The Criterion of Causation in Swedish Tort Law
and the European Harmonisation Process
Mårten Schultz[1]
Introduction
In similarity with most other legal systems, a general requirement for liability for damages to apply under Swedish law is that the act or omission for which the defendant is blamed, is causally connected with the harm or loss in question. No matter how severe the negligence is, or how extensive the loss is, there is no liability for the defendant if the plaintiff does not establish causal connection – causation – between the negligent act and the loss. More specifically, it can be said that the requirement of causation, as it is understood in Swedish law, presupposes a distinction. Partly the plaintiff needs to show a factual causal connection and partly that the factual causal connection should be regarded – from a legal point of view - as being of such degree of foreseeability that it is reasonable to use it as the basis of liability. The first requirement is often determined with the help of the ephemeral theory of conditions – a term which is perhaps more well-known among legal scholars than among practicing lawyers - whereas the other requirement is usually examined within the framework of the generally known theory or doctrine of adequacy.[2] The following presentation deals only the first aspect, i.e. the determination of factual causal connection, and the ambition is only to point to a few factors deserving attention, mainly in light of the on-going harmonisation process, or the Europeanisation as Lando and others have dubbed it, of European private law.[3]
Causation in Swedish Tort Law and The European Dimension
In Sweden the requirement of causal connection does not follow from any provisions of the Swedish Tort Liability Act, but is generally said to belong to the unwritten general and fundamental tort law principles.[4] The absence of statutory support is perhaps not so surprising when we realise that the Tort Liability Act is a skeleton act, and that the Swedish legislator never intended to codify tort liability law in its entirety. As regards the question of causation the legislator’s silence on the matter is perhaps especially understandable, taking into consideration the complexity of the causation discussion and elusive character of the concept of causation.[5]
If general Swedish tort law legislation is silent on the question of causation, a number of continental private law codifications are more explicit on the matter. This fact must be kept in mind in the discussion on the on-going efforts to harmonise legislation in the field of private law in Europe, for instance as expressed in several projects led by the most prominent European academics. The aim of these projects is to work out detailed proposals for the codification of private law, including tort law. One of the ideas behind these projects is that the proposals could be used as a basis for the legislative process when the time is considered ripe for further harmonisation of European private law. The most comprehensive of these projects – well-known to the readers of this review – is the Study Group on a European Civil Code, with professor Christian von Bar in Osnabrück in Germany as chairman.
The requirement of causation is one of the most important questions that will need to be accounted for in the Study Group’s proposals for the codification of tort liability law and an issue which will certainly become the object of heated discussion and fierce debate among European tort law academics. From a Swedish point of view it is of great importance to reflect upon how the formulation of the requirement of causation in a proposal to tort law legislation at European level will influence and cohere with the Swedish law of damages in general – irrespective of whether the formulation is worked out by the Study Group or by the EU Commission. Unfortunately, Swedish law does not seem to be well equipped for such evaluation.
Causation analysis in Swedish jurisprudential literature
As mentioned, the Swedish Tort Liability Act lacks a general rule on causation. The absence of a clear rule on causation would not pose a problem if the concept of causation itself was thoroughly explored and well-developed within the law: It does not really matter if the fundamental principles or rules remain uncodified, as long as they are familiar and generally recognised in legal sources other than legislation. However, this is unfortunately not the case in Sweden. Not only has the criterion of causation not been manifested in the form of any codified rule. Worse is that there is no agreement as to how the concept should be understood in Swedish law, neither does there exist a generally accepted theory or doctrine of causation.
The jurisprudential studies made in the area have concentrated primarily on the instruments of evaluation, such as the theory of adequacy and its competitors, which are employed after the proper determination of factual causation has been made.[6] The exceptions to the general characteristic, the few comprehensive inquires on causation that have been published, excellent as they are, (Peczenik’s ‘Causes and Damages’ should be especially mentioned in this context) seem to be considered too theoretical and difficult to follow to be of real interest to the practicing lawyer.[7]
In the more general works, such as Hellner’s and Johansson’s text book, which must regarded as the standard work on tort law in Sweden, the concept of causation is indeed examined, but the presentation is restricted to pointing out different possible lines of reasoning, without taking any stance supporting any specific way of approach.[8] A reader interested in more robust guidelines on how to deal with causal problems searches in vain. And if one reads the recently published commentary on the Tort Liability Act one finds insightful comments concerning the determination of adequacy, but as regards the factual determination of causation the commentary is restricted to just a few lines which address primarily the question of evidence.[9]
The above-shown presentation does by no means pretend to give a complete picture of how the causation criterion is perceived in the jurisprudential literature. However, the works, which must be regarded as the most important contributions in the area today, can be seen as an indication of the fact that no uniform and generally accepted causation theory exists in Swedish law, and that the causation concept itself is rather diffuse. There is a common consensus that a factual causal connection must exist between the negligent act and the loss if liability for compensation is to apply, but the exact understanding of this criterion is shrouded in mystery.
Lundstedt and the scepticism towards causation theories
The general view in Swedish law seems to be that the concept of causation is so elusive and complicated that it is not worth the trouble to burrow into the theoretical and philosophical problems and issues which always arise on closer inspection. A number of authors – ‘authors’ because it is mainly within jurisprudence that the subject gets any explicit attention at all – have chosen instead to mention in passing the arguments made relevant by the “theory of conditions”, without any thorough analysis, before proceeding to the more concrete aspects of liability analysis. To generalise, one can say that the Swedish attitude seems to be that the increasingly sophisticated approaches to analyse causation that have dominated the international discussion – such as those by Tony Honoré, Jane Stapleton or Richard W. Wright to mention but a few names – are considered too theoretical and unworldly to be of any use to at all Swedish law. It is felt that issues of causation can be decided instead in a more pragmatic way by means of determination of reasonableness ‘from case to case’.
This negative attitude towards theories, affecting perhaps primarily theories of a more philosophical character, has been aptly characterised by Jan Kleineman as the anti-intellectual attitude in Swedish tort liability law.[10] In the area of causation the chief supporter of this view is Vilhelm Lundstedt, the great theoretician of tort liability law and private law in the first half of the 20th Century, and a prominent supporter of the Uppsala school of thought. Lundstedt’s attitude towards the need and merits of theoretical analysis of the concept of causation was decidedly negative. In his withering broadside against the Scandinavian tort theory, Kritik av nordiska skadeståndsläror[11] (Criticism of Nordic tort doctrines) Lundstedt unleashed a volley of angry strictures on what he called the doctrine of causation: ‘One may on the whole say that an account of the causal doctrines in a legal work is lessmotiva ted than an account of, for instance, anatomy, surgery, medicine or the political economic rules on price determination, a presentation on seed germination, a technical presentation on how a certain machine ought to be constructed, or of professional skills necessary for horse-breeding, agriculture and stock-farming, etc.’[12] We have not yet encountered any studies on seed germination in tort liability law publications, but neither has much interest been displayed in the causation theories. Lundstedt’s scepticism towards theoretical analysis of the concept of causation fell into good ground in Swedish jurisprudence. The effects of this absence of a more sophisticated understanding of causation is made clear not least in the Europeanisation context.
The need for theoretical analysis of causation
The sceptical attitude towards causation theories in Sweden entails a risk for the European codification process, where encounters with other legal systems and traditions make it necessary to compromise with ones prevalent understanding of the law. How many Swedish jurists can meaningfully participate in a discussion with, say, English or Dutch jurists, with the causation theory at their finger tips, as to whether the shortcomings of the condition sine qua non-doctrine may be overcome or not? A Swedish jurist will be hard put to even start to describe the general view of the concept of causation as it is perceived in Sweden. The point I am trying to make is that if the meaning of the causation criterion is not defined or elucidated in Swedish law, then we will not even know what it means to lose it, or what it would mean for Swedish tort liability law in general if a new codification of the causation criterion were to be externally formulated. The necessity of self-awareness is amplified in the context of European co-operation.
It is, however, not only from the point of view of European harmonisation that a more profound insight into causation theories may fulfil an important function. If we look at the issue of the need for analysis from a more traditional perspective, it appears that even applied legal studies would often benefit from more sophisticated causation theories, presuming they are simple, follow the precepts of pragmatism, that they are properly substantiated and therefore tenable from both a legal-dogmatic and philosophical point of view.[13]
An obvious point of departure, which is sometimes forgotten in more advanced theoretical studies of causation, is that in the majority of cases no theoretical models are necessary for the determination of the causal connection. If Johansson hits Andersson with a hammer on the head, and Andersson’s frontal bone is immediately thereafter found fractured, and no other factors come into play, the judge will hardly have need for any sophisticated models of analysis: The causal connection is obvious in this case, and can be answered intuitively. It is in the more complicated cases that the need for theoretical causation models arises, in which cases these models could be a welcome aid for the judge in making a decision. Examples of situations in which complicated problems of causation may come into play (and have done so in practice) are cases of environmental damage (an example is the case of the poisoned fish in NJA 1981, p. 622) or injuries caused by medication (as in the so-called LEO-case, NJA 1982, p. 421). Being open to causation theories as a useful instrument for the difficult cases does not mean that one should too much faith in their ability to solve practical problems of tort liability. Theories per se can never give any answers to difficult questions. The only thing they can do is to provide a structure for the formulation of arguments – a methodology – and to create a framework for the way in which ‘correct’ questions can be asked. This alone can be of sufficient importance in a complicated case.
The theory of conditions and its history
Swedish legal scholars may be sceptical towards theoretical models for the determination of causation, but it would not be fair to say that the question has not met with any interest at all. On the contrary, one can find plenty of references to the so called “theory of conditions” (in Swedish: betingelseläran). This term includes several modes of reasoning, which have one thing in common: they are all based on the conceptual pair of necessary and sufficient conditions, which are used for the purpose of causal analysis. The history of the theory of conditions as it is used in the Swedish literature is obscure. There are some indications that the theory of conditions has its root already in Roman law. A prominent specialist in Roman law, Reinhard Zimmerman, claims, on the other hand, that the Romans did not use anything like the modern causation theory, relying instead on commonsense argumentation.[14] In Anglo-American law one refers often to John Stuart Mill (who was influenced by David Hume) as the father of the theory of conditions, whereas Swedish authors refer often to German jurists in addition to Mill. The term ‘theory of conditions’ as such should be considered as a continental construction, probably attributable to the German theoreticians.[15]