1. Terminology and concept
Strict liability means liability for injury or damage to another person without fault, ie intention or negligence. The legal ‘responsibility’ does not require the tortfeasor to neglect his duty of care consciously or unconsciously. The term Gefährdungshaftung is used in Germany, but no precise counterpart to this term exists in other jurisdictions. The French describe liability without fault as responsabilité du fait des choses, while in the countries of the common law, the term ‘strict liability’ is used.
These differences are not merely of a terminological nature but an expression of factual differences. All three concepts—Gefährdungshaftung, responsabilité du fait des choses and strict liability—share as their common feature that fault is not required. However, the concepts do differ in terms of the element of attribution that is supposed to replace the prerequisite of fault. While the concepts of responsabilité du fait des choses and strict liability do not address this directly, the concept of Gefährdungshaftung (literally, ‘liability for endangerment’) makes the element of attribution explicit in its name: the defendant is liable because he created or controlled a source of danger that poses an increased risk of harm to others. The purview of strict liability in this sense primarily concerns the operation of, and control over, technical equipment or installations, but also the keeping of animals. The French doctrine of custodian’s liability is different because, according to the dominant view, it is not limited to dangerous objects but applies to objects of any kind. For this reason, custodian’s liability certainly qualifies as a case of liability without fault but not as one of ‘strict liability’ in the technical sense of the term, ie responsibility for a source of danger.
Where clear statutory or judge-made rules are missing, the distinction between fault-based and strict liability is difficult to draw, and it remains controversial in many cases. Based on the practical outcomes, the two concepts are more closely related than their theoretical polarity would suggest. The standard of care, which is crucial for the determination of fault-based liability, is contingent upon the magnitude of the harm in question and the probability of it occurring. Therefore, the operation of a source of danger, in itself, tends to increase the number of precautions the operator must take in order to avoid liability. The more serious the danger, the closer fault-based liability moves towards strict liability. Moreover, strict liability merely abandons fault on the part of the tortfeasor but still measures the conduct of the victim against the standard of due care under the concept of contributory negligence. For these reasons, scholars tend to highlight the commonalities instead of the differences between the two liability regimes and tend to talk, not of two distinct categories, but of a grey area between strict and fault-based liability. Nonetheless, an important conceptual and practical difference remains: fault-based liability is contingent upon proof of fault, ie intention or negligence, so that a loss that occurs even though the defendant has taken due care, is borne by the claimant (casum sentit dominus). Under a strict liability regime, conversely, the person responsible for the danger is liable for all damages regardless of his fault.
2. Economic rationale
The distinction between fault-based and strict liability has gained support from the field of economics (see economic analysis of European private law). As the economic analysis of liability law has revealed, the concepts of liability for fault and strict liability are equally suited to generate incentives to take the efficient amount of care, ie to take precautions that cost less than the harm averted by these same precautions. Fault-based liability fails, however, in controlling the amount of harmful activities, while strict liability also has the potential to influence the ‘level of activity’ of a would-be defendant. Under strict liability, the injurer has to bear not only the per unit costs of the activity in question plus the costs of taking care but also the costs of the residual loss that is caused even though due care has been taken. Thus, the calculation of the actor includes the total social costs of the activity in question. A rational actor will weigh up costs and benefits of activities in deciding whether to engage in one activity or another, eg in deciding between the railway and a motorcar as a means of transportation. For this reason strict liability should be preferred in situations where the loss caused by the activity in question cannot be eliminated or reduced to a negligible amount through due care, ie by taking precautions that cost less than the harm in question is worth. Given these conditions, it is important to provide the actor with incentives only to engage in the respective activity where the utility derived from it exceeds all direct and indirect costs, including the loss suffered by third parties. Conversely, the law should settle upon fault-based liability where the only decision of the actor that needs to be influenced is the choice between different levels of care because the taking of due care essentially suffices to protect third parties from harm. Here, the law must only see to it that courts fix the level of care correctly. This distinction ties in nicely with the idea, developed within the common law of torts and the German law of delict, that strict liability should be limited to abnormally dangerous activities.
3. National systems of strict liability
In the area of fault-based liability, the European jurisdictions seem relatively homogeneous, while the area of strict liability displays great differences. French law accords strict liability a broad scope; the English common law is far more reluctant, while German law takes an approach somewhere between the two. The great variations in strict liability are rooted in history and also in political choices. Fault-based liability has been an institution of European legal systems since the days of antiquity, as it developed from the Roman lex Aquilia (law of torts/delict, general and lex Aquilia). It is true that Roman law had established actions that, today, would be brought under the heading of strict liability, eg with respect to damage caused by domestic or wild animals (actio de pauperie, edictum de feris), with regard to objects thrown from houses onto the street (action de deiectis vel effusis) or certain types of innkeeper liability (actio de damno aut furto adversus nautas, caupones, stabularios). Most of these actions, however, were eroded over the course of time and either integrated into the general fault liability or placed into special niches. In its modern guise, strict liability is a product of the industrial revolution, ie the late 19th century. It is a reaction to the development and proliferation of mechanical and technical installations and appliances.
The birth of Gefährdungshaftung in Germany took place in 1838, when Friedrich Carl von Savigny drafted the Prussian Railway Operators Act (Preußisches Gesetz über die Eisenbahn-Unternehmungen). It imposed strict liability on railway operators for damage caused as a result of operating trains. Despite this early and prominent precursor, the German Civil Code (Bürgerliches Gesetzbuch (BGB)) settled on the fault requirement, with the single exception of § 833 BGB regarding the liability for damage caused by animals. As a result, strict liability became the domain of special statutes outside the Civil Code. The legislature, on its part, eschewed the creation of a general clause of strict liability and instead followed the principle of enumeration by creating a catalogue of specific statutes, each of them introducing strict liability for a particular category of activity or installation. This was the case for the operation of trains, cars, aeroplanes, nuclear power plants and facilities hazardous to the environment (see environmental liability). The list of strict liability activities that was so created is constantly being extended as advances in science and technology create new sources of danger, eg—recently—in the area of genetic engineering. Still, the fragmented character of strict liability remains unchanged. German courts remain faithful to the political prerogative by refusing to create new heads of strict liability, or even to extend existing ones by way of analogy (German Federal Supreme Court, BGH 25 January 1971, BGHZ 55, 229). This position is in stark contrast to the Austrian case law that subscribes to the application by analogy of provisions ordering strict liability (Austrian Supreme Court of Justice, OGH 11 October 1995, JBl 1996, 446).
English law is even more reserved when it comes to strict liability. It was recognized in the middle of the 19th century in the House of Lords’ decision of Rylands v Fletcher  LR 3 HL 330, where the defendant was held strictly liable for damage to a neighbour’s property caused by substances which had escaped from a facility that the defendant had maintained on his land. While it appeared for some time that the rule in Rylands v Fletcher would develop into a general clause of strict liability, this has eventually proved not to be the case. The House of Lords, as it used to be called, has refused to renege on the requirement that a relation to the land must exist and, thus, deliberately avoided extending the rule of strict liability to dangerous facilities or activities of any kind (Read v Lyons  AC 156 (HL)). More recently, the House of Lords has even required that the damage must be foreseeable, which moves the rule in Rylands v Fletcher very close to fault-based liability (Cambridge Water Co v Eastern Counties Leather Plc  2 AC 264 (HL)). The Court was of the opinion that it was the task of the legislature—and not of the courts—to create categories of strict liability (Transco Plc v Stockport Metropolitan Borough Council  UKHL 61). In fact, Parliament has passed a number of statutes each of them subjecting a class of technical appliances or installations to strict liability, such as for the operation of nuclear or waste disposal facilities. In stark contrast to the legal situation in most continental European countries, the operation of motor vehicles is not subject to strict liability. The liability for having caused a traffic accident remains exclusively the domain of fault-based liability.
Article 1385 of the French Code civil contains the classic case of strict liability: the keeping of animals. Of greater importance today is Art 1384 Code civil, dealing with the liability of the ‘custodian’ (gardien de chose). The liability of the custodian does not require fault and is therefore ‘strict’, but it does not require the control of a ‘source of danger’. Even in France, strict liability in the true sense of the term, ie for the operation of technical appliances and installations, remains the domain of the legislature. Among the special statutes, the loi Badinter of 1985 stands out as particularly important. It subjects the operator of a motor car to strict liability for traffic accidents vis-à-vis non-motorized victims like pedestrians, and even abandons the fault element as a measure for apportioning the damages between two or more motorists who were involved in an accident.
4. Perspectives of harmonization
Time and again, scholars of comparative law have called for a general clause of strict liability for keepers of a source of danger, in order to replace the patchwork of special statutes. The current proposals for harmonization of the European law of torts reflect the choice between these two models, ie the general clause on one hand and the principle of enumeration on the other.
The European Group on Tort Law (Principles of European Tort Law) responded to the call for a general clause of strict liability. According to Art 5:101 PETL, a person is strictly liable if he carries out an abnormally dangerous activity. An activity is deemed to be abnormally dangerous if it creates a foreseeable and highly significant risk of damage even when all due care is exercised in its management. An activity is not abnormally dangerous when it is a matter of common usage. As a result, motor vehicles are excluded from the scope of the general clause, which significantly reduces its practical impact. The Austrian draft on the reform of the law of damages (Entwurf zur Reform des Schadensersatzrechts) by and large follows the example set by the PETL. It was published in 2005, with a revised version issued in 2008 ((2008) Juristische Blätter 365).
The authors of the Draft Common Frame of Reference (DCFR) declined to subscribe to a general clause. Instead, they opted for the principle of enumeration, which was spelt out within the general provisions on non-contractual liability instead of being shirked off to special statutes. The third chapter of the sixth book of the DCFR defines categories of strict liability for keepers of ‘immovables’ which are in an unsafe condition (Art VI.-3:202 DCFR), for keepers of animals (Art VI.-3:203 DCFR) and motor vehicles (Art VI.-3:205 DCFR), and also for ‘keepers’ of substances and operators of installations dangerous to their environment (Art VI.-3:206 DCFR). Moreover, Art VI.-3:204 DCFR codifies the provisions of the Product Liability Directive (Dir 85/374).
A common feature of both regimes is that their provisions on strict liability, be it a general clause or an enumeration of particular categories, are not exhaustive, but rather contain clauses that allow for national laws to establish further cases of strict liability. According to Art 5:102 PETL, national laws may introduce further heads of strict liability, regardless of whether the activity in question satisfies the requirements of Art 5:101 PETL, ie that it is not abnormally dangerous or a matter of common usage. This allows most Member States to retain strict liability for the operation of motor vehicles. The DCFR (Common Frame of Reference) solves the problem of motor vehicles itself but does not desist in including an opening clause for the benefit of the Member States (Art VI.-3:207 DCFR). Prima facie, it may be surprising that the projects on unification have not managed conclusively to exhaust the categories of strict liability, whether through a catalogue of sources of danger or a general clause. This failure is evidence of how diverse the opinions on the proper scope of strict liability still remain. Even within relatively small groups of scholars it is apparently impossible to achieve a consensus on which activities qualify for strict liability and which should better be left to the fault principle.
5. European Union law: products liability
The European Union has not touched upon non-contractual liability very often. Apart from the specialized domain of Member States’ liability for breaches of Union law, the Product Liability Directive is a case in point. The drafters of the Directive intended the introduction of strict liability in the sense that it was their goal to hold the manufacturer strictly liable for any damages caused by defects in the products he put on the market. However, a closer analysis reveals that the requirement of negligence has merely been shifted into the concept of a product defect under Art 6 of the Directive. The defectiveness of a product, ie its non-conformity with the standard of safety that the consumer is entitled to expect, can only be established on the basis of an investigation of the safety measures taken by the manufacturer. In the areas of design defects and marketing defects (failure to warn), the definition of defect is merely a disguised version of the standard of safety a diligent manufacturer would have observed. Arguably, however, strict liability—in its technical sense—does exist for manufacturing defects (production flaws) because the manufacturer is also liable for the ‘outliers’, which could not have been avoided even if due care had been taken.
The scope of protection of the Directive and the national legislation it has engendered is limited to death, personal injury and damage to property (Art 9). Harm to the product itself and to property that had been used for commercial purposes are not covered by the Directive. Furthermore, with regard to damage to property, a deductible of €500 applies. In spite of its fragmented protective perimeter, the European Court of Justice (ECJ) has held that the Directive has ‘fully harmonized’ the law on product liability and, as a result, is binding upon the Member States, with the result that national lawmakers must not deviate from the Directive, neither by tightening nor by relaxing its standards (ECJ Case C-402/03 – Bilka  ECR I-199). Against this background, it serves as a cushion that the Directive does not touch upon the national regimes of contractual liability and liability in tort, which remain applicable alongside the scheme of the Directive (Art 13 Dir; ECJ Case C-52/00 – Commission v France  ECR I-3827).
6. International conventions
Specific categories of strict liability for particular sources of danger are also to be found in international conventions. This is particularly true for environmental liability. Beyond this, aviation should be mentioned, as the recent Montréal Convention on the Unification of Certain Rules for International Air Carriage has rearranged and modernized the principles of liability for airline accidents. Under the convention, the carrier is subject to strict liability for the death or bodily injury of a passenger caused by an accident on board the aircraft or in the course of embarking or disembarking. The liability is capped at 100,000 Special Drawing Rights (around €120,000) for each passenger. Beyond this cap, the Convention provides for fault-based liability with a rebuttable presumption of fault, just as provided by its precursor, the Warsaw Convention of 1929. For air carriage within a Member State, the Air Carriers Liability Directive (Dir 2027/97) provides that personal injuries are to be compensated according to the Montréal Convention and also makes additions for damage to luggage.
The strict liability of innkeepers (see innkeepers’ liability) for damage to the guests’ luggage has its roots in Roman law (namely the receptum cauponum). Nowadays, the 1962 Paris Convention on the Liability of Hotel-Keepers concerning the property of their guests applies. Whether the liability of innkeepers vis-à-vis their guests is a true case of non-contractual liability or rather one of liability for breach of contract ‘as defined by statute’ is debatable.
7. Conflict of laws
The conflict of laws rules regarding non-contractual liability have been harmonized by the Rome II Regulation. It applies to claims for damages, regardless of whether they are based on fault or on categories of strict liability. The law applicable to an international dispute is generally that of the country where the damage occurred. In addition, the Rome II Regulation supplies special rules for a range of cases such as environmental liability and products liability.
Steven Shavell, ‘Strict Liability versus Negligence’ (1980) 9 JLS 1; Reinhard Zimmermann, The Law of Obligations (1990) 1095 ff; Christian von Bar, The Common European Law of Torts, vol II (2000) 333 ff; Bernhard A Koch and Helmut Koziol (eds), Unification of Tort Law: Strict Liability (2002); Gerhard Wagner, ‘Grundstrukturen des Europäischen Deliktsrechts’ in Reinhard Zimmermann (ed), Grundstrukturen des Europäischen Deliktsrechts (2003) 189; Nils Jansen, Die Struktur des Haftungsrechts (2003); Franz Werro and Vernon Valentine Palmer (eds), The Boundaries of Strict Liability in European Tort Law (2004); Gert Brüggemeier, Common Principles of Tort Law (2004) 82 ff; Cees van Dam, European Tort Law (2006) 255 ff; Reinhard Zimmermann, ‘Innkeepers’ Liability—Die Entwicklung der Gastwirtshaftung in England’ in Festschrift Claus-Wilhelm Canaris (2007) 1435; Bernhard Koch, ‘Other Strict Liabilities’ in Helmut Koziol and Reiner Schulze (eds), Tort Law of the European Community (2008) 135.