Strict liability in tort law
by BILL W. DUFWA1
There was a young lady of Riga Who rode with a smile on a tiger; They returned from the ride with the lady inside
And the smile on the face of the tiger.
A. Strict liability2 — negligence law
1. One of Jean Limpens' most important contributions to the understanding of tort law and its development was the second chapter ofthat volume of the International Encyclopedia of Comparative Law, which covers the law of torts (vol.XI).3 In this work, Limpens starts from one of the central elements of tort liability: one's own act. The technique was simple but certainly arduous to apply in detail. It proved to have great merits. Thanks to it, we have a better understanding of the differences between the tort laws of various countries.
2. One's own act may have been negligent, and so a liability may befounded on negligence law. One's own act may, however, also lead toliability without negligence, namely if there exists strict liability for the given situation. Even though he has not been negligent, a driver maybe liable for his way of driving, provided that there is strict liability fordrivers. Therefore, when Limpens discussed liability for one's own act, he might very well have let it embody both liability built on negligence law and strict liability.
3. However, he did not. From the international point of view Limpens distinguished three major trends in tort law, all considering
the number of general rules: the single rule approach (adopted by most countries whose legal systems are based on French law, as also by the socialist countries, Scandinavia, and existing in the modern law of the Arab world), restricted pluralism (found in certain countrieswhose legal systems are based on German law) and unrestricted pluralism (the common law and the traditional Mohammedan law).They all had as their starting point the concept of fault.4 According to this, Limpens' survey had its centre of gravity fixed on the fault principle.5 Strict liability was treated only exceptionally — it was left to be handled in more detail by other authors of the same volume of the International Encyclopedia of Comparative Law.6
4. Limpens' way of treating tort law is worthy of our attention. Although there is a clear difference between negligence law and strictliability, and although this difference is commonly accepted, there is atendency for them to be confused in some respect or other in modernlaw. Perhaps the most obvious example of this tendency concerns the question of cause in law.
5. It often happens that causation in tort law is regarded in the same way, whether the liability discussed is based on fault or not. But the truth is that causation in the field of strict liability might be something quite different from causation in negligence law.7 The owner of a dog,who locks up the dog properly, but who due to strict liability is nevertheless liable for the damage caused by the animal after it has been let out by another person, can hardly be said to have "caused" the accident, at least if you make use of normal language.8 Behind strict liability one finds risk considerations of quite a different kind from those one meets in negligence law; and this can occur also when the causation question is decided.
6. The matter now being considered is not only a theoretical one. It can be highly practical. This may be illustrated by a case decided by the Swedish Supreme Court in 1982.9 Patients claimed that they had been harmed by a drug. The courts had to decide two questions, firstly: was there a causal relationship between the drug and the injuries, and secondly: should strict liability be introduced in this field
of tort law, where negligence law hitherto reigned? The SupremeCourt treated these questions in the order mentioned: firstly thecausation question, secondly the strict liability issue. The court foundthat there was no causal connection between the drug and the injuries.Therefore it never had to decide on the issue of strict liability. The causation issue seems to have been hard to decide — two judges of six also came to a different conclusion. But if the order of the questions had been reversed, and if the Supreme Court had come to the conclusion that strict liability should be introduced — then the causation question might have been considered differently. For one may say that if there is reason enough to introduce a strict liability in favour of the victim, there is also reason to relax the demand for proof of causation. There is no certain evidence that Swedish courts would argue in that way.10 It is even possible that they might reflect to the contrary: a heavy responsibility entailed by strict liability ought to be compensated for in favour of the defendant by raising the demand for proof of the causal relationship. Nevertheless there is a reason for emphasizing the possibility of mitigating the demand for proof that strict liability can bring about.
7. Litigation coincidences seem to have decided the order in which the two main questions were decided in the 1982 case. In another case, however, decided by the Supreme Court in 198311 the question of how far the causation issue could be influenced by strict liability considerations appears to have been brought to a head, although it is doubtful whether the plaintiff and the defendant were aware of the problem. An elk had run out onto a road and caused a traffic accident. Since about forty minutes a dog had been in contact with the elk. The dog, trained for elk-hunting, had broken itself loose. According to Swedish law, a dogowner bears a strict liability for all damage that the animal might cause.12 Invoking this kind of liability, the victim alleged that the dog had induced the elk to run out onto the road and claimed indemnification from the owner. Two causation issues were brought to the fore, firstly: had the dog caused the elk to run out on to the road (causation in fact), secondly: had the owner, as a matter of law, caused the accident through his ownership of the dog (remoteness of damage). Since the Supreme Court, as well as the County Court, replied to the first question in the negative, it never had to takea decision of the second one. The Court of Appeal, however, answered
the first issue in the affirmative.13 Concerning the matter of remoteness of damage, the court agreed that the damage to the car was not a typical consequence of the fact that the dog had broken itself loose. But the court did not stop here. It invoked the travaux préparatoires ofthe law rule regulating strict liability for a dog owner. According tothese one purpose of the liability was to bring home to the owner that a dog should be kept under such control that damage could beprevented. Another aim was that dog owners would cover their liability to a greater extent with an insurance. Against this background the court found it reasonable to assume that the purpose of the legislator had been that even less common kinds of damage caused by dogs should be indemnified. That the dog, running around, had been unattached and had followed the elk was therefore considered to be insuch a causal relationship with the elk's jumping out onto the road that liability was accepted. In short: with the aid of strict liability the court came to the conclusion that the causation link was established.
This way of reasoning gives proof not only of the independent way in which a Swedish court is able to treat tort law questions. It also illustrates how it can bring Swedish tort law more into line with the modern discussion and how strict liability can also be invoked intraditional questions (such as the causation ones) as a weapon to improve the situation of the victim.14
8. Total liberation from the principles of the law of negligence is advantagous to the rules of strict liability. The latter ought to be builtup quite independently15. They can be tailored to the special activity they concern. Questions of who shall bear the risk, how the demand for causation shall be constructed and other details can be decided
within each type of liability. In this way strict liability, in combination with liability insurance, can be an important and flexible instrument for the development of tort law.16 This is clear not least in connection with some of the goals that are set up for tort law. Compensation, loss spreading and general deterrence are goals that from time to time are supposed to be achieved best with the help of strict liability.17
B. Development of tort law through strict liability
9. Strict liability can however be looked upon from another point of view. Does switching over to it always mean that tort law is developed? Is such a step always an advantage? Is it the only alternative? Are we, to use the words of P. S. Atiyah, getting value for money?18
Of course, it is not possible to give a general answer to these questions, valid for all countries at all times. Some views can however be given.
10. Most people take it for granted that strict liability is much easier to handle than negligence law, since it presupposes the absence of the difficult issue of fault. But strict liability also has its problems.
Already the question of how to introduce strict liability involves considerable problems.19 In Western Germany20 and Switzerland21the task of establishing a new kind of strict liability has been principally given to the legislature. Austria22 and France23 are countries where the courts are supposed to be permitted to introduce strict liability; in Sweden this is also the case, although there is nowadays a tendency to
leave this task to the legislature.24 But even if the direction of a certain country in this respect is easy to find, there still might be introductory questions that are difficult to answer, as for example if it is better tosharpen the liability in some more traditional way: by raising the demands for care that one might put upon the tort feasor,25 by invoking principles making the proof of fault easier for the victim (such as the doctrine of res ipsa loquitur), or by working with analogies from other fields where strict liability already exists.26
Another type of problem concerns the more technical aspect of the scope of the liability and the defences against it. What kind of activity should be covered by this kind of liability?27 Should there be any limits to non-pecuniary harm or to the amount of compensation?28Are default of the plaintiff or Act of God recognised as defenses? What will the relationship between strict liability and the negligence law bein a certain field; for example, can the victim refer to the negligence law when this is to his favour? Questions like these show that strict liability is not always so easy to grasp as one might think. The courts might have as many difficulties applying the rules concerning strict liability as applying those of negligence law.29
From a theoretical point of view it is easy to find an obvious weakness of strict liability: its lack of a consistent theory. The rules are, contrary to those concerning the fault principle, split up and difficult to survey.
Much of the discussion around strict liability in Europe has had a tendency to concern only the more technical aspects mentioned above.30 To be able to answer the question if we are getting value formoney (9 supra) it is however necessary to pay attention to other issues, such as economic considerations. Is strict liability effective from an economic point of view? Is this kind of liability better able to influence the behaviour of enterprises? The answer to these questions have not been unanimous. To the former some scholars have answered in the affirmative,31 others no.32 To the latter the pattern is the same33 except that there is also another group that has taken an intermediate position.34 The discussion so far has been very complexand hard for a judge to follow;35 at least in this field the observation that the influence of scholars on the development of tort law is large probably has an exception, at least in Europe.36
11. From the victim's standpoint it might seem as if the introduction of strict liability must always be the best solution. This however need not be the case. There might be other better solutions. These need not only be alternatives to strict liability. They might even come
into conflict with this kind of liability. To illustrate this, it may be suitable to call to mind the development which has taken place in Swedish tort law during the last decade.37
12. In 1972 a general Tort Damages Act came into force in Sweden. The conditions of liability in the Act are based on fault though some general provisions in the Act also apply to strict liability situations. One of the purposes of the law was to improve the victim's right to compensation.
In the preparatory work of the Act the future of tort law was discussed.38 Not only the negligence rule was criticized. Further developments through the introduction of new kinds of strict liability were in principle, considered inconvenient.
It was admitted that strict liability had led to changes in favour of several categories of victims. But many of the arguments that were raised against the negligence rule were considered valid also against strict liability. The main aspects of the criticism were the following.
Due to its connection with a criterion of risk or danger, also strict liability rules meant that the question of the right to compensation was considered in view of how the damage had occurred. From a social point of view, fair and economically rational spreading of the losses could not be achieved through strict liability. Even within the framework of the rules of strict liability, contributory negligence on the part of the victim might mitigate the liability, in principle in the same way as by application of the negligence rule. Strict liability did not guarantee that compensation would really be paid out. It was considered nearly impossible to find general criteria that were acceptable from a social point of view for a demarcation of the fields in which strict liability should apply. Danger was not possible to measure with any degree of exactness. The dividing line between a dangerous and non-dangerous activity was considered arbitrary. The criterion of danger was completely irrelevant if one applied views of social security to problems of compensation. From the point of view of the victimsit made no difference whether the liability arose within an activity that was dangerous or not.
In return, the future was considered to lie in insurance arrange-
ments which worked directly in favour of the victim.39 In a system of this kind, a first party insurance, it was much easier to avoid the limitations on the protection of the victim that followed on a system of strict liability with its demand for an identification of a cause. A strict liability might even prevent or at least make it more difficult to transfer to compensation systems of an insurance character. The criticism that was thus directed against strict liability concerned personal injuries above all, but it was also considered valid on the whole for property damage.
13. The later development of Swedish compensation law demonstrates that the recognition of no-fault insurance40 was a realistic alternative to strict liability. Just after the Tort Damages Act had come into force a compensation system for work injuries — security insurance —was established: employers under took to bear liability regardless of fault through a voluntary insurance scheme.41 In 1975 a new traffic accident compensation scheme was adopted by the legislature; in important parts, as for example concerning personal injuries, it can be considered a no-fault insurance. 1978 the drug insurance scheme came into force: manufacturers and importers of drugs promised to pay for injuries whether or not they had been negligent.42 Other arrangements, partly coming close to the no-fault insurance, have also been developed in Sweden.43
14. Even if the development of Swedish compensation law has thus been much in favour of no-fault insurance, the arguments against strict
liability that were presented in the preparatory work of the Tort Damages Act must be considered untenable. Apprehensions seem to have been exaggerated. As was later pointed out in the Swedish discussion, the argument that strict liability would prevent or make it difficult to introduce a more radical insurance system was contrary to earlier Swedish experiences.44 No-fault insurance and strict liability are not opposed to each other. The one does not exclude the other.
It was also stated in the travaux préparatoires that a new kind of strict liability was not completely ruled out. If it should be considered necessary, strict liability might be initiated; some fields were even mentioned where this might happen. Just after the law had come into force, a commission was also set up to reform the rules of product liability. According to its directives this commission had two alternatives: establishing no-fault insurance or introducing strict liability. The commission chose a no-fault insurance concerning personal injuries due to drugs45 but, in principle,46 strict liability for all other kinds of injuries.47
15. The monotonous arguments against strict liability, as well as those against tort law on the whole which were raised in the preparatory work of the Tort Damages Act, illustrates a tendency in modern compensation law to simplify complex matters that can hardly be simplified, at least not without risk of misunderstanding. Within this trend, obvious not only in legislation but also in the works of scholars,48 all forces are concentrated on one system (in this case a first party insurance) at the same time as others (here tort law) are ruledout. The strive for system and rationalism is driven too hard.49
What one forgets is that both systems, not only the criticized one (here strict liability) but also the one aimed at (here a non-fault scheme), may have weaknesses. The causation issue for example (cf.
5—7 supra) might be hard to decide also in a system of no-fault.50 Even if it on the whole might operate more generously to the victim, than a tort system does, it does have a primitive way of functioning that might be disadvantagous to him.51
One overlooks that both systems may be useful in a society for different kind of damages (see 17—19 infra). One disregards that they might even exist together in the frame of one system of compensation (see 20 infra).
A less rigid and more practical approach, closer to the realities, is needed.52 And this concerns not only the details in question,53 but also the more general discussion about the goals of compensation law.54
C. Strict liability and other systems
16. In 1965, Geneviève Viney published Le declin de la responsabilité civile (Paris). The author made an important analysis of the French tort law and its relationship to the various compensation systems that were built on insurance. Evaluating the role of tort law in the future, Viney proposed a mixed system: "la coexistence d'un système completd'indemnisation collective avec une véritable responsabilité individuelle".55 Viney's proposal had a firm basis. Mixed systems are common today not only in France but also in other countries, not least in Sweden.56
17. The coexistence of strict liability and no-fault insurance systems raises many questions. The first one is not the simplest. If we have tomake a choice between the two — which one ought we to prefer?
Some kinds of injuries or damages are certainly more suitable for insurance arrangements than others. Looking at Swedish experiences in the product liability field, drugs turned out to be a suitable case for no-fault insurance. But efforts to achieve the same result for other products have been made in vain.
The question now discussed has nothing to do with the size of damages. Strict liability might be as good an instrument to handle large-scale damages as the no-fault insurance is. In environmental law, where damage can be considerable, an insurance arrangement has been contemplated in Sweden instead of strict liability, but the pursuit has hitherto been fruitless. An insurance system turned out to be too expensive and too difficult to administer in comparison with strict liability.57 And one should not only look at the amount of the damages. American scholars have found that tort law on the whole might be a useful instrument for dealing with catastrophic accidents.58
18. As already mentioned, a no-fault insurance system was established in Sweden in 1978 to cover injuries due to drugs. About the same time strict liability was introduced in Western Germany for the same kind of injuries.59 Would it not be possible to compare these two compensation systems so as to be able to get an answer to our question?
Some in teresting results may certainly come out of such a comparison. Obvious differences may show us what is bad and what is good. But the details are one thing, the overall picture another. A definite choice depends on facts that are difficult to grasp: underlying evaluations, the will to harmonize the system with other existing systems, institutional conditions and so on. Comparative law has hard work to do in the compensation field.60 It is naive to believe that the choice
between a strict liability and a no-fault scheme is the same in one country as in another.61
19. In Sweden the definite choice in favour of a system of no-fault insurance was taken after having considered the possibility of strict liability. The emphasis was placed on the victim. It was found that it was important that everyone who was entitled to compensation actually received it, even if this meant that some people should receive compensation without being entitled to it.
But, after all, everything depends on how the rules are applied. It is one thing to be generous when discussing the causation issue: is the injury due to a drug? It is quite a different thing to compensate these injuries, when they have been accepted as drug injuries. The Swedish drug insurance scheme has been applied in such a way that one can assert that the chance of receiving compensation is not greater thanabout 40%; only more serious and rare injuries can be compensated.62 Of course it is easy to be generous in the matter of cause, if you are so restrictive in granting compensation when a final decision is made.
This shows that general principles can never be a good basis for deciding which compensation system to adopt. One has to follow up the system in detail and in its practical applications in order to arrive at any sort of conclusion.
20. The question of the relationship between strict liability and no-fault insurance might include not only a choice between the two in the frame of a given type of damage. A much more complex issue arises when both systems are combined in one single compensation arrangement; to use no-fault insurance does not exclude the possibility of using tort principles within the insurance. This model was used in the framework of the Swedish Traffic Damages Act, which came into force in 1976 and which included no-fault insurance. The main issue in this application of tort rules was the amount of compensation. It was considered necessary for the traffic accident victim to receive full compensation according to the tort law principle whereby the compensation awarded shall be of such a magnitude that it will restore him to the same situation in which he would have been if the accidenthad not taken place (restitutio in integrum). But even as regards the
conditions of compensation, for example concerning the causation issue, tort principles have been used. It was stated in the preparatory work of the Act, that the obligation for the insurer to pay compensation to a traffic accident victim corresponds to strict liability in tort.
In this case the stress without doubt was put on the no-fault element of the system. A mixed arrangement might however also have its emphasis on tort law principles.63
21. Of course there are limits also for a mixed system. The legislature must avoid creating too complex a set of rules. From this point of view the Swedish Traffic Accident Act was criticized.64 And the same objection was raised against the result of the "overall strategy" that was proposed in England in 1979 by the Pearson commission.65
In the frame of a no-fault system to refer to a lot of "general" tort law principles may save time for the legislature. But if these are not clear — and is this ever the case? — it nevertheless sooner or later might give rise to a lot of problems for the courts. Also from this point of view the Swedish Act was criticized.66
The English experience shows that there has to be a deeper harmonisation of no-fault and tort compensation. There has to be some kind of principle that underlies the mixed system. "Pragmatism without principle"67 cannot be accepted. Too many compromises may be dangerous.68
22. Strict liability has been and is still one of the most burning issues of tort law. The fight is conducted in three contexts.
The first, classical but still brought to the fore, concerns the relationship to negligence law. Shall strict liability be introduced or not? Closely connected to this question is the issue of whether both kinds of liability follow the same pattern as regards the general conditions, above all the one of causation.
The second one concerns economic considerations; is it or is it not a good way to attain ccrtain economic goals?
The third fight is against non-tort schemes. It has already become "a conventional reform wisdom" 69 that tort should be deserted infavour of a no-fault insurance. But is this always wise?
23. It is difficult to deny the importance of this insurance as well as its role in the future. One thing, however, is to accept this kind of compensation arrangement and perhaps admit that it "will not go away".70 Another is to condemn strict liability all along the line as an effective instrument to achieve an improvement in the victim's situation. This kind of liability also has its advantages and might very wellbe used in the future to develop compensation law. The situation might be such that it is quite impossible to effect a non-tort scheme. Is it then better to refuse to make any change because of a matter of principle than to make this kind of change in favour of the victim? After all half a loaf is better than no bread at all.71
24. Whether the sophistication this approach involves is good or not is another question. Compensation law is extremly complex and will probably continue to be so. It is not excluded that main features with all it has of unjustice and of split will remain, simply because a more radical change will be too arduous and without support of the individuals themselves.
25. Perhaps there has been a tendency in the discussion to simplify conditions that in reality cannot be simplified. The question is not whether to introduce strict liability or a no-fault insurance. Evidently, both may be used at the same time for different types of damages and even for the same type of damage.
So, on the whole, caution is recommended. If one goes in for a certain system all along the line, one might be seriously disappointed by the realities. As happened to the smiling lady of Riga, who thoughts he could dominate the tiger: one might be swallowed by what one thought was mastered.