Jump to Content Jump to Main Navigation
The Law of International Responsibility edited by Crawford, James; Pellet, Alain; Olleson, Simon; Parlett (Assistant), Kate (20th May 2010)

Part II International Responsibility—Development and Relation with Other Laws, Ch.14 The Concept of ‘Soft Responsibility’?

Jean-Marc Sorel

From: The Law of International Responsibility

Edited By: James Crawford, Alain Pellet, Simon Olleson, Kate Parlett (Assistant)

Responsibility of states — Wrongful acts — Reparations — Soft law

(p. 165) Chapter 14  The Concept of ‘Soft Responsibility’?

The question of ‘soft responsibility’ remains largely a doctrinal subject which can only be approached through inferences from scattered and informal practice. The starting postulate is clear: the universe of ‘soft law’ does not have an appropriate form of responsibility—itself ‘soft’—as a matter of positive law. This hypothesis is confirmed by the ILC’s silence: nothing in the ILC Articles supports directly or even indirectly the concept of soft responsibility. Articles 1 and 2 define internationally wrongful acts as actions or omissions attributable to a State and constituting a breach of an international obligation. These provisions are part of a framework of hard law, traditional law, in which it is the breach of an obligation combined with a wrongful act that engages responsibility.

Thus it is necessary to examine ‘soft responsibility’ with extreme caution. First, it is useful to reconsider the semantic aspect, since the terms used (in English, French, and other languages) seem to create confusion between concepts often not clearly formalized. Once the concept of ‘soft responsibility’ has been clarified, it is essential to consider it in its context, the only one in which it makes sense: that of ‘soft law’. Nevertheless, the nebula of soft law having been omitted from the ILC’s draft, we have to note that the monolithic character of international responsibility leaves aside many nuances in situations of responsibility, particularly in the sphere of certain international organizations.

Responsibility versus liability?

It appears that the French language is incapable of sufficiently translating all the nuances of the term ‘responsibility’. At least this is what Zemanek notes when he distinguishes between the words ‘responsibility’ and ‘liability’:

In some languages, like in French, one term (responsabilité) is used for all sources of the duty to repair damage, whatever the cause. In others, like in German, one term (Verantwortlichkeit) designates the consequences of an offence while another (Haftung) describes the duty to compensate. In English the duality of terms exists, too. International law adopted ‘responsibility’ to denote the (p. 166) consequences arising from the violation of an international obligation. The term ‘liability’ has only recently become part of international law, perhaps with the 1972 Convention on International Liability for Damages Caused by Space Objects, meaning the duty to compensate the damage in the absence of a violation of international law.1

The author pushes the analysis further by reference to such concepts as liability for fault, liability for result, objective liability, strict liability etc.2 But overall he suggests that responsibility is the consequence attached to the breach of an international obligation, whereas liability only correspond to the duty of reparation or of compensation in the absence of a breach of obligation (liability for acts not prohibited by international law).

It is true that in the dictionary the term ‘liability’ is defined to include obligation, commitment, responsibility; as international lawyers we would have to prefer ‘commitment’ to avoid confusion both with ‘responsibility’ as we understand it and ‘obligation’ (since the breach of an obligation is not required).3 The ordinary meanings of the term ‘responsibility’ cause even more confusion. If we take the moral meaning of ‘responsibility’ as being ‘to hold somebody responsible’, the legal meaning of responsibility would be ‘liability’ to which we must add ‘accountability’ for financial responsibility. Above all, responsibility and liability seem to correspond to different legal fields: civil including contractual liability; penal or criminal responsibility).4 This leads to a dual wrongfulness which the ILC did not contemplate.

One can infer from these semantic subtleties that the term responsibility corresponds to that used in the ILC Articles, whereas the term liability would be closer to the vast category of objective responsibility. But it is still necessary to agree on the sense of the latter. Indeed, objective responsibility and other types of responsibility (no fault, responsibility for risk or causal responsibility, etc) are often amalgamated but the meanings are very diverse. Objective responsibility might correspond to responsibility for a wrongful act where ‘fault’ is not a component element. But it is also understood as a responsibility in which only the wrongful act has to be considered independently of its harmful effects.5 Absolute responsibility (or liability), in which the damage is almost the only foundation of responsibility, is sometimes regarded as a synonym of objective responsibility. Causal responsibility could be a form of objective responsibility in which the obligation to compensate arises only from the causal link between the action and the damage.6 Finally, responsibility for risk is a kind of responsibility in tort in which compensation is due solely because of the occurrence of damage, even in the absence of any wrongful act.7 This last meaning brings us closer to ‘liability’ because it may concern activities not prohibited by international law but involving a risk of causing damage, implying an obligation of due diligence and an obligation to compensate. This is the sense of the ILC Draft Articles on the Prevention of Transboundary Harm from Hazardous Activities, in which article 1 provides that it (p. 167) applies to ‘activities not prohibited by international law which involve a risk of causing significant transboundary harm through their physical consequences’.8 The goal was to generalize the regime of numerous activities which had only been subject to individual conventional treatment (pollution of the sea, spacecraft, peaceful use of nuclear energy).9 Nevertheless, responsibility for risk is difficult to establish as the long gestation of the ILC’s Draft Articles demonstrates. The text seems to insist more on the principle of due diligence than the utility of a new regime. Responsibility for infringing rights as distinct from mere interests remains the rule.

In summary, we may note that, if responsibility corresponds to our responsibility in hard law, liability is also part of this universe when it corresponds to conventional agreements, to the recognition of a customary rule or to the recognition of a rule of progressive development of international law through codification in the context of responsibility for risk. Thus it is unnecessary to distinguish these two concepts in terms of soft or hard law. Both can concern legal obligations likely to engage responsibility. The distinction lies primarily in whether or not a breach of an obligation exists. In the case of responsibility, a breach of an obligation is considered a wrongful act; whereas in the case of liability, nonobservance of an obligation (of reparation or of due diligence) is not assimilated with a prior wrongful act. It is the adjective ‘soft’ which makes the difference in both cases. Thus we need to re-engage in this sphere to understand how a ‘soft responsibility’ or ‘soft liability’ could respond to ‘soft law’.

‘Soft law’ and ‘soft responsibility’

The scope of ‘soft law’ is vast and vague at the same time. This uncertainty is perfectly characterized by the remark of Seidl-Hohenveldern: ‘Soft law is something more than “law without any obligation”.’10 In fact, it refers to two phenomena: the loosely binding character of a legal norm deprived of an obligation in its classical meaning, and the pre-legal value of some of the norms. Thus ‘soft law’ refers to ‘non-law’ and ‘pre-law’ at the same time. According to the very critical view of Weil, this would not be sufficient to create ‘law’.11 Due to a ‘simplifying rigour’,12 he excludes the phenomenon of ‘soft law’ from the field of law, recognizing as points of reference only voluntarism, ideological neutrality, and positivism.13 This view has been somewhat overridden by writers who reject the combination of voluntarism and positivism on the ground that its realism is paradoxical and its claim to purity illusory.14 But Weil’s view has also, and especially, been rejected in the practice; an impressive number of measures can be considered part of soft law. Among these, the so-called ‘gentlemen’s agreement’ is characterized as an ‘agreement of honour that does not entail any legal obligation for the parties while binding them morally’.15 In the mode of informal political agreements, suppletive or normative, these have some legal value though (p. 168) not legally binding in themselves. The work of the Institute of International Law on this subject fails to capture this idea:16 a text may have some normative character without being peremptory or even binding. This enables us to distinguish rules and principles, breach of which is constitutive of a wrongful act (action or omission attributable to a legal person and constituting a breach of an international obligation legally sanctioned), from texts with legal effect although not comprising sanctioned obligations; it is the latter which forms the category of soft law. In summary, soft law is characterized by an absence of sanctioned obligations, but not by an absence of law. The true question is: can a legal person be found responsible for not respecting a legal text that does not include sanctioned obligations?

It is undeniable that soft law is defined by its form, and even more by the intention than the form. In this sense, the Stand-By Arrangements of the IMF are a good illustration of the subtleties of soft law: they may be taken as examples here.17 The question of the responsibility of the State towards the Fund, but also and more especially of the Fund towards the State, in the process for obtaining IMF financial aid (see eg, the Argentine crisis), had often been evoked; but no-one has managed to define strictly the nature of the support given (Stand-By Arrangements) or the content of the obligation incurred (whether of result or conduct). Thus we definitely have to turn toward the intention of the parties to seek a possible form of implicit responsibility without a contractual relationship—an absence which lies more in the intention or discourse of the Fund than in the form adopted. Clearly, there is no right undermined but, as with sanctions within the IMF, there is a form of soft and graduated liability. The process of a ‘letter of intent’ (a unilateral decision of the State wishing to have recourse to the IMF) and the decision of the Fund (which constitutes what is improperly described as a Stand-By Arrangement) are not legally linked, and there is no right recognized as such. Moreover, the IMF’s refusal of responsibility is extended to third parties who might seek to rely on this Arrangement, for example to grant credits through private bank loans. Thus a relationship exists between the IMF and the State (without legal obligation) which creates a kind of commitment of mutual interest.18 This phenomenon recalls article 9(2) of the ILC’s Draft Articles on Prevention of Transboundary Harm from Hazardous Activities where it is specified that the States, via a notification proceeding, shall seek solutions based on an ‘equitable balance of interests’ (an expression specified in article 10).

As can be seen from this example, situations of soft law are not likely immediately to form the basis of the responsibility of a subject of international law, even for risk (since there is obviously no obligation of due diligence or of reparation). In the case of the IMF, it would be necessary for the State (in the most plausible case) to challenge the responsibility of the organization (the decision is made by a restricted body, the Executive Board) and not by the member States. This must be difficult to do in practice. Moreover it would be especially difficult to establish a direct causal relation between the normative activity of the organization and any damage inflicted on the State. Nevertheless, just as soft law is not generally synonymous with the absence of law, so the absence of any ‘immediate’ (p. 169) responsibility is not synonymous with the complete lack of any kind of responsibility. Responsibility will simply have to be sought outside the traditional framework, in related but not assimilated principles which range from good faith to estoppel, while respecting the permissive value of the texts and the principle of non-contradiction.

Good faith is the first principle to consider. Concerning the Stand-By Arrangements of the IMF, for example, it is essential to go beyond the presentation ‘proposal-answer’ which appears prima facie to suggest a reciprocity of commitments (‘promise-guarantee’). By doing so, one can see that the operational content is not devoid of consequences even if the State seems to make a promise to itself which places it face-to-face with its own responsibility. Indeed, the ‘promise’ is supported by the guarantee of the IMF; the actors are not isolated. Moreover, good faith implies that any breach involves consequences, even if the situation is simply that of two parallel unilateral acts. The responsibility targeted here is the one pertaining to conduct which is not described as mandatory but which is supported by a presumption of good faith. This good faith implies a moral commitment based on the recognition of a ‘defect’ (and not a fault or an illicit act) and driven by the principle of reciprocity. Of course, in an implicit way, there is the idea of risk related to the expected effect that might emerge from a certain commitment (what the Anglo-Saxons describe as justifying effects). Indeed, although not constraining, the commitments of soft law give rise to expectations whose consequences must be appreciated in this particular sphere in which the permissive value of the norm and the principle of non-contradiction belong.

Alternatively, it may be that estoppel could be invoked to reflect these expectations. If this form of responsibility cannot be immediately included within the framework of soft law, the concerted non-conventional act might have created expectations which can allow the partners to invoke the estoppel principle as a form of ‘muffled’ responsibility. It is a notion inspired by the Anglo-Saxon law of evidence which expresses the principle of compatibility. This last principle is so defined by Martin.19 Thus estoppel corresponds to a non-conventional situation having effects on third parties and enables us to override the requirements—which seemed insuperable—of the fixed relationship between a legal obligation, its breach and its sanction. It is not a coincidence if the IMF is protecting itself against the private loan agreements in the tripartite relationship between itself, the State, and private banks. Indeed, if the private lender turned out to be unsatisfied with his commitment, the State concerned could turn against the program suggested by the Fund thereby leading to a chain of causation perhaps rising to the level of estoppel, ie a form of ‘soft responsibility’. Nevertheless, this concept remains badly defined and hardly useable, as is demonstrated by the hesitations of the International Court in this regard.20

Notwithstanding some particular cases, the recognition of the inclusion of a ‘soft responsibility’ is not founded because the general regime of responsibility envisaged by the ILC for States is characterized by two elements: it is interstate and it understands hard law in a univocal way.

(p. 170) Soft responsibility vis-à-vis international responsibility

It would be wrong to say that the ILC ignored soft law—it simply did not consider it as part of the form to be given to the Articles, while recognizing that their adoption in the form of a declaration would oblige States opposed to it to demonstrate that they are not bound by it, and while affirming that soft law can have a decisive impact in international relations and in State practice.21 But these positions have no repercussions in terms of the content of the text itself: international responsibility. The result is that responsibility in international law does not contain the nuances and subtleties which exist in domestic law because of the boundaries of sovereignty. This can be seen in the commentary to article 1 which notes that the term ‘délit’ was set aside in favour of the ‘fait internationalement illicite’ to avoid confusion with its particular meaning in domestic law.22

Thus we observe an orthodox legal framework faced with the heterodoxy of real situations. Responsibility is not flexible and adjustable according to varying behaviours. ‘Soft responsibility’ is thus overridden by the rigidity of the wrongful character of the act as the foundation of the international responsibility of the State. The idea of risk is avoided (except in particular conventions and in related projects—limited as these are—of the ILC). Some critics hold that this is a good thing since it is for States to face up to their own responsibilities in the matter of risk-taking.23 This is so even though the same author seems to think, a few years later, that responsibility is refined according to the character of the obligations violated—even where this is an obligation to compensate—while preserving its unity due to the requirement of breach of obligation even when the underlying activities are not prohibited by international law. By suspicion or mistrust, we nevertheless compel ourselves to a simplification of responsibility: no distinction between civil, penal, contractual, tortious responsibility, or between obligations of conduct or of result with respect to the breach of an obligation.24

The construct of international responsibility (it is indeed a construct, something new based on materials provided) is a fine one. What remains unknown is what its practical use is going to be; if the real regime of responsibility will not be left by the wayside, as it already has been in the sphere of private agents (particularly within the framework of oil pollution). The Montreal Protocol of 1987 or the Vienna Convention on the Protection of the Ozone Layer had already set up a form of ‘soft’ responsibility which does not speak its name, in the form of a non-compliance procedure. The Kyoto Protocol reproduces the idea, also under the name of a non-compliance procedure.25 We note its softened definition of the wrongful act which in consequence assists more the offender than would sanctions. Thus conventional practice seems to soften what appears to be forbidden by the general framework.

Another obstacle is added to this monolithic framework: the avoidance of the question of international organizations. The IMF example (among many others) shows that soft law lies largely within the field of international organizations which, as we know, can be (p. 171) held as responsible, if some legal framework for implementation can be located. Article 57 leaves aside all questions concerning the responsibility of an organization and of a State for the behaviour of an organization. Of course, the ILC recognized that the Articles could have included those cases where a State is responsible because of its implication in the organization’s behaviour or because of its status as a member. But the controversy seemed too difficult to solve and the ILC left this subject aside, even if this rejection applies only to ‘derived’ or ‘secondary’ obligations and does not exclude the State responsibility for its own conduct.26 Paradoxically, this leads the organizations to adopt a kind of ‘soft responsibility’ by means of an empirical practice. This is notably the case regarding the UN in its peacekeeping operations. Lacking an adequate framework, the UN has constructed, in a pragmatic way, a self-assessment of harms caused in an original and uncategorized system.27 Nevertheless, in other institutions, like the IMF, the organization continues to disregard its responsibility, preferring the self-control and self-responsibility of the State, which makes promises only to itself and thus is responsible for its own damage. For the IMF to shoulder the risks it places on the State, these programmes would undoubtedly need to reach a postulate of infallibility in economic matters, which is far from being the case. The IMF shows a remarkable propensity to protect itself against its own responsibility.

Thus the recognition of ‘soft responsibility’ remains a timely subject, particularly regarding the legal framework that needs to be created for the implementation of the responsibility of international organizations. This situation is regrettable given that the judge and the sanction are not the best guarantees for the execution of an obligation in international law. On the other hand, they are the best guarantees when a balance of interests is reached. But an ‘interest’ is not an ‘obligation’ and international law is only aware of the latter in the field of responsibility. What remains is for the process to evolve towards the crystallization of norms which will perhaps be recognized in the future and therefore will become projects of progressive development of international law. In the meantime let us hope that the soft will not change into the hard but will remain what it is intrinsically. If not, we would be back at the starting point.

Further reading

  • L Condorelli, ‘L’imputation à l’Etat d’un fait internationalement illicite: solutions classiques et nouvelles tendances’ (1984-VI) 189 Recueil des cours 9
  • P-M Dupuy, ‘Le fait générateur de la responsabilité internationale des Etats’ (1984-V) 188 Recueil des cours 9
  • P-M Eisemann, ‘Le Gentlemen’s agreement comme source du droit international’ (1979) 2 Journal du Droit International 326
  • U Fastenrath, ‘Relative Normativity in International Law’ (1993) 4 EJIL 305
  • B Graefrath, ‘Responsibility and Damages Caused: Relationship between Responsibility and Damages’ (1984-II) 185 Recueil des cours 9
  • C Nègre, La responsabilité internationale pour les atteintes massives à l’environnement (Thèse, Paris X-Nanterre, 2003)
  • J-M Sorel, ‘Sur quelques aspects juridiques de la conditionnalité du F.M.I. et leurs conséquences’ (1996) 7 EJIL 42
  • P Weil, ‘Vers une normativité relative en droit international?’ (1982) 88 RGDIP 5
  • K Zemanek, ‘Causes and Forms of International Liability’, in G Schwarzenberger, B Cheng, & ED Brown (eds) Contemporary problems of International Law: Essays in honour of Georg Schwarzenberger (London, Stevens and Sons, 1988), 319(p. 172)


K Zemanek, ‘Causes and Forms of International Liability’, in Contemporary Problems of International Law: Essays in honour of Georg Schwarzenberger (London, Stevens and Sons, 1988), 319.

Ibid, 327.

Commerce, Finance, Law Dictionary—English, German, French (Lausanne, Robert & Herbst, 1955).

Robert & Collins French Dictionary (5th edn, Paris, Dictionnaires Le Robert, 2000).

A Pellet, ‘Remarques sur une révolution inachevée, le projet d’articles de la C.D.I. sur la responsabilité des Etats’ (1996) 42 AFDI 12.

See J Barboza, ‘La responsabilité “causale” à la Commission du Droit international’ (1998) 44 AFDI 513–522.

Concerning these distinctions, see J Salmon et al, Dictionnaire de Droit international public (Brussels, Bruylant, 2001), 994–1000.

ILC Yearbook 2001, Vol II(2), 144.

See C-A Kiss, ‘La réparation pour atteinte à l’environnement’, in SFDI, La responsabilité dans le système international (Paris, Pedone, 1991), 225.

10  I Seidl-Hohenveldern, ‘International Economic Soft Law’ (1979-II) 163 Recueil des cours 194.

11  P Weil, ‘Vers une normativité relative en droit international?’ (1982) 88 RGDIP 5.

12  Ibid, 44.

13  Ibid, 19.

14  See U Fastenrath, ‘Relative Normativity in International Law’ (1993) 4 EJIL 305.

15  P-M Eisemann, ‘Le Gentlemen’s agreement comme source du droit international’ (1979) 2 Journal du Droit International 326, 327.

16  See the study ‘Distinction entre textes internationaux de portée juridique et textes dépourvus de portée juridique’, Annuaire de l’Institut (Paris, Pedone, 1983), Vol 60-I, 166; Vol 60-II, 117, 284.

17  See J-M Sorel, ‘Sur quelques aspects juridiques de la conditionnalité et leurs conséquences’ (1997) 7 EJIL 42.

18  On the distinction between the breach of obligation and infringement of interests, see B Graefrath, ‘Responsibility and Damage Caused: Relationship Between Responsibility and Damages’ (1984-II) 185 Recueil des cours 9, esp 47ff.

19  A Martin, L’estoppel en droit international public (Paris, Pedone, 1979), 260.

20  The concept was recognized by the ICJ in North Sea Continental Shelf (Federal Republic of Germany/Netherlands), Judgment, ICJ Reports 1969, p 3, 26 (para 30), where it was distinguished from acquiescence due to the existence of detriment suffered by the party relying upon it. It was also relied upon in Delimitation of the Maritime Boundary in the Gulf of Maine Area (Canada v United States of America), Judgment, ICJ Reports 1984, p 246, 305 (para 130), where it was assimilated to the idea of preclusion, and in Elettronica Sicula SpA (ELSI) (United States of America v Italy), Judgment, ICJ Reports 1989, p 15, 44 (para 54), where it was said to be a possible consequence of silence.

21  Report of the ILC, 53rd Session, 2001, A/56/10, 24 (para 64).

22  The word ‘fait’ was also preferred to ‘acte’ which does not include omissions in the French language, which is not the case of ‘act’ in English, ibid, 70 (para 8).

23  P-M Dupuy, ‘Le fait générateur de la responsabilité internationale des Etats’ (1984-V) 188 Recueil des cours 9, 108.

24  See also art 12, ARSIWA which makes clear that there is a breach of an obligation ‘regardless of its … character’; see also Commentary to art 12, para 11, ILC Yearbook 2001, Vol II, 54.

25  C Nègre, La responsabilité internationale pour les atteintes massives à l’environnement (Thèse, Paris X-Nanterre, 2003), 394.

26  Commentary to art 57, esp para 5.

27  Ibid.