Unfair Exemption Clauses An Economic Analysis* 


by professor HEIN KÖTZ


Economics is not my field. All I can boast of is that, as a lawyer and interested outsider, I have for a number of years followed the development of a new approach to the law that is now commonly called "economic analysis of law". It is an approach based on the conviction that it makes sense to ask whether legal rules can be better understood and explained, and perhaps improved, if one assumes that they are, orought to be, designed to promote the economic goal of efficiency in terms of improving the allocation of scarce resources. Economic utilitarianism underlying this approach is nowadays criticized by many as, in the words of Dworkin1, "unjust in its consequences, because it perpetuates poverty as a means to efficiency, and deficient in its theory of human nature, because it sees individuals as self-interestedatoms of society, rather than as inherently social beings whose sense of community is an essential part of their sense of self'.
    This may well be so. Unfortunately, however, legal theory is not my field either, and I hope therefore that you will suffer for a few minutes the more pedestrian view I would like to adapt, i.e., that the best test of a theory is its utility in predicting or explaining reality. In order tolay a basis for a discussion on this level I would like to demonstratethe modus operandi and perhaps the usefulness of the new approach by way of a practical example.
    1. With the rise of the consumer movement in the early seventies most European countries enacted legislation conferring upon the court a power to invalidate contract terms found to be "unfair","unconscionable", "inequitable" or "unreasonable". S. 36 of the


* This is the text of a talk I had the honour and pleasure of presenting to a seminar at the Uppsala Law Faculty on March 23, 1987. I wish to express my appreciation to Professors Anders Agell and Torgny Håstad for their kindness in arranging the talk. Footnotes have been kept to an absolute minimum, and I have not attempted to convert a talk into an article.

1 Dworkin Taking Rights Seriously (1977) X.


474 Hein KötzSwedish Contracts Act as amended in 1976 is a typical example. It provides that the judge may set aside or modify a contract term if he finds it unfair having regard to the contents of the contract, thecircumstances at the time of making the contract, later events and other circumstances. Similar rules have been enacted in Denmark, Norway, Austria, West Germany, the United Kingdom and the United States of America. Some of these statutes are applicable only where the unfair contract term forms part of a standard-form contract. Evenwhere this is not so, experience shows that most, if not all, terms foundto be unreasonable by the courts were "standardized", in the sense that they formed part of a ready-made "pre-fabricated" contract form drawn up by one party for a large number of contracts. There are two questions I would like to analyze from the economic point of view:
    Firstly, Is there an economic justification for legislative measures of the type described?
    Secondly, Is it possible to develop economic criteria for determining whether a given standard-form clause is "unconscionable" or "unreasonable"?
    2. As to the first question, there is no doubt that the justification for the statutes which are of interest here is generally seen in the inequality of bargaining power of the parties. According to traditional wisdom, standard contracts are typically used by enterprises with strong bargaining power. The weaker party, in need of the goods or services, is frequently not in a position to shop around for better terms, either because the author of the standard contract has a monopoly or because all competitors use the same standard terms. For this reason, so the argument runs, the party's contractual intention is but a subjection, more or less voluntary, to terms dictated by the stronger party2,and the overall result is that standard contracts, in the words of Kessler, "could thus become effective instruments in the hands of powerful industrial and commercial overlords enabling them to impose a new feudal order of their own making upon a vast host of vassals"3.
    There may indeed be little doubt that standardized contracts are omnipresent today in the field both of consumer and business transactions, and it is hardly an overstatement to say, as one writer has put it, that "agreements in which the terms are actually negotiated by the parties now belong to the past"4. On the other hand, if inequality of


2 Kessler, Contracts of Adhesion, Some Thoughts about Freedom of Contract, 43 ColL. Rev. 629 (1943) at p. 632.

3 Kessler (preceding note), at p. 640. See also, in a similar vein, Lord Denning's statement in George Mitchell v. Finney Lock Seeds [1983] Q.B. 284 (C.A.), at p. 297.

4 Deutch, Unfair Contracts, The Doctrine of Unconscionabilitv (1977) 1.


Unfair Exemption Clauses 475bargaining power were really the sole raison d'être for standardized contracts, such inequality would have to be equally omnipresent in all areas of economic activity, and that is a statement I find hard to accept. It would also follow from the traditional explanation that contract terms proposed by one party would always be cut down to size by individual negotiation provided that the proponent operates ina highly competitive market or the bargaining power of both sides is roughly equal on other grounds. The available evidence points the other way. There are many cases in which clearly unfair standard terms have been accepted by a party whose bargaining power was equal or even superior to that of the other side, and there is no doubt that standardized contracts dominate even in branches of industry where there exists fierce competition. Even in these situations, pre fabricated terms are accepted without discussion, not because a powerful capitalist has forced them down the throat of an unwary or helpless customer but because it simply does not make sense, economically speaking, for the customer to waste time and energy on a process of negotiation or of shopping around for better terms. If somebody parks his car or has his clothes dry-cleaned, he will of course neither discuss the other side's exemption clause nor shop around for better terms. But this is so, not because the other side is a "powerful industrial or commercial overlord", but simply because the cost of shifting the risk back to the other side is much greater than the potential harm multiplied by the probability that harm will occur at all. I do not contest that standard-form contracts are often proposed by parties whose market position has monopolistic characteristics. But I think that the basic reason why individual negotiating over standard terms is so rare is that the "transaction costs", i.e. the costs of shopping around for better terms and of custom-tailoring each transaction are prohibitive.
    What, then, is the economic justification for this type of legislation? In general, of course, economists favour freedom of contract since contract is the most important mechanism by which scarce resources can be moved to what are considered the most valuable uses. But economists have always conceded that freedom of contract can only lead to the desired result of maximizing the net satisfactions in a given society if certain conditions are satisfied. One of these conditions is that the contracting parties are rational in the sense that they know what they are doing and what best serves their interest. For this reason there is a clear economic justification for a rule under which contracts are void if made by minors or if induced by fraud, undue influence or duress. Another condition is the absence of monopolies or


476 Hein Kötzoverwhelming bargaining power, and it is for this reason that economists have no objections in principle against statutes seeking to outlaw agreements in restraint of trade, to lower barriers to market entry, or to subject mergers to some sort of governmental control. A further condition is the absence of prohibitive transaction costs. Free contractual exchange can lead to the desired results only where negotiations are, economically speaking, feasible. Where this is not so a party will swallow almost anything offered by the other side, not because the other side is a monopolist but because the cost of negotiating exceeds the benefits to be obtained through the negotiating process. This is exactly the situation of a party confronted with a comprehensive set of ready-made standard-form clauses proposed by the other side. Ofcourse, a statute that allows judicial control of such clauses interferes with freedom of contract. Yet this is economically justifiable because what the statute seeks to do is to correct market failure occasioned by prohibitive transaction costs5.
    3. From this analysis we learn how to proceed to answer the second question, in other words, to identify criteria for invalidating exemption clauses which are more precise and more specific than the well meant, if trivial formulae used by the legislature, such as "unconscionable", "unreasonable", "unfair" and so on. We must ask the question: What rule would the parties have agreed upon had the transaction cost been zero? In other words: How would the parties have allocated the contractual risk in question if they had lived in an ideal world in which all the information required for a rational solution ofthis problem had been available to them at no cost? If we have ananswer to this hypothetical question we then compare it to the contract term in the real world. If there is a difference we call the term "unconscionable", if only to please the lawyers, and strike it down as invalid.
    Let us speculate for a moment how parties would allocate contractual risks in the ideal world. Let us assume that a buyer and a seller are about to enter into a contract for the sale of an explosive chemicalat a tentative price of 1000. There is one point on which there is not yet an agreement, and that is the question who should bear the risk of an accidental explosion of the chemical. In an ideal world each party would determine with great care what the expense of avoiding the risk


5 Cf. the more detailed analysis of this problem by Adams, Ökonomische Analyse des Gesetzes zur Regelung des Rechts der Allgemeinen Geschäftsbedingungen, in: Ansprüche, Eigentums- und Verfügungsrechte, Schriften des Vereins fur Socialpolitik (Hrsg. Heumann 1983) 655 ff.; Schäfer/Ott, Lehrbuch der ökonomischen Analyse des Zivilrechts (1986) 323 ff.; Behrens, Die ökonomischen Grundlagen des Rechts (1986) 155ff., in particular 170—172. 

Unfair Exemption Clauses 477of explosion would be. Suppose that the seller came to the conclusion that it would cost him 100 to take safety measures that eliminate the risk. In that event he would be willing to bear the risk if the buyer agreed to pay 1100. This offer would be accepted by the buyer if the safety measures he might take to eliminate the risk cost him more than 100. If the cost to him is only 80 he would not accept the offer. He would assume the risk himself and pay only 1000. True, the overall cost to him would be 1080. But he would still be better off than if the risk had been borne by the seller since in that case the contract price would have been 1100. It is remarkable that this method of allocating the risk does not only maximize the private advantage of the parties but serves also the overall interest of society. From a general point of view, we must be interested in assigning the risk to the party who is able to eliminate it at lower expense than the other party. On a first level, we therefore reach the conclusion that a contractual risk of a certain description will in the ideal world be assumed by the party who is able, at lower cost than the other party, to prevent it from materializing.
    We must now refine this analysis a little further. You will remember that it was in my example at a cost of 80 that the buyer was able to eliminate the explosion risk. This investment makes sense only if the expected cost of the risky event amounts to more than 80. What happens if this cost is no more than 20? Suppose that in the event of anexplosion damages of 20000 will result, and that the probability of an explosion is 1:1000. In this case the cost of the risk amounts to 20000 multiplied by 1/1000 which is equal to 20. In this situation, the buyer would not invest 80 in safety measures because no reasonable person would sacrifice 80 in order to save 20. In short, there may exist risks that will not be avoided because the avoidance cost exceeds the avoidance benefits. A similar situation arises where avoidance is not only very costly but totally impossible. In both situations the parties are facing a risk that neither party will prevent because risk prevention would either be wasteful or impossible. How will these risks be allocated in an ideal world? To find the answer it is necessary to understand the fundamental concept of risk aversion. Compare a 100 per cent chance of having to pay 10 with a 1 per cent chance of having to pay 1000. Although the expected cost is the same in both cases most people will prefer the first alternative. They will prefer a certain cost of 10 to the risk of having to pay 1000 if the probability that the risk will materialize is 1/100. In other words, they will insure, and the omnipresence of insurance is powerful evidence that risk aversion is extremely common. Now let us go back to our case. The parties are


35—37-166 Svensk Juristtidning


478 Hein Kötzfacing a risk that neither party is willing or able to prevent. In this case both parties will want to insure. Therefore, each party will determine what his expense of insuring against the risk would be. In the final result, the risk will be assumed by the party who is able, at lower cost than the other party, to buy insurance coverage. The provisional conclusion is therefore this: In an ideal world, contractual risks will be assigned to the party who is able, at lower cost than the other party, to eliminate the risk by taking preventive measures. If no such measures will be taken either because the risk is unavoidable or the avoidance cost exceeds the avoidance benefits the risk will be accepted by the party who is able, at lower cost than the other party, to cover it by insurance.
    4. The proof of the pudding lies in the eating, and the true test of a theory lies in its utility in predicting or explaining reality. The reality I propose to use as a testing ground for the theory are two cases decided by the German Federal Court of Justice. In both cases a shipowner had agreed to carry a cargo to a certain port. In both cases the shipper sued the shipowner for damages to the cargo caused by the negligence of the shipowner or his crew. In both cases the shipowner based his defence on an exemption clause. In both cases the decision depended on the validity of the clause.
    In the first case sheet metal had been damaged because the hatch covers had been leaky, and seawater had during the voyage penetrated into the hold. In other words: the ship had at the time when it was placed at the shipper's disposal been unfit for the transportation of sheet metal. I have no doubt that the risk of the ship's initial lack of fitness would in an ideal world have been assumed by the shipowner. It is he who is familiar with his ship's condition and with the risks to which a cargo of sheet metal is exposed if the hatch covers are not watertight. He also knows the various safety measures to protect the cargo, and he can easily select the least expensive measures. The shipper, on the other hand, knows nothing about the ship's condition. Even if he obtained the relevant information at considerable cost he would not be in a position to take the appropriate measures since he has no power to accomplish anything on board another person's ship. For this reason it is clearly the shipowner who is able, at lower cost than the shipper, to eliminate the risk of the ship's unfitness to carry the agreed cargo. It follows that this risk would in the ideal world have been assumed by the shipowner. The exemption clause, by assigning the risk to the shipper, deviates from the risk allocation that the parties would have agreed upon had the transaction cost been zero.


Unfair Exemption Clauses 479Accordingly, the exemption clause must be invalidated, and this is indeed the result reached by the German Federal Court of Justice6.
    In the second case a cargo of salt had been damaged because owing to the crew's negligence the ship had during its voyage collided with a pier, and water had trickled into the hold. How would the risk of anegligent failure by the crew to use reasonable care in handling the ship during its voyage be allocated in the ideal world? The shipper can of course do nothing to eliminate this risk. How about the shipowner? After all, he is the employer of his crew. He is in a position to select the most competent people, to train them and fire them if they appear to be incompetent. On the other hand, we must not overestimate the shipowner's possibilities to eliminate blunders and mistakes of his crew. In selecting the captain, the officers and the crew a shipowner will of course ensure that they hold the necessary certificates, have stayed clear of the criminal law, have done a satisfactory job in their prior employments, and are not too fond of the bottle. He will do all this anyway if only because he wants to protect his ship (rather than the cargo). What else can a shipowner do to weed out applicants who are more likely than others to commit the occasional mistake? Such mistakes are, as we all know, inevitable, and this means, unavoidable. Arguably, no clear decision can therefore be made on the first level because the risk is more or less unavoidable for both parties.
    We must therefore move to the second level and ask, Which party is able, at lower cost than the other party, to provide insurance coverage? It would seem that this risk is more easily and more cheaply insured by the shipper. He has full information on the cargo, on its value and on the consequential damages that might result from its loss. He is therefore in a better position than the shipowner to buy custom-tailored transport insurance protection. True, the shipowner might buy liability insurence. But this is likely to be more expensive. Fixing the maximum coverage of the policy would be difficult since the shipowner knows little about the value of the cargo and about his potential liability. The liability insurer's duty to pay depends on the liability of the insured. This is as a rule more costly to determine than loss or damage to the cargo which suffices to trigger a transportinsurer's duty to pay. Liability insurance would not cover loss or damage to the cargo if caused by force majeure or act of God. Since this risk would have to be covered by the shipper, wasteful double insurance might follow. There is therefore evidence to show that it is


6 Entscheidungen des Bundesgerichtshofs in Zivilsachen vol. 49, 356; cf. also Bundesgerichtshof, Neue Juristische Wochenschrift 1973, 1878. 

480 Hein Kötzless costly for the shipper to provide insurance coverage. The parties would therefore in an ideal world have agreed to shift the risk to the shipper. Since this is exactly what the exemption clause amounts to itshould be allowed to stand, and the shipper's damages action should be dismissed. This is the result reached by the Federal Court of Justice7.


7 Bundesgerichtshof, Neue Juristische Wochenschrift 1973, 2107.