The reversal of the burden of proof in the Principles of European Tort Law A comparison with Dutch tort law and civil procedure rules

Nobelstraat 2a, 3512 EN Utrecht, the Netherlands; i.giesen@uu.nl). This paper is a revised version of a contribution that was previously published in Dutch in 2007 AV&S, no. 41, pp. 271-276, in a series of articles that compared the PETL with Dutch tort law. Many thanks to Ton Hartlief and Siewert Lindenbergh for their critical comments on an earlier draft of that article. 1 See European Group on Tort Law, Principles of European Tort Law. Text and Commentary, 2005 (this work will be quoted below as: EGTL (author), Article x:xxx, Comment x); J. Spier & O.A. Haazen, ‘The European Group on Tort Law (Tilburg Group) and the European Principles of Tort Law’, 1999 ZEuP, pp. 469 et seq., as well as <www.egtl.org>. Further reading is also listed below. 2 Cf. H. Koziol, ‘Die “Principles of European Tort Law” der “European Group of Tort Law”’, 2004 ZEuP, p. 234.

The reversal of the burden of proof in the Principles of European Tort Law A comparison with Dutch tort law and civil procedure rules Ivo Giesen *

Introduction
In 2005, the European Group on Tort Law (EGTL), in a widely acclaimed effort to contribute to the further harmonisation of the law of tort in Europe, published its Principles of European Tort Law (hereafter: the PETL or Principles). 1 The PETL are meant to serve as common principles of 'European tort law' (if that already exists at all) and as the starting point for the future discussion on the possible harmonisation or even unification of tort law in Europe. 2 Under these Principles, especially in cases of liability based on fault, the burden of proving fault is in some instances 'relaxed' or even 'mitigated' through the acceptance of a reversal of that burden of proof.For one thing, the mere possibility of such a reversal is already somewhat surprising since the PETL are predominantly devoted to substantive law issues; the PETL do, however, also contain a number of specific provisions on the subject of the burden of proof even though this is quite generally regarded as a procedural law topic.Next, the inclusion in the PETL of such burden of proof rules is highly relevant since in practice a reversal of the burden of proof leads to a tightening of liability.Of course, such a rule might (still) be accepted but, given its effect on the substantive outcome of tort cases, this should only be the case if and when such a decision can be normatively justified.Hereafter, I will try to show that these issues as to the burden of proof merit our full attention because the much needed normative justification for the choices made in the PETL is not always sufficiently forthcoming.I will do so by contrasting the Principles with the rules on the burden of proof in Dutch tort law.
In the following contribution I will thus analyse the burden of proof rules in the PETL not only from a more technical point of view, but also from the perspective of the possible influence they might have on the substantive outcome of tort cases.To highlight their content, importance and possible inspirational force for a future European tort law, I will contrast them with their counterparts under Dutch tort law.I will try to answer the question whether the choices made in the PETL are justifiable and whether the Dutch tort system can -or maybe even should -seek inspiration from these PETL.
In doing so, I will first provide the reader with the text of the Principles involved (Section 2), and then further analyse and explain these Principles, including a discussion of the relevant Dutch law in that regard (Sections 3-10).In the concluding section I will determine whether the Dutch tort law system -or any other system for that matter -can 'benefit' or learn from the apportionment of the burden of proof in the PETL (Section 11).In that evaluation I will also provide an assessment of the subject in its entirety from the perspective of whether we are on the right track towards a 'European tort law' if we follow the principles as they now stand (if and to the extent that such a European tort law might indeed be considered desirable).

The text of the PETL as regards the burden of proof
In the PETL two provisions have been devoted to the subject of this paper, i.e. the reversal of the burden of proof in certain cases of liability arising from fault on the part of the person addressed. 3s we will later discern, these rules serve as an exception to the internationally self-evident and (therefore?) non-articulated general rule that the burden of proof in the case of fault is borne by the claimant. 4The provisions laid down in Section 2 of Chapter 4 PETL read as follows: 'Article 4:201 Reversal of the burden of proving fault in general (1) The burden of proving fault may be reversed in the light of the gravity of the danger presented by the activity.
(2) The gravity of the danger is determined according to the seriousness of possible damage in such cases as well as the likelihood that such damage might actually occur.''Article 4:202 Enterprise liability (1) A person pursuing a lasting enterprise for economic or professional purposes who uses auxiliaries or technical equipment is liable for any harm caused by a defect of such enterprise or of its output unless he proves that he has conformed to the required standard of conduct.
(2) "Defect" is any deviation from standards that are reasonably to be expected from the enterprise or from its products or services.'Preceding these provisions, in the chapter on 'damage', we also find a provision on 'proof'.It concerns Article 2:105 PETL which states: 'Article 2:105 Proof of damage Damage must be proved according to normal procedural standards.The court may estimate the extent of the damage where proof of the exact amount would be too difficult or too costly.'Below I will discuss and explain the first two provisions in particular (in Sections 3-9), but subsequently I will also briefly touch upon the third provision quoted above (Section 10). 5

Article 4:201 PETL and what they do not say… but should have said
It is important to stress from the outset in any analysis of the burden of proof rules in the PETL that the allocation of the burden of proof itself is an aspect that is not explicitly included in these Principles, namely a generally worded provision such as 'the burden of proving the requirements regarding liability lies with the party that is relying on this remedy', or a better formulated variation thereof.This would have clarified that also the PETL are based on the general rule regarding the burden of the proof.This general rule (that the claimant must prove that the requirements for liability have been met 6 ) is followed everywhere and is considered self-evident.The possible reversals of this burden of proof, which are discussed in the PETL and in this paper, form exceptions to such a general rule.There is, however, no trace of such a general rule at all in the PETL. 7This implies that such a rule must be deduced from the fact that Article 4:201 does address the reversal of that burden of proof, thus implicitly stating that there is some rule to reverse, to deviate from. 8lthough this general rule is thus taken for granted by the drafters of the PETL, I would advocate that it should be made more explicit after all.Since the PETL do not state from which starting point (i.e., which general rule) one can or must start reasoning when embarking upon (an argument for) a reversal of the burden of proof, we miss out on information that might be useful when studying such an exception to that general rule. 9The question as to how such a general rule on the division of the burden of proof would have been formulated, if dealt with at all, must then be solved by following the obvious explanation, namely that the drafters of the PETL undoubtedly would have had the customary general rule described above in mind had they indeed taken action in this regard.

Reversal of the burden of proof in light of the 'gravity of the danger'
After having cleared this hurdle we can now focus on the envisaged possibilities of reversing the burden of proof.I will then restrict myself to the burden of proof regarding fault, or fout as a Dutch tort lawyer would call it.
Article 4:201 Paragraph 1 PETL opens up the possibility of a reversal if 'the gravity of the danger presented by the activity' gives rise to this.One has to choose one's words carefully here, because the provision also does this.The burden of proof may, as it is stated, be reversed if the danger involved in the activity is sufficiently great.I will return to this later in more detail (in Sections 6-7).
The aim of the provision, according to the Explanatory Comments by the drafters of the PETL, 10 is to build a bridge between the fault-based liability of PETL and strict liability in the sense of Article 5:101 PETL.If, for whatever reason, there is not (yet) any latitude or necessity for strict liability, the burden of proof may still lead to a certain tightening up of liability.This idea is obviously not a new one; the burden of proof is often used to narrow the 'gulf' between fault-based liability and strict liability and to tighten up liability. 11Its incorporation in a general (statutory) rule, however, and not just for certain specific cases, does seem to be exceptional. 12hether this exceptional situation will turn out to be a fortunate choice requires more analysis, however.I will therefore return to (the text of) Article 4:201 PETL and will make three more remarks thereon.

The rationale behind Article 4:201 PETL
First of all, I would like to discuss the rationale behind the rule, the reason(s) for its incorporation in the Principles.This rationale is closely related to the more usual reasons put forward to justify the reversal of the burden of proof, namely that it seeks to improve the position of the plaintiff.It does so in particular by meeting the victim halfway if the application of the general rule regarding the allocation of the burden of proof would result in unreasonable difficulties for that plaintiff due to the technical or organizational complexity of the defendant's activities (and resulting in facts which are difficult to prove due to a lack of clarity). 13The reasons mentioned for accepting the enterprise liability of Article 4:202 PETL could also have been used here (such as the 'beneficiary pays' principle, the channelling of liability or insurance coverage as a result thereof). 14As far as I am concerned, it would have made the foundations for the reversal of the burden of proof much stronger because it would have made legal policy and normative considerations arguments (and not just the mainly factual arguments, which vary depending on the case) decisive. 15As far as I am concerned, normative points of view, and not factual situations, should in the first instance govern the allocation of the burden of proof and the reversal of this burden, if any. 16The reason why I arrive at this conclusion is that the reversal of the burden of proof leads to a tightening up of liability and this must be capable of being normatively justified.

A discretionary reversal of the burden of proof: What is the law?
A second striking aspect is that this reversal of the burden of proof implies that the court is vested with discretionary powers, 17 including all the freedom that this entails.In itself, there is nothing untoward in this as our national laws are riddled with discretionary powers and the principle of reasonableness and fairness mentioned in Article 150 of the Code of Civil Procedure (Wetboek van Burgerlijke Rechtsvordering) enables the Dutch courts to use their discretionary powers to the full.But having said this, there is still a certain degree of friction.It is disturbing that discretionary powers such as these -and the same applies to the reasonableness and fairness exception in Article 150 Code of Civil Procedure -provide so few guidelines regarding the application of the provision in question. 18The court may reverse the burden of proof, but does not have to do so, not even if the required conditions are satisfied. 19But what, then, can the citizen seeking justice expect?What will he be able to rely on beforehand?Will there be a reversal of the burden of proof, and if yes, in which case? 20 Wagner therefore speaks of the 'concert of vagueness' after having listed a number of comparable 'may' provisions found in the PETL. 21Are the criteria mentioned and used in these Articles (more about this later) decisive for their application or not, he wonders.I agree with him. 22It would have been better, in my view, to introduce a rule which provides conditions that are as clearly formulated as possible and which allow the courts to reverse the burden of proof in certain cases -which then also imply that they must do so in those cases.That rule would then resemble a special rule regarding the burden of proof in the sense of Article 150, second sentence Code of Civil Procedure. 23

The conditions for applicability, if known, are difficult to apply
I have argued that these types of discretionary provisions lead to uncertainty as to whether or not the criteria or conditions mentioned for their application are decisive.An additional factor -and that is my third point -is that the 'norm' regarding the 'gravity of the danger presented by the activity' is unclear.Although Paragraph 2 of Article 4:201 lays down a more detailed description of the term 'gravity of the danger', this description once more leads the user of the PETL 24 to a (weighing of a) number of factors, in this case the likelihood that such damage might actually occur and the gravity of such possible damage.If the activity at stake often leads to damage or even serious damage, a reversal of the burden of proof can be an option.This will apply all the more so if there is often damage of a serious kind. 25The latter case is no longer, as far as I am concerned, a situation which is halfway between a normal risk (and thus a regular fault-based liability) or an abnormally high risk (which is subject to strict liability) for which the reversal of the burden of proof has been created, 26 but a case in which strict liability should be seriously considered.After all, if this form of liability is not relevant in such a case (often damage and serious damage), then when will it be?
Apart from this, it is striking that two of the factors that in the Netherlands also determine -through the rules laid down in the so-called Cellar Trapdoor case (Kelderluikarrest) 27 -the existence and scope of the duty of care are operative here at a different level, namely regarding the issue whether the reversal of the burden of proof may be acceptable.The fact that the same concepts (or factors) appear in several places is in keeping with the philosophy behind the PETL 28 and should therefore not come as a surprise. 29This does not alter the fact, however, that such a weighing of factors, whatever they may be, is always tricky -and this case is no different -and rather unpredictable in any case as far as the outcome is concerned.It only leads to an increase in the uncertainty involved in the application of this 'may' provision. 30In my estimation the reception of the PETL will not benefit from this. 31

Article 4:202 (Enterprise liability): The bone of contention
Having said that, it is probably a good idea to discuss the second principle quoted above (Article 4:202 PETL), a Principle which is almost certain to lead to scepticism here and there.This is almost certain since the provision at issue, providing rules regarding enterprise liability, has given rise to controversy even before its introduction.Apart from the fact that this category is more or less still unknown in the legal systems of the world, 32 one of the members of the European Group on Tort Law (EGTL), its founder Jaap Spier, has explicitly stated that he cannot agree with the provision. 33To say that this is an intriguing state of affairs is an understatement. 34he provision lays down, putting it succinctly and without going into detail concerning the definitions of the terms used, 35 that if harm is caused as a result of 'fault' 36 on the part of an enterprise or its output, the person operating the enterprise is liable, unless he proves that he has conformed to the required standard of conduct.It implies a reversal of the burden of proof regarding 'fault', and therefore a reversal as far as the duty of care (which means that of a reasonable person, see Article 4:102 PETL) is concerned. 37The Explanatory Comments once again make it clear, in no uncertain terms, that this rule is not about strict liability (which is also evidenced by its inclusion in Chapter 4 of PETL). 38he reason for the incorporation of this provision appears to have been the possibility of spreading the damage and the 'beneficiary pays' principle, 39 though also and in particular when we look at the Explanatory Comments, the desire or necessity to meet the claimant halfway if evidentiary problems concerning the exact cause of his damage, though being unclear, can be traced back to an activity carried out by the enterprise in question (in one word: channelization). 40ll the victim has to do is to prove that the cause of the damage lies within the sphere of the enterprise due to a 'defect' on the part of that enterprise or its output. 41Such a reversal of the burden of proof (and de facto the tightening of liability) is unknown in present Dutch law, but seems to be defensible 42 although radical and, what is more important, probably superfluous.
The rule can be defended on substantive grounds, based on the arguments just mentioned and because proof to the contrary is indeed possible.But it requires a difficult weighing of legal policy (i.e. policy choices) which will not be readily made in the same manner in the Netherlands.Besides, to my mind this rule is actually superfluous as the use of presumptions of fact in these cases will often lead to the same result, which means that they provide a solution for the most urgent cases of (evidentiary) difficulties. 43My suggestion that this rule is or at least could be superfluous becomes all the more convincing, I believe, if one also takes into consideration that the drafters of the provision sought to restrict its scope (which is due, in all probability, to their disagreement concerning this provision).Liability (of an enterprise) can usually also be based on one's own fault, the fault of a member of one's staff or on the strict liability rule of Article 5:101 PETL, according to the Explanatory Comments. 44Was it, considering all this, indeed wise to retain this bone of contention and source of disagreement?And if so, why then?Was there a strong intention (at least among some of the members of the EGTL) to create something new?And which members of the group were in favour of this?And do they not fear, like Spier, 45 that in this way the liability of enterprises will be unduly extended (for example, by giving the concept of 'fault' a broad interpretation).Unfortunately, a great deal remains unclear in this respect.

A possible, though well-concealed (extra) reason?
One possible (extra) reason to opt for the incorporation of Article 4:202 PETL might be found in the following.The examples that are used to clarify the operation and scope of this Article regularly 46 include the example of something going wrong in a hospital after which the patient (e.g. because he/she was anaesthetised) appears to be unable to prove what has happened.The patient can then rely on Article 4:202 in such a case, or so it is said.The link with the idea of a central liability of a hospital, which is known in the Netherlands from Article 7:462 Dutch Civil Code (Burgerlijk Wetboek; the hospital itself, not only the doctor, is also liable for fault), and which is incorporated in a more restricted version in the 'Principles of European Law on Service Contracts' (PEL SC) by the 'Study Group on a European Civil Code' is then easily made, although the form is slightly different here (a reversal of the burden of proof). 47Is this, a central liability, perhaps what those within the EGTL really wanted?
At first sight, I admit, this does not seem plausible because it would have been much easier in that case to simply write that down in the PETL.On the other hand, such central liability is virtually unknown outside the Netherlands 48 and originates (in many countries at least) from liability under contract law, which is not covered by the PETL.Perhaps this led to the idea that if the drafters wanted to address this issue, it should remain 'hidden'.Further speculations on this subject, because that is what they are, are of course pointless so I will not delve any further in this respect, but if such central liability is all that is left after the smoke of battle surrounding Article 4:202 PETL has cleared, then its incorporation has, after all, been useful in my opinion.Indeed, such provisions are very desirable because it is not just that a victim seeking to recover damages has to stumble over barriers unrelated to the main issue at stake (was there fault?), such as: who did exactly what in a particular organization such as a hospital at a certain moment in time.There may be many reasons for not accepting liability, but this type of obstacle (the sometimes Kafkaesque 'formalities') must be eliminated as far as possible.Even then, enough problems remain, particularly in the case of medical liabilities, which are often difficult to overcome such as the burden of proof regarding fault and the necessary causal connection.

Article 2:105: Proof of the damage
Besides fault and the causal connection, (the extent of) the damage must also be proven.The Dutch counterpart of Article 2:105 PETL is Article 6:97 Dutch Civil Code.If necessary, the damage which a party has suffered may be estimated.As such there are no surprises here, although the conditions under which such an estimate is permitted are slightly different.According to Dutch law, an estimate of the damage is allowed if it is not possible to provide an exact amount; the Principles allow this if proof of the exact amount is too difficult or too costly.Both conditions will often apply simultaneously, though this is not necessarily always the case.If the amount can be determined, although it is costly to do so, the Dutch court will feel more restrained in coming to a justified estimate than the court which applies Article 2:105 PETL.
The fact that an estimate of the damage is permitted means that the evidence regarding the amount of damages is approached rather flexibly.We should not forget, however, that such an estimate effectively comes second according to the PETL rules.The main rule is namely that 'damage must be proven according to normal procedural standards'.This normal procedural standard refers to national legal rules, which means they differ in each country. 49The reason why this rule has nevertheless been incorporated is, firstly, to make it clear that (the remainder of) the law of evidence is 'outside the scope' of the Principles, and, secondly, to provide a uniform arrangement for one aspect (the possibility of an estimate). 50It is praiseworthy that this lastmentioned possibility has been used, but it is still odd to see this sudden reference to 'the normal procedural standard'.It is odd for two reasons.First of all, there is this sudden reference in the PETL to national laws, which incidentally also occurs in Article 5:102 PETL (regarding 'dangerous activities').A reference to national law, however, as a result of rules created by a group without any democratic legitimization remains odd. 51Because which or whose national laws do they actually mean?And is it possible to combine these national laws with the application of the PETL?Does the country's appropriate legislator or court also allow this?Or is that irrelevant?It seems as if the drafters of the PETL were overplaying their hand here and that is regretful because that reflects or could reflect on the often valuable ideas that they wish to diffuse.
The second aspect is closely connected with this.What is it that suddenly prompts them to say something about this procedural aspect, when that does not happen, in principle, anywhere else?Indeed, more than that, why should they do so when, at the same time, it is indicated that procedural aspects are outside the scope of the Principles? 52It does not seem to be consistent.Either incorporate (a number of) procedural aspects, such as some that are closely related to the substantive issues of the case, e.g.rules on the burden of proof 53 or rules on the limitation of actions, or do not do so.Whatever one does, do it completely or do not do it at all, no exceptions made.And that brings us back, as a matter of fact, to the issue with which I began this paper, which is the absence (since it is so obvious) of a general rule on the allocation of the burden of proof.If exceptions are incorporated, the general rule must also be laid down.A single reference to the 'normal procedural standard' does not adequately resolve this deficiency as far as I am concerned.

Evaluation: The PETL on allocating the burden of proof as seen from a Dutch perspective
Are the choices that have been made in the PETL appropriate?Generally speaking, I think that it is recommendable that a new system of tort law takes into account that the substantive law of tort is not all-decisive, and that the formal side of the matter, such as rules on the allocation of the burden of proof, which are generally considered to be procedural in nature, are also essential.After all, to use a cliché, 'being right is not always the same as being proved right, since you will have to prove that you are right'.This applies at the European as well as the national level.In addition to this, the allocation of the burden of proof, and more broadly, the entire law of evidence, is an instrument that can be used to attain the desired solutions, which means that their incorporation in rules on liability (enabling lawyers to use them) can only be applauded.The very fact that attention is being paid to the burden of proof is thus a very fortunate choice.Unfortunately, the details of this sound point of assumption have become somewhat flawed down the road of drafting the Principles, due to a lack of internal consistency.Eventually, it was decided to focus on substantive law only. 54And so there is no general rule dedicated to the allocation of the burden of proof in the Principles, but only exceptions and a sudden reference to 'the national procedural standard' without an adequate explanation of what that actually means.This is of course regretful. 55e that as it may, can the specific rules on the burden of proof as laid down in the PETL still serve as a source of inspiration?The possibility of estimating the extent of the damage in Article 2:105 PETL is no novelty; it will not arouse any enthusiasm, or any opposition.The reversal of the burden of proof in case of a certain degree of danger or 'fault' occurring within an enterprise can definitely be called innovative (and therefore, perhaps, inspiring).Is there anything to gain in this respect?The Austrians think so, because they have envisaged using this Principle for the revision of their tort law. 56 am inclined to say that that is something that is unlikely to happen for Dutch tort law. 57fter all the commotion regarding the so-called 'reversal rule' as regards the burden of proof on causation and the many court rulings that needed to be dedicated to this subject (without really bringing us any further), 58 the Dutch Supreme Court (Hoge Raad) is unlikely to embark on any new adventures concerning the allocation of the burden of proof.And what about the legislator?There is not much to be expected from that side either.First of all because a revision of Dutch tort law is not on the agenda.This is not odd, considering the (relatively) short period that has passed since the introduction of the new Civil Code in 1992, although it would not be amiss to reflect on the subject of tort law, once again, in a fundamental way. 59And secondly, the introduction of Article 150 of the Code of Civil Procedure, with its wide scope, means that there is no need for action on the part of the legislator.If necessary, the law may be amended (by the courts, it is true) within the framework of the existing Dutch legislation.
Apart from the issue of who might opt for the PETL in this regard, is it perhaps worthwhile from a more substantive perspective of the Principles to move in this direction?I have previously indicated that I think that the conditions for a reversal of the burden of proof in case of a 'certain gravity of danger' are still lacking in clarity and its possible foundation is poor.Moreover, there are too many uncertainties surrounding this Principle for it to serve as a guiding example.I consider the Principle on enterprise liability (as stated above) to be radical, though defensible, but, in all probability, particularly superfluous in addition to the possibility to work with presumptions.Therefore, I cannot wholeheartedly say that the choices made in the PETL do constitute the right way forward towards a 'European tort law', if so desired. 60ll that is then left as a possible source of inspiration that can be used for the further development of 'the' European tort law is the idea -which was unfortunately not followed in the PETL -to lay down consistent rules on the allocation of the burden of proof as part of the (newly developed) system of substantive rules on European tort law, including the general rule as well as the exceptions to that general rule.This line of approach would not only be justified by the (practical) importance of those rules on evidence; it would also have made the entire system easier to apply in practice.And that is something that all tort lawyers, in the end, will want to achieve if they want to do something useful when they are thinking about, working on, and developing principles of European tort law.