The European Commission has initiated a comprehensive review of the European standardisation system. This article claims that the Commission’s initial analysis is incomplete with regard to the difficulty of salvaging a formal legal role for European standards. It describes how the standards nonetheless can keep an important role, though through the back door.Summary:

The following article analyses the legal predicament in which the European Court of Justice has placed the EU legislator and the Commission by considering referenced standards to be part of the EU legal order. It shows that the current system of referencing standards is not tenable. International and European standards can hardly fulfil requirements derived from the rule of law principle. In addition to the known alternatives of European technical specifications, the article points out another one that has already been chosen in some EU member states: the use of dynamic, undefined legal terms such as ‘state of the art’, which leaves it up to the legal practitioner to find applicable standards and to use them as a concretisation of the dynamic, undefined legal term. As a result, the author recommends, as default practice, using such undefined legal terms and only implicitly referring to standards through the back door. In addition, recommendations under Article 292 TFEU or corrective technical specifications can be adopted by the Commission to steer the use of standards. Self-standing technical specifications should be adopted from the outset if it is clear that the undefined legal terms are causing the practice of economic operators to drift too far apart.

I. The European Commission has started a comprehensive review of the European standardisation system. Before revising the European standardisation system, the current situation should be adequately analysed. The initial analysis provided by the Commission in the document ‘Call for evidence for an impact assessment’ is good in principle. However, it does not provide the necessary clarity with regard to the predicament in which the European Court of Justice has placed the European legislator and the Commission with its James Elliott ruling of 2016¹. By stating that European standards are part of the European legal order, the rule of law requirements also extend to European standards. However, both the European and, above all, the international standardisation system is not designed to meet the requirements of the rule of law principle. The rule of law principle includes, in particular, the requirement of legal certainty. It must be clear to those subject to the law which requirements they must fulfil.

Since European and international standards are based on a network of so-called dynamic normative references, it is not clear at the time a standard is cited in EU legal acts which text is to be applied by the parties subject to the law, since the referenced standards contain themselves dynamic references to the latest version of a standard with a specific number, and this standard is subject to revisions and amendments. An economic operator endeavouring to comply with standards must therefore pursue a vague ‘moving target’. In addition, multiple transition periods create uncertainty. Overall, such a system does not fulfil the requirement of certainty as part of the principle of the rule of law. The rule of law in the form of the principle of certainty requires that legal texts only refer to static texts or at least very controlled text structures that may not be further amended without the consent of the legislator or at least the Commission.

II. The only way to salvage the European and international standardisation system as source of EU law would be to prescribe the application of a standard in a specific version in such a way that the normative references in the referenced standard, which are intended to be dynamic in the ISO/IEC or CEN standardisation system, are not applied dynamically but statically, i.e. the chains of references to the versions of the referenced standards that were valid at the time of the formal adoption of the EU legal act are frozen. However, this is unrealistic, as the version of the standards referred to at the time of reference by the EU legislator or the Commission can usually no longer be easily consulted after a few years and practitioners always apply the latest versions of the standards anyway. The units of the EU Commission that wanted to enforce a static interpretation of the chain of references are therefore likely to have largely failed in practice. A further disadvantage of such a static interpretation of the chain of references would be that contradictions would arise in international trade with the standards as they apply to other countries, as in most other countries the international standards are applied with their dynamic references. One of the major advantages of legislation based on standardisation would therefore be reduced by a static reference chain interpretation in the EU.

III. Even if it were possible to reinterpret the dynamically conceived reference chains in the standardisation world statically, a further problem would arise: the net-like network of normative references inevitably results in contradictions upon closer analysis, unless all norms in the reference chain are updated simultaneously, which would not be feasible. Here is a fictitious but realistic example: Standard A from 2025 refers (normatively) to standard B from 2023 and elsewhere to standard C from 2019. Standard B from 2023 in turn refers to standard D from 2021, while standard C from 2019 refers to standard D from 2015. The person subject to the law should therefore apply standard D of 2021 via one reference chain and standard D of 2015 via the other reference chain. If the two versions differ, which is to be assumed, this results in a contradiction and thus legal uncertainty. Such contradictory reference chains are found above all in standards that belong to a particular family of standards, but not only there; even very fundamental standards such as those on quality assurance or risk assessment are referenced by many dozens of other standards via parallel chains of references that refer to four different versions. Contradictions arise not only through new versions (recognisable by a different year), but also through supplementary amendments. This and further deficiencies are described in more detail in this article of mine.

IV. The European standardisation system operates under the constant shadow of potential scrutiny by the European Court of Justice (ECJ). One day, the Court may conclude that, given the status of European standards as part of EU law—particularly in light of Regulation 1/58—standards that have not been translated into all official EU languages cannot have binding legal effect. While the ECJ has held in past rulings that not all documents must be translated into every EU language, it has consistently affirmed that legal acts imposing obligations are only valid when available in the language understood by those to whom they apply or in the respective official language². Together with Regulation 1/58, this jurisprudence, though developed primarily for judicial and administrative proceedings, presents a significant systemic risk for the European standardisation framework: the vast majority of European standards have never been translated into all official EU languages.

V. The predicament described in Section IV. cannot be solved at all within the current legal system. The predicament described in Sections I. to III. can only be resolved to a very limited extent within the current legal system. Instead of generally referring to standards or citing them in the Official Journal, the European Commission would have to filter out the few standards that are exceptionally not embedded in the system of dynamic normative reference chains or reference networks and would only be allowed to refer to these. This filtering process leaves only a very small number of standards. Of the few remaining standards, others that contain other legal defects are likely to be eliminated, such as definitions that deviate from the legal definitions, intentional or erroneous reduction of legal obligations, discriminatory addition of further obligations, other internal contradictions or contradictions with EU law, misuse of standards for the purpose of excluding competitors via unfair requirements or interpretations, imprecise terms or formulations contrary to the rule of law, the granting of discretion to the user of the law. If all legal exclusion criteria are applied cumulatively, as has been practised at times in the medical devices and construction products sector, for example, virtually no standards can be referenced and thus integrated into the European legal system. We therefore end up in a dead end.  A review of recent European standards undertaken by the author has confirmed that the described deficiencies are still prevalent today.

VI. There are mainly two alternatives to the current system, one known and one largely ignored in the European legal framework:

1. The European Commission can adopt selfstanding harmonised technical specifications by means of implementing or delegated acts, as is already taking place in various legislative sectors. This route is the most legally secure, but also the most time-consuming. Moreover, it may lead to the EU distancing itself from the international mainstream in terms of its technical requirements, because the international standards only serve as inspiration, but no longer form the regulatory focus themselves; especially as the mere copying of standards in the technical specifications can lead to copyright problems.

2. The standards could also be deliberately removed from the EU law, but implicitly referred to by the backdoor.

a) This could be done if the legislator and the Commission no longer refer to the standards, but instead use abstract but dynamic terms such as ‘state of the art’ or ‘best available techniques³’. The legal practitioner, from the engineer of a manufacturer or service provider to the judge at the European Court of Justice, will then have to ask themselves what is meant by ‘state of the art’. And in this context, the European and international standards will automatically come back into focus, in their current and thus updated version. In the absence of good alternatives, the legal practitioner will ultimately apply precisely those standards that the legislator or the Commission has decided not to refer to formally. This leads to the same result, but the link of legal interpretation has been added. The standards are only incorporated by way of legal interpretation, not by way of legislation. This means that the standards no longer have to be scrutinised according to the principles of the rule of law. The path described under 2. has been followed in various EU countries. It has therefore been tried and tested in practice and is not fundamentally new.

b) The disadvantage of the approach described under 2. is, of course, a certain degree of legal uncertainty if there are various possible sources for the ‘state of the art’ or ‘best available technique’; and, on the other hand, that the legislator or Commission no longer has any permanent ex ante control. If the standards develop in a direction that is not desired, the Commission has to intervene ex post with its own technical specifications. Ultimately, therefore, a form of indirect control is also required. However, such a system is much easier to manage than the current one and certainly much simpler than a system based entirely on harmonised technical specifications from the EU Commission, which probably requires the most manpower.

c) Incidentally, the EU Commission could also steer this system described in 2. in a soft way: through so-called recommendations in accordance with Article 292 TFEU. By means of recommendations, the Commission could make clear how it believes the legal practitioner should deal with the standards and alternative sources for determining the ‘state of the art’ or ‘best available techniques’. Of course, it cannot do this in a binding manner. However, in practice, in the vast majority of cases the legal user and in particular the notified bodies will follow the recommendation, just as non-legal guidance documents are already applied almost like law. In order to avoid becoming legally liable for the content of the standards, the Commission could also limit itself to advising against the application of a standard in certain cases or to providing instructions for the application of standards, but not to positively recommending the use of certain standards; the recommendation would be made indirectly through the absence of a critical comment.

d) As a result, the Commission would have got rid of the onerous, ultimately unfulfillable responsibility for the rule of law control of standards with such a system as described in 2. The regulatory collapse risk created by standards that have been referred to but that do not fulfil legal requirements would be reduced⁴. At the same time, the flexibility and speed of the traditional standardisation organisations, which can be observed at least in many areas, would continue to be used. Finally, the Commission would retain the option of using its own technical specifications to overwrite standards that are inappropriate or contradict the legal requirements. This could result in a hybrid system in which the administrative and control effort is minimised and at the same time flexibility and internationalisation are maximised through the de facto inclusion of standards. The rule of law problem of access to translations would also at least be minimised. Finally, the system would also be accelerated insofar as the international (ISO/IEC) standards would no longer have to be converted into European standards first.

e) Critics may object that in the system described under 2. the industry would no longer have the certainty of having fulfilled certain legal requirements by applying the standard. However, in the current situation, it does not either, because the presumption of conformity usually only exists in relation to certain legal requirements, but not in relation to all of them. In addition, European standards repeatedly contain erroneous statements regarding the coverage of legal requirements or openly contradict them, so that a conscientious economic operator or notified body must first check the legal requirements. And finally, what is wrong with the fact that legal practitioners have to read the law more carefully if the standards ultimately cannot provide certainty of legal conformity? To read the law more carefully would be more conducive to compliance with the law.

f) Critics may also raise the issue that, in the system outlined in 2., the industry may lack sufficient incentives to utilise standards in the first place. Nevertheless, adherence to standards would offer the greatest probability of meeting the ‘state of the art’ or ‘best available technique’ requirement. Furthermore, standards have been observed to be beneficial, if not mandatory, in other regulatory systems. Consequently, there will be sustained interest in the utilisation of these standards. In instances where there is inadequate utilisation of standards, procedural advantages such as reduced conformity assessment obligations can be employed to incentivise their application. Similarly, the French “use or explain” approach could be incorporated into EU law. Regarding the variety of alternative means available to encourage the use of standards, please see again this article of mine.

g) However, the system described in 2. should not be applied in all cases. If the interpretation of the undefined, dynamic legal concepts diverges too far, the Commission must be able to intervene with its own harmonised technical specifications. The boundary between the cases where harmonised technical specifications and where the approach described in 2. is better will emerge over time and will certainly change over time. That is not a bad thing. After all, the approach described in 2. is only intended to extend the range of instruments available to the Commission and the legislator. In many cases, it will be completely sufficient, so that the complex harmonised technical specifications (or even the complex current reference system) can be dispensed with. Therein lies the gain. Legislators and the Commission would have another channel for achieving uniform economic practices with minimum effort. This would save considerable administrative resources overall and the industry would be less burdened by the undefined legal terms and the associated only indirect reference to the standards. For example, in such a system it would be definitively clear that the notified bodies would not be allowed to stubbornly insist on the application of the standards in atypical constellations, which they still often do today.

VII. Irrespective of the legal constraints discussed above, the Commission should be able to have the state of the art for new, innovative areas defined by private standardisation organisations, scientific institutes and other contractors. In the past, European standardisation organisations have often been unable or unwilling to develop certain urgently needed standards. It is therefore important that the Commission itself can award contracts to organisations other than the traditional standardisation organisations in order to quickly close gaps in the state of the art. Once the state of the art has been developed in this alternative way, the Commission should again retain all options in its hands: either only refer to the state of the art in legislation or use its own technical specifications to make the state of the art binding for those applying the law.

VIII. The following chart gives an overview on the differences between the three main ways to use technical specifications:

IX. Finally, let us take a look back, which makes the above proposals appear to be a logical continuation:

1. Thirty years ago, there were only the classic references to industry standards and the voluntary standards cited in the Official Journal, which were intended to trigger a presumption of conformity with legal requirements. Despite the obvious shortcomings of the latter system at the time, any thought of alternatives was frowned upon – it contradicted the prevailing New Approach ideology of the time to even consider regulatory alternatives; the lobbying of the European standardisation organisations, which were well funded by the Commission, and the industry’s interest in controlling technical requirements did the rest.
2. The first alternative, technical specifications to supplement or replace harmonised standards, saw the light of day in 1998 in Directive 98/79/EC on in vitro diagnostic medical devices and from there spread to more and more product areas via legislation on ordinary medical devices and fertilisers. Now this instrument is even to be generalised for all product areas as part of the 2025 Omnibus IV Simplification Package. At least this frees the Commission to a certain extent from the grip of the European standardisation organisations.
3. However, the liberation of the Commission from the grip of industry and the European standardisation organisations can still be further pursued. The proposals made in this article serve this purpose. Of course, it cannot be a question of taking less account of the legitimate interests of industry. It is only a matter of remaining capable of acting when impertinent reasons prevent the development of standards. For example, the author bitterly recalls how CEN was unable to standardise the colour coding of hospital gas pipes despite the Commission’s strongest intervention due to the resistance of a single manufacturer who regarded Germany and Austria as his monopoly territory⁵. The lack of standardisation has repeatedly led to confusion with fatal consequences in hospitals across Europe. Until today. This should not and does not have to be the case. That is why we need alternatives to commissioning the European standardisation organisations.
4. Greater candour is also needed when it comes to the legal oversight of standards in the current legal setting. For decades, the responsibility for determining whether certain European standards may be incorporated into EU law has fallen to low- to mid-level officials within the European Commission. These officials find themselves in a difficult position—caught between the political pressure exerted by industry and certain Member States and the demands of the legal order. This is neither a fair nor a sustainable arrangement. It requires structural reform of the kind outlined in this article. Until such reform is undertaken, the European Commission, as guardian of the Treaties, should take a clear and public stance in defence of the legal order. 

1 Case C-613/14. Strictly speaking, the predicament existed prior to the ruling and has simply been brought to light by it. Some product legislation sectors within the European Commission had already applied legal criteria to standards prior to 2010.

2 See e.g. cases Case C-276/01 Steffensen; C-443/03 Leffler; C-161/06 Skoma-Lux s.r.o. v Celní ředitelství Olomouc; C-14/07 Weiss und Partner;  e C-243/19 A v B and Others; C-566/19 P, Ryanair v Commission.

3 The second term is more stringent.

4 Should a medium-to-large-scale study be conducted that proves the investigated standards do not fulfil legal requirements, any economic operator concerned about standards could challenge their legal validity and thus their application. If the regulatory system is dependent on the standards, this would lead to its disintegration.

5 The respective manufacturer also sold gas pipes with the colour coding applied in other EU Member States. However, the manufacturer’s decision to hinder harmonisation ensured that he remained the only manufacturer selling gas pipes with the colour coding traditionally preferred in Germany and Austria.