s ‘new governance’ approach.