Those versed in the subject of soft law are aware that there is no universally accepted definition of soft law in the literature. Indeed, the sheer diversity of phenomena that fall under the expression “soft law” renders the establishment of criteria for preference between definitions a challenging task. In fact, some authors argue that it is impossible to define soft law, given the diversity of phenomena that fall under the label of soft law.

1. Some approaches to defining soft law

A first approach to defining soft law involves deriving a definition from linguistic convention, particularly how legal scholars understand the term. However, despite the availability of definitions from judicial organs, the term “soft law” remains predominantly used in academic discourse rather than in legal practice.

A second approach might involve the determination of an exhaustive list of all soft law instruments, akin to the manner in which the term “European parliamentarian” is delineated by reference to each individual member of the institution. Nevertheless, this approach is fated to meet with failure for two fundamental reasons. Firstly, the incessant generation of novel soft law documents would demand regular revisions to the definition, thereby rendering it impractical and unwieldy. Secondly, such a definition would be excessively lengthy, making it difficult to manage effectively.

A third alternative could be to attempt the identification of not each instrument that can be identified as soft law, but only the types of soft law instruments, for example, communications, recommendations, advice, plans, charters, guides, etc. However, this approach to defining soft law does not appear to be very promising either, for at least two reasons. Firstly, there is a considerable number of atypical soft law instruments (i.e. those whose creation is not foreseen by any norm), which would raise doubts as to whether a particular type of instrument should be included in the list. Secondly, in order to determine the inclusion (or exclusion) of a given instrument within the definition, it is necessary to identify the characteristics that must be fulfilled for an instrument to be designated as soft law in the first place. This requires the delineation of the characteristics necessary for an instrument to be classified as soft law.

This third strategy appears to have inherent limitations, which necessitates the exploration of a fourth alternative: namely, the identification of the characteristics (sufficient conditions; or necessary and sufficient conditions) that an instrument (or a type of instrument) would have to satisfy in order to be considered soft law. According to Terpan, this strategy encompasses two distinct types of definitions. Firstly, there are those that attribute multiple characteristics as conditions (sufficient or necessary and sufficient) for an entity to be considered as soft law. Secondly, a school of thought exists which attributes a single characteristic to soft law. In this way, Terpan identifies a singular characteristic that distinguishes all soft law: “law without obligation” (Terpan, 2023, p. 46).

2. Two important definitions of soft law

Snyder and Senden are regarded as leading exponents of the latter type, with Snyder defining soft law as “rules of conduct which, in principle, have no legally binding force but which nevertheless may have practical effects” (1994, p. 2). In contrast, Senden’s definition is more nuanced, characterising soft law as a set of “rules of conduct that are laid down in instruments which have not been attributed legally binding force as such, but nevertheless may have certain (indirect) legal effects, and that are aimed at and may produce practical effects” (Senden, 2004, p. 4). As the discrepancies between these two definitions are minor, they will be considered together for the purposes of this discussion.

Contrary to Terpan’s assertion, it is evident that these definitions are not limited to a single property; rather, they are highly complex. Firstly, both definitions allude to the concept that soft law encompasses a set of behavioural norms. Secondly, both definitions allude to the fact that soft law has no binding force (Snyder), or does not have binding force in itself (Senden), due to the type of instrument used to give it shape. Thirdly, both definitions also point out that soft law can have practical (Snyder) and/or legal indirect effects (Senden). In the following, I will analyse each of these elements separately.

3. Soft law as a set of rules

The question of whether all soft law consists of rules of conduct is a complex one. If by ‘rules of conduct’ we mean exclusively those rules that categorise actions as either prohibited, obligatory or permissive (prescriptive norms), then, to maintain consistency, it would be concluded that soft law consists solely of permissions. This is because both definitions seem to state that these rules do not establish obligations. This is somewhat perplexing, however, given that it is not uncommon to encounter rules that stipulate obligations within instruments conventionally termed soft law. It should be noted that the argument does not claim that Senden and Snyder do not take account of rules of obligation or prohibition in their theories; rather, it is argued that their definitions are reductive in this respect.

However, the scope of soft law extends beyond prescriptive norms, encompassing two additional categories of entities. Firstly, soft law encompasses rules that establish definitions or allocate competences, which can be conceptualised as constitutive rules. These rules do not create obligations or prohibitions but establish definitions or confer powers on specific subjects. Secondly, soft law instruments frequently include other types of directives that are of lesser force than norms, such as opinions, recommendations and advice. This category comprises a wide array of directives that do not obligate to carry out the specified action, but only establish a type of action as advised, recommended, etc. For example, some road signs do not require you to drive below a certain speed, but only recommend a certain speed. Senden and Snyder do not entirely disregard these entities, but their definitions of soft law do not consider these elements.

4. The non-binding nature of soft law

The question thus arises as to whether all soft law is necessarily non-obligatory or non-binding. In order to answer this question, it is first necessary to establish the meaning of “non-obligatory” and “non-binding”. The term ‘obligation’ can be understood as a rule that establishes an action as obligatory. However, the term ‘binding’ is not synonymous with ‘obligatory’, as obligations can be imposed by rules that are not considered to be binding. A pertinent illustration of this phenomenon can be observed in the rules of classical Roman law, which designated specific behaviours as obligatory (e.g. the prohibition of marriage between patricians and commoners), currently non-binding. The notion of bindingness within legal systems does not equate to the anticipation of sanctions and the associated enforcement mechanisms. This is evidenced by the existence of numerous rules within contemporary legal systems that, despite the absence of provisions for sanctions, appear to be binding. This phenomenon is evident, for example, in the rules that stipulate the obligation of fidelity to one’s spouse (Art. 44 of the Spanish Civil Code) or those legal obligations addressed to individuals holding the highest public office (eg Article 62 of the Spanish Constitution requires the King of Spain to sanction and promulgate laws, but there is no sanction for failing to do so).

According to one approach, the concept of bindingness in soft law can be interpreted as signifying the characteristic of a rule that invalidates the judicial or administrative decision that does not apply it, provided there is an opportunity to implement it. Soft law is not generally considered binding, although there are some exceptions to this rule. The possibility of a soft law directive being regarded as binding arises when the hard law refers to it, although this is not the most interesting case; there are instances where the application of soft law directives has been deemed a necessary condition for the validity of decisions. For instance, the European Commission Notice on Cooperation within the Network of Competition Authorities (art. 2.1, 2004/C 101/03) stipulates that the network of competition authorities is responsible for allocating litigious cases to either a national authority or the European Commission (and there is no indication in any legal norm of any referral to this joint declaration). The convergence of the behaviour of legal agents, and particularly of judicial and administrative bodies, has the potential to result in a soft law directive becoming generally binding.

However, the question must be posed: when soft law is not binding, does its use become irrelevant in terms of its potential to provide a justification for a decision? An affirmative answer, it must be noted, is not particularly satisfactory. On the contrary, soft law is frequently employed to provide a justification for judicial and administrative decisions, and it does not appear that its use is mere ritual; rather, it justifies such decisions in areas of discretion. In summary, there appears to be that  the absence of binding status does not negate its importance in the justification of decisions. The question then arises as to the capacity of soft law to generate legal consequences.

5. The effects of soft law

Snyder posits that soft law engenders practical effects which can be understood as the modification of beliefs and behaviours in accordance with such norms. However, it can be contended that any norm, whether hard or soft, possesses the capacity to generate practical effects. In essence, any rule has the potential to exert legal effects. To illustrate this point, one may consider entering an illegal car race with a set of rules that permit car-to-car collisions.  In the 18th and 19th centuries, when a judge had to determine civil liability for damage caused during a duel, the court would refer to the rules of dueling, which could not be considered as coming from any authority (for example, in Spain, the book Lances entre caballeros by Julio Urbina y Ceballos-Escalera) and, in fact, were illegal.

Senden’s position is more precise in its delineation of the potential for soft law to engender indirect legal consequences. That is to say, the capacity of a soft law directive to effect a change in legal position is inherently limited; however, it may, under certain circumstances, give rise to legal effects. The assertion that all soft law engenders direct legal effects would undoubtedly be contentious. Nevertheless, certain components of soft law may, in fact, exert legal consequences that extend beyond their incorporation or referral by hard law.

In order to illustrate this point, consider the European Commission Notice on the enforcement of State aid law by national courts [2009] OJ C85/1. This instrument is the means by which the Commission announced how it would understand the term “state aid”. One possible interpretation of this Notice is that the Commission was merely announcing, informing, how it would understand the term in future state aid proceedings. However, public authorities rarely engage solely in the practice of informing member states or citizens about how they intend to interpret a term in the future.

Alternatively, the Commission Notice could be interpreted as constituting a rule. While it is the CJEU that has the ultimate authority in determining the interpretation of EU law, this does not negate the fact that the definition of state aid constitutes a rule. This is because when the administration defines a term such as ‘death’ or others, it does not constitute scientific truth or describe anything. Rather, it establishes the way in which the term must be interpreted in order for a legal decision to be considered valid. If the definition is deemed to be unlawful, the courts may declare the definition of death to be unlawful. In the same way that national courts are empowered to annul an administrative norm, the CJEU should be equally capable of annulling the rule formulated by the Commission through the Notice. Consequently, it can be posited that the rule delineating “state aid” is endowed with direct legal consequences until the point of its annulment.

Posted by ALVARO NUÑEZ VAQUERO

—

Álvaro Núñez Vaquero is Professor of Legal Theory at the University of Murcia, Spain. A more extensive analysis in Spanish of this blog post can be found at: https://revistaseug.ugr.es/index.php/acfs/article/view/31479/28693. This work forms part of the project Nuevas formas de creación y aplicación del Derecho, which is financed by the Autonomous Community of the Region of Murcia through the call for project grants for the development of scientific and technical research by competitive groups, included in the Programa Regional de Fomento de la Investigación Científica y Técnica de Excelencia (Plan de Actuación 2002) of the Fundación Séneca, Agencia de Ciencia y Tecnología de la Región de Murcia (Science and Technology Agency of the Region of Murcia).

—

References:

Senden, L. (2004). Soft law and in European Community law. Oxford: Hart Pu­blishing.

Snyder, F. (1994). Soft Law and Institutional Practice in the European Community, EUI Working Paper LAW No. 93/5.

Terpan, F. (2023). The definition of soft law. En Eliantonio, M., Korkea-aho, E. y Mörth, U. (eds.) Research Handbook on Soft Law (pp. 43-55). Cheltenham: Elgar.