Classifying precontractual liability: a comparative analysis

Vol21 No 2 June 2001

Legal Studies Classifying precontractual liability: a comparative analysis Joachim Dietrich LLB (QLD) (Hons), PhD (ANU)* Senior Lecturer, Faculty of Law, Australian National University

The common law has solved questions of liability arising in the context of precontractual negotiations by resort to a range of different doctrines and approaches, adopting in effect 'piecemeal' solutions to questions of precontractual liability. Consequently, debate has arisen as to how best to classify or categorise claims for precontractual work and as to which doctrines are best suited to solving problems arising from anticipated contracts. The purpose of this article is to consider this question of how best to classify (cases of) precontractual liability. The initial focus will be on the ongoing debate as to whether principles of contract law or principles of unjust enrichment can better solve problems of precontractual liability. I will be suggesting that unjust enrichment theory offers little by way of explanation of cases of precontractual liability and, indeed, draws on principles of contract law in determining questions of liability for precontractual services rendered, though it does so byformulating those principles under different guises. Irrespective, however, of the doctrines utilised by the common law to impose liability, it is possible to identify a number of common elements unijjing all cases of precontractual liability. In identifying such common elements of liability, it is necessary to draw on principles of both contract and tort law. How, then, should cases ofprecontractual liability best be classified.? A consideration of the issue of classification of precontractual liability from a perspective of German civil law will demonstrate that a better understanding of cases of precontractual liability will be gained by classifying such cases as lying between the existing categories of contract and tort.

* This article was written during a research semester spent, for the most part, utilising the facilities of the Rechtswissenschaftliche Fakultat of the University of Vienna, Austria. I wish to thank Professor H Koziol of the Institut fur Zivilrecht for his warm welcome to me, as well as his kindness and assistance throughout my period of stay. In particular,

154 Legal Studies 1 INTRODUCTION In a world of increasingly complex transactions, parties to contractual negotiations may carry out work (or otherwise act to their detriment)’ in anticipation of a contract which never eventuates. In some circumstances, such work performed by a plaintiff- gives rise to legal liability (‘precontractual’ liability), often in the form of an obligation upon the defendant to pay for the reasonable value of such work. The problem is a familiar one: typically, parties are in advanced stages of negotiating a contract and the plaintiff (or more usually, both parties)’ believes an agreement will be finalised in the future. The plaintiff‘s expectation that the contract will be concluded ‘in due c o u r ~ eor ’ ~in the ordinary course of negotiations5 is the result of the defendant’s conduct or the circumstances surrounding the negotiation process. The plaintiff expends time, skill, money, materials and the like in anticipation of the contract eventuating, and suffers detriment when no contract is concluded. Although the subject matter 4 this article mostly concerns such ‘typical’ cases of precontractual liability arising from failed precontractual negotiations,6 it is suggested that the comments made and conclusions drawn have equal application to incomplete contracts generally, including where at least one party believes an agreement has been reached but no sufficiently cerruin agreement is discernible.’ It is the lack of an objectively discernible, complete and reasonably certain agreement

I wish to thank Professor Koziol for his careful reading of an earlier draft of this article, for his helpful suggestions and for clarifying my numerous misunderstandings of Austrian and German law. Thanks also go to Professor P Schlechtriem of the University of Freiburg, Germany, for his helpful comments. as well as to the anonymous reviewers of the article for their suggestions. Finally, special thanks go to Birgit Hofer for her assistance in translating German texts. Any remaining errors are, of course, my own. 1. In exceptional circumstances, a plaintiff may not have suffered anyPnancial detriment as a result of relying on a precontractual arrangement but, none the less, a defendant may be held liable merely for having obtained an advantage as a result of the breach of a precontractual arrangement: see further below, s 4(b). 2. For convenience, I have continued to use the terms ‘plaintiffk’ to refer to claimant/s throughout this article. 3. See eg Brewer Street Investments Ltd v Barclciys Woollen Co Ltd [ 19541 1 QB 428 at 429. 4. TurrifSConstruction Ltd regalia Knitting Mills Ltd [ 197212 Lloyd’s Rep 234 at 259. 5. In order for precontractual liability to arise, arguably the plaintiffs expectation need not necessarily be that the contract is an absolute certainty. See C F Cauchi ‘The Protection of the Reliance Interest and Anticipated Contracts Which Fail To Materialise’ (1981) 19 UWOLR 237,268, formulating the requirement that there needs to be merely a ‘legitimate expectation’ of a contract. 6. Included herein is the situation where the parties have reached agreement on all essential matters but the parties still require that certain formal steps be taken before they are to be legally bound. 7. Eg Stinchcombe 1’ Thomas [ I9571 VR 509. A recent example is Easar Antennas Ltd v Racal Defence Electronics Ltd (7 September 2000, unreported, HC, Hart J , www.law.cam.ac.uk/restitution/archive/englcases/easat.htm). Supportingthe nexus between uncertain and anticipated contracts,justifying their common consideration under the rubric of ‘incompleteness’, see J W Carter and D J Harland Contract Law in Australia (Sydney: Butterworths, 3rd edn, 1996) p 73, and William Lace? (Hounslow) Ltd v Davis [ 19571 2 All ER 712 at 719.

Classifying precontractual liability: a comparative analysis 155 which explains why the plaintiff‘s efforts ultimately prove to be fruitless and why no contructud obligations can be enforced in relation t h e r e t ~ . ~ Of course, work done in anticipation of an incomplete contract usually will not give rise to liability of any kind.” It has been suggested, however, that the ‘courts have shown an increasing willingness to impose precontractual liability’” and those decisions which have imposed such liability are largely uncontroversial in their results. McKendrick has pointed out that the ‘difficulty with the topic lies, not so much in the results which the courts reach, but in the reasoning which gets them to that conclusion’.I* Part of this difficulty stems from the fact that a variety of rules and doctrines have been utilised by the courts to impose precontractual liability, including resort to contract law itself, such as by enforcing an express or implied ‘ancillary’ contract in relation to work done in anticipation of an incomplete ‘main’ contract;I3 resort to equitable principles such as estoppel14 and the law of constructive trusts;15 resort to rules within the law of quasi-contract, restitution, or unjust enrichment (whatever it should now be called); or, perhaps more rarely, resort to tort law.I6Consequently, debate has arisen as to how best to classify or categorise claims for precontractual work and as to which doctrines are best suited to solving problems arising from anticipated contracts.

8. See, however, further below, section 3(a). 9. It is possible to draw a distinction between incomplete contracts and defective contracts. A defective contract can be defined as one in which the parties have reached what is a complete and certain agreement, but they have failed to comply with a statutory or common law rule which renders ‘defective’ what would otherwise be an enforceable contract. Although some of the discussion herein applies equally to defective contracts (and, in any case, the distinction between incomplete and defective contracts can not be sharply drawn and a ‘contract’ may be ineffective on both grounds, as eg in Deacon v Adams (1982) 55 NSR (2d) 218), it must be noted that other considerationsare relevant to determining liability under defective contracts, eg thepolicy of the statute rendering a contract unenforceable. See generally J Dietrich Restitution: A New Perspective (Sydney: Federation Press, 1998) ch 6, and specifically pp 107-122. 10. Cf E McKendrick ‘Workdone in Anticipation of a Contract which does not Materialise’ in W R Cornish, R Nolan, J O’Sullivan and G Virgo (eds) Restitution: Past, Present & Future (Oxford: Hart Publishing, 1998) p 167. 11. E A Famsworth ‘PrecontractualLiability and Preliminary Agreements: Fair Dealing and Failed Negotiations’ (1987) 87 Col LR 217, 218-219. See also G Jones ‘ClaimsArising Out of Anticipated Contracts Which Do Not Materialise’ (1980) 18 UWOLR447.448. 12. McKendrick, above n 10, p 163. 13. See eg Brewer St Investments Ltd v Barclays Woollen Co Ltd [ 19541 1 QB 428 and generally J W Carter and M P Furmston ‘Good Faith and Fairness in the Negotiation of Contracts’ Parts I & I1 (1994) 8 JCL 1; 93, 106-1 17. 14. See eg Waltons Stores (Interstate) Ltd v Maher (1988) 164 CLR 387; MTA v Waverly TrunsitPty Ltd [1991] VR 176 andgenerally CarterandFurmston,above n 13, pp 100-103. Cases of proprietary estoppel are also relevant. Many such cases involve either incomplete or defective contracts, often involvingagreements which are less than fully formed because of the domestic context within which they are made. See eg Giumelli v Giumelli( 1999) 196 CLR 101. 15. See eg Banner Homes Group plc v Luff Developments Ltd [2000] 2 All ER 117, discussed further below, nn 69 and 76 and text to nn 101-103. 16. See eg Box v Midland Bank [1979] 2 Lloyd’s Rep 391; Esso Petroleum Co Ltd v Mardon [ 19761QB 801 and generally Carter and Furrnston, above n 13, pp 94-98.

156 Legal Studies The purpose of this article is to consider this question of how best to classify (cases of) precontractual liability. After focusing specifically on the question of whether principles of contract law or the principle of unjust enrichment can better solve problems of precontractual liability,” the issue of classification of such liability will be considered from a perspective of European (particularly German) civil law. From this comparative study, conclusions will be drawn which, it is suggested, are helpful in allowing us better to understand precontractual liability at common law. McKendrick has suggested that it matters ‘a lot’ into which categories we place such claims.” I will be suggesting that classification is important, but that we must not see the issue in terms of competition between one discrete category or another. Instead, it will be suggested that cases of precontractual liability can best he understood if they are seen to lie between existing categories and, perhaps controversially, I will be suggesting that the two categories between which cases of precontractual liability lie are not contract and unjust enrichment but, rather, contract and tort. First, however, it is necessary to consider the legal background to claims for precontractual work at common law.

2 ESTABLISHING PRECONTRACTUAL LIABILITY Common law courts continue to adhere to a general rule protecting the freedom of parties to negotiate contracts without the risk of incurring legal liability, a view well encapsulated by Sheppasd J in Saberno Pty Ltd v North Sydnqy MC: ‘It has long been the law that parties are free to negotiate such contract as they may choose to enter into. IJntil such contract comes about, they are in negotiation only. Each is at liberty, no matter how capricious his reason, to break off negotiations at any time. If that occurs that is the end of the matter and, generally speaking, neither party will be under any liability to the other.’Iy One ‘exception’ to this freedom to negotiate without risk of liability is where the parties themselves have sought expressly to regulate the future directions of their negotiations (and the consequences of the failure of such negotiations) by entering into a ‘process’ contract governing their negotiations or tendering processes. Such process contracts may take a variety of forms’” and, if enforceable, give rise to contractual (and not precontractual) liability if breached. It should

17. Others see the issue in different terms, such as a ‘competition’between restitution and estoppel. See eg D Y K Fung Pre-contractual Liability Rights and Remedies: Restitution and Promissory Esroppel (Asia: Sweet & Maxwell, 1999). Although estoppel is undoubtedly a significant mechanism for imposing liability in this context, it will be argued below that the principles underlying the imposition of liability via estoppel are derivative from principles of contract and tort law. 18. McKendrick,above n 10,p 164, and see also K Barker ‘Review’(1999) 19 LS 415,417. Contra D Davis ‘Restitution and Equitable Wrongs: An Australian Analogue’ in F Rose (ed) Consensus Ad Idem: Essays on the Law of Contract in Honour of Guenter Treitel (London: Sweet & Maxwell, 1996) p 176. 19. [ 19771 2 NSWLR 880 at 900. Cf R M Goode ‘England Report’ in Formation of conrracrs andprecontractual liabiliry (International Chamber of Commerce, 1990) p 58. 24). See Goode, above n 19, pp 55-56 and E A Farnsworth ‘General Report’ in ibid, pp 22-25.

Classifying precontractual liability: a comparative analysis 157 be noted that such process contracts may either expressly require negotiations to be carried out in good faith” or else, like contracts generally under American law and, perhaps, under Australian law,” be subject to an implied duty of good faith and fair dealing in the per3cormance of contracts. Such an implied duty of good faith and fair dealing, however, has not explicitly2’ been recognised as extending to precontractual negotiations t h e r n ~ e l v e s . ~ ~ None the less, even in the absence of a contract governing the negotiation process, cases arise in which a plaintiff has performed work of ‘such magnitude and of such a nature’2’ that the plaintiff can no longer be said to have taken the risk of such work not being recompensed should the contract negotiations fail in the circumstances in which they do.26In Subemo, for example, the plaintiff was requested to prepare considerable planning work after it had ‘successfully’ tendered for a contract for the lease and redevelopment of certain land. The defendant decided not to proceed with the project, but it was held liable to compensate the plaintiff for the work carried out. Sheppard J pointed out that: ‘as substantial and continuing work had been done over a period of three years . . . [it would] be unthinkable that the plaintiff would have been prepared to do what it did, if it thought that the defendant might change its mind about proceeding with the proposal.’” 21. Such contractswill most likely be unenforceable under English law (see Walford v Miles [ 199212 AC 128, though note the tort liability imposed in that case for misrepresentation, compensating the plaintiff‘s reliance loss), but likely be enforceable in Australia (see eg CoalCliffCollieriesPtyLtdvSijehamaPtyLtd(1991)24NSWLR 1 at21-27,perKirbyP), Canada, New Zealand and the United States. See generally N Cohen ‘Pre-Contractual Duties: Two Freedoms and the Contract to Negotiate’ in J Beatson & D Friedmann (eds) Good Faith and Fault in Contract Law (Oxford: Clarendon Press, 1995). In particular, Cohen emphasises the distinction between a process contract to negotiate in good faith and a ‘result-oriented’ contract to enter into a contract. As Cohen points out, giving certain content to the latter type of contract is much more difficult than to the former. See also J M Paterson ‘The Contract to Negotiate in Good Faith: Recognition and Enforcement’ ( 1 996) 10 JCL 120, and A F Mason ‘Contract,Good Faith and Equitable Standards in Fair Dealing’ (2000) 1 I6 LQR 66.80-8 1. 22. See eg Hughes Aircraft Systems International v Airservices Australia (1997) 146 ALR 1 at 36-37, per Finn J; Hughes Bros Pty Ltd v Trustees of the Roman Catholic Churchfor the Archdiocese of Sydney ( 1993) 3 1 NSWLR 9 1 ; and Alcatel Australia Ltd v Scarcella & Others ( 1 998) 44 NSWLR 349; but contrast Service Station Association Ltd v Berg Bennett & Associates Pty Ltd ( 1993) 1 17 ALR 393 at 406, per Gummow J. 23. Cohen, above n 21, pp 4 1 4 2 , however, considers that developments in the United States in the law of promissory estoppel have ‘opened an avenue for recognition of anoncontractual promise to negotiate in good faith’, and that United States law is thus ‘on the verge of recognising an explicit duty to act in good faith in negotiations’. 24. See Farnsworth, above n 20, p 2 1, Mason, above n 2 I , pp 69 and 77ff and Carter and Furmston, above n 13. See, however, Livingstone v Roskilly [ 19921 3 NZLR 230 at 237, per Thomas J. 25. See McKendrick, above n 10, p 167. 26. The reason for the failure of the contract negotiations is of importance because precontractual liability will depend on the nature and extent of the risk assumed by, or imposed by law upon, each patty. See further below. 27. [I9771 2 NSWLR 880 at 901. Substantial work or detriment carried out in circumstances such as these must be contrasted with the type of work a plaintiff may ‘be expected to do without charge when tendering’: William Lacey (Hounslow)Ltd v Davis [I9571 2 All ER 712 at 716.

158 Legal Studies Allowing the plaintiff‘s claim, Sheppard J awarded a quantum meruit for the reasonable value of the services performed by the plaintiff, a claim which Sheppard J classified as lying in restitution. As already noted above, however, a variety of other rules and doctrines have also been utilised by the courts in order to impose precontractual liability: thus, inroads into the preclusion against precontractual liability have been made in a fairly ‘piecemeal’?*fashion. One consequence of this diversity of judicial approaches has been a (largely academic) debate as to which approaches are of greater utility in solving precontractual liability problems and in explaining past decisions. A recent manifestation of this debate can be seen in the exchange between McKendrick and Hedley as to whether ‘unjust enrichment’ or ‘contract’ better explain cases. Their respective views will be canvassed below. Like Hedley, I will be arguing that unjust enrichment (which is said by some to explain many or all cases of liability in quasi-contractjrestitution) has little to offer by way of explanation of cases imposing precontractual liability, on the basis of quasi-contractlrestitution or otherwise. Further, like Hedley, I will be arguing that the principles of contract law are of considerable importance in explaining many cases of precontractual liability. Indeed, I will suggest that unjust enrichment theory itself draws on principles of contract law in determining questions of liability for services rendered, though it does so by formulating those principles under different guises. I will go on to suggest that principles of tort law are also of importance in explaining cases of precontractual liability. These conclusions, it will be suggested, apply generally to cases imposing precontractual liability, irrespective of the doctrines utilised to impose such liability.”

3 CONTRACT AND UNJUST ENRICHMENT McKendrick in his article ‘Work done in Anticipation of a Contract which does not Materialise’30has suggested that two main doctrines apply to cases of anticipated contracts, namely contract law itself and unjust enrichment. Hedley has summarised the burden of the article thus: ‘From [McKendrick’s] account, the balance between the two is fairly equal. Both approaches explain a certain amount; both have difficulties; each has something to contribute.’3’

It is necessary, therefore, to consider how these two approaches operate in the context of anticipated contracts.

28. Cohen, above n 21, p 32. 29. The remedial consequences which flow from the use of these different doctrinal mechanisms m q differ, but see further below. 30. Above n 10. 31. S Hedley ‘Work done in Anticipation of a Contract which does not Materialise: A Response’ in W R Cornish, R Nolan, J O’Sullivan and G Virgo (eds) Restitution: Past, Present & Future (Oxford: Hart Publishing, 1998) p 195.

Classifying precontractual liability: a comparative analysis 159 (a) Contract One way in which contractual reasoning is applied by the courts is by enforcing a contract anticipated by parties, despite the absence of a complete agreement in relation to all matters. In such cases, the courts effectively enforce an objectively discernible intention to contract, whilst accepting that some (inessential) terms of that agreement need to be implied. An example is provided by Butler Machine Tool Co Ltd v Ex-cell-0 Corporation (England) Ltd.32The case involved a so-called ‘battle of the forms’, with each party to the transaction utilising its own standard forms, despite the inconsistency of the terms of those forms as to the important matter of price. Nevertheless, each party proceeded on the basis that its terms and conditions governed the agreement to construct and sell a machine tool. Upon completion of the work, a dispute arose as to the price payable. The English Court of Appeal held that a contract existed on the terms of one of the party’s form. The court simply recognised that the parties entered into a consensual relationship intended to have legal force - that work was to be performed by one party and paid for by the other - and sought to give effect to those intentions. In one sense, perhaps, the reasoning of the court was artificial: not in finding a contract as such, but in resorting to traditional offer and acceptance analysis to try to overcome the difficulties of the incompatible terms in each party’s forms. Instead, it may have been more appropriate to consider the transaction as a whole in order to imply a term that the purchaser pay a reasonable price.33Such a technique is consistent with an increasing preparedness by common law courts to fill gaps in contract^,^^ including even as to the price payable by accepting that the parties must have intended a reasonable price where none was specified or objectively as~ertainable.~~ Cases such as Butler Machine Tool can be said to fall outside the realm of precontractual liability, since the contract anticipated (in broad terms) by the parties is in fact held to have been concluded, albeit in the absence of a concluded agreement in relation to all (non-essential) terms. However, the courts have also used contractual techniques in order to impose what is

32. [I9791 1 WLR401, [I9791 1 AllER965. 33. Cf the approach of Lord Denning in that case. See also Sir Anthony Mason and S J Gageler ‘The Contract’ in P D Finn (ed) Essays on Contract (Sydney: Law Book Co, 1987) p 1 1 , for other possible solutions. 34. See eg Carter and Harland, above n 7, pp 78-84, and S N Ball ‘Work Camed out in Pursuance of Letters of Intent -Contract or Restitution?’ (1983) 99 LQR 572,584, who has described this as the ‘realistic ability to construct a reasonable contract from the bare bones of agreement’. Carter and Furmston, above n 13, p 12, consider that a ‘generous approach of the courts [to implying terms] is essential today’. Note, however, the unwillingness of the House of Lords in WuLford v Miles [ 19921 2 AC 128, to ‘fill in’ the incomplete ‘lockout’ agreement in that case, and instead to treat it as uncertain; for criticisms, see eg Paterson, above n 21, and R P Buckley ‘Walford v Miles: False Certainty About Uncertainty - An Australian Perspective’ (1993) 6 JCL 58. We are concerned here with contracts other than for the sale of goods, in relation to which a term as to reasonable price, in the absence of agreement, is implied by various Sale of Goods legislation. 35. See eg Way v Latilla [ 19571 3 All ER 759, in which the House of Lords considered the parties’ previous negotiations in implying what was a reasonable price payable for the services rendered. See discussion in Dietrich, above n 9, pp 132-133, and also Hedley, above n 31, p 196.

160 Legal Studies genuinely precontractual liability by enforcing express or implied contracts (to pay for services rendered) ancillary or collateral to the incomplete main contracts. Thus, in Brewer St Investments Ltd v Burcluys Woollen Co Ltd,” for example, Somervell and Romer LJJ considered that the defendant’s (express) undertaking to pay for alterations made by the plaintiff to its own property was enforceable, despite the fact that an anticipated contract for the lease of the plaintiffs premises was never finalised. In cases such as Brewer St, contract is the basis of liability, as the courts enforce the parties’ clear intentions that the services performed were not to be gratuitous so that the recipient ‘cannot sensibly deny that those benefits are meant to be paid for’.” Yet the use of such contractual technique has been criticised by McKendrick and others for its artificiality in the finding of contractual intentions and for thus extending the boundaries of contract law too far.’x Reacting to such criticisms, Hedley has said: ‘But why should contract be excluded? Why should the parties’ hope that they would make a big contract prevent a finding that they have infuct made a more modest one? Why should the courts ignore good evidence of a contract? The courts can simply enforce the agreement the parties actually made, vague though it may be, and leave aside the parties’ fantasies as to agreements which they might have reached in other circumstances.’” Critics such as McKendrick argue that instead of contract, the law of restitution or unjust enrichment is often better suited to resolving questions of liability for work carried out in anticipation of ~ o n t r a c t . ~It” is necessary, therefore, to consider the principle of unjust enrichment and the detailed theories it has spawned. According to such theories, in order to establish a claim under the law of restitution or unjust enrichment, one needs to establish a number of elements, two of which are for our purposes of critical importance, namely that 36. [ 19541 1 QB 428. See also TuriffConstr-uctionsLtd v Regalia Knitting Mills Ltd [ 19721 2 Lloyd’s Rep 234, and cf Bluckpool and Fylds Aero Club Ltd I? Blackpool BC [ 19901 1 WLR 1195. 37. Hedley, above n 3 I , p 195. 38. See eg McKendrick, above n 10. For references to other critics and criticisms and for a more detailed discussion of the relative merits of contract and unjust enrichment in this context, see Dietrich, above n 9, pp 128- 133. 39. Hedley, above n 31, p 197. 40. Of course, restitution clearly has a role to play where a plaintiff transfers money in anticipationof acontract which does not eventuate. Such money (or it equivalent value) may be readily returnable should the contract not eventuate, subject perhaps, to any change of position defence operating to limit a defendant’s liability to the enrichment surviving in his or her hands. (It is questionable,however, whether such achange of position defence necessarily ought to be available in the context of precontractual transfers of money: see Dietrich, above n 9, pp 133-1 35.) The cases under consideration here, however, involve the performance of services,the incursion of expenses, and other detrimental reliance on the part of a plaintiff. Services need not even have been conferred upon defendants or, if they have, need not be unequivocally enriching. It is in the context of liability for services generally that unjust enrichment theory faces particular difficultiesas an explanatory concept. For an excellent summary of some of the difficulties, see S Hedley ‘Restitution: Contract’s Twin?’ in F Rose (ed) Failure of Contracts:contractual,restitutionup andproprietav consequences (Oxford: Hart Publishing, 1997)particularly pp 266273, and see further below.

Classifying precontractual liability: a comparative analysis 161 the defendant must have been enriched and that such an enrichment has been unjustly obtained (at the expense of the plaintiff). Let us consider these two elements in turn.

(b) Enrichment Much has been written on the valuation of enrichment or benefit where the provision of services are at issue. The debate is too complex and wide-ranging to be discussed in depth here.4’For present purposes, it suffices to say that given the difficulties which arise, a number of tests of enrichment have been formulated. Leaving aside cases in which a tangible benefit, such as a valuable end-product or the saving of necessary expense, has been these tests adopt a subjective approach to services as benefit. They focus on defendantrelated matters such as the conduct of the defendant in order to establish an acknowledgment on his or her part that the services performed by the plaintiff are of value. In other words, it is necessary to show that the defendant has either bargained for, or requested?’ or acquiesced in the conferral of, or ‘freely accepted’44the services. Whichever of these tests is adopted, and there is much disagreement as to which are appropriate, the very existence of benefit is thus established by reasoning which appears to be contractual in its essence: that services are of benefit where there is an express or implied (as objectively determined by the defendant’s words or conduct) acknowledgment or intention that such services are not to be considered to be conferred gratuitously. It could be said that the tests applied emphasise the notion of an underlying, essentially consensual bargain. Ironically, the test of enrichment which has been subject to the most criticism by unjust enrichment theorists themselves, namely that of ‘free has been criticised precisely because it may not identify conduct which demonstrates sufficiently such an intention to accept that the services are, indeed, of value (and to be paid 41. For a detailed discussion, see Dietrich, above n 9, pp 50-77. 42. It should be noted that in many of the cases under consideration, there exists no such tangible benefit, the plaintiff‘s efforts having been wasted, for example, as a result of the failure to finalise a contract. Further, even if such a benefit exists, it is not necessarily the appropriate measure of any liability. Instead, compensation of reliance loss ought to be the minimum measure of remedial relief where the elements of precontractual liability (considered below) are satisfied. Consider, however, Van den Berg v Giles [I9791 1 NZLR 1 1 I , in which the remedy granted was restitution of the benefit conferred; it is suggested that this was not the appropriate measure of relief in the circumstances. 43. Cf McKendrick, above n 10,pp 175-177, and A Burrows Law ofRestiturion (London: Butterworths, 1993) pp 14-15. 44. For a comprehensive list of different suggested tests of enrichment, see Hedley, above n 40, p 267 and references cited therein. 45. See eg criticism by A S Burrows ‘Free Acceptance and the Law of Restitution’ (1988) 104 LQR 376; and M Gamer ‘The Role of Subjective Benefit in the Law of Unjust Enrichment’ (1990) 10 OJLS 42. 46. It should be added that even if one accepts such subjective tests of enrichment as having explanatory value, some cases imposing precontractual liability, ostensibly in restitution or quasi-contract, cannot be said to involve any benefit to defendants in any form. See further below, n 74.

162 Legal Studies It should also be added that the remedial response to a finding of precontractual liability, irrespective of whether imposed on the basis of contract or restitution, often takes the form of a quantum meruit for the ‘reasonable’ value for the services rendered.” In determining the measure of such reasonable value, courts have often referred to the intentions of the parties as to the price payable for the services, such as was done by the House of Lords in Wuy v L . ~ t i l l aSuch .~~ references to the parties’ intentioris in order to determine the appropriate measure of relief, it is suggested, irrespective of whether contract or restitution is the basis of relief, reconfirms the essentially contractual approach of the courts. (c) The ‘unjust factor’

According to unjust enrichment theory, in order to obtain restitution, a plaintiff needs to show not only that the defendant has been enriched, but that the enrichment is ‘unjust’. This requires the identification of some ‘unjust factor’ justifying restitution. In the context of failed contractual negotiations, there is some debate as to what the relevant unjust factor may be. Two possibilities have been suggested. The first of these is that the defendant has requested or freely accepted the work.49As already indicated above, the concept of free acceptance focuses on the acknowledgment on the part of a defendant that the services are of benefit and are to be paid for, such as where a ‘recipient of services knows that a benefit is being offered to him non-gratuitously and where he, having the opportunity to reject, elects to accept’.s”Such an unjust factor again refers us back to the intention of the defendant that the work was not to be gratuitous. Presumably, such conduct may also justify an expectation on the part of the plaintiff that the work will be paid for, so that we can say that there is an intention on the part of both parties that the work is to be remunerated. Disguising such reasoning in terms of an unjust factor merely hides its essentially contractual nature.5’ 47. For more detailed discussion of the valuation of quantum meruit, see E McKendrick ‘The Battle of The Forms and The Law of Restitution’ (1988) 8 OJLS 197,218. 48. [ 19371 3 All ER 759. Although the decision in Way v k t i l l a has been described by Birks as ‘entirely satisfactory’, he considers the contractual reasoning utilised to achieve it as ‘artificial’: Introduction to the LUM’ofRestitutiori (Oxford: Clarendon Press, 1985) p 272. With respect, it is difficult to see how the House of Lords could have reached the result it did without such contractual reasoning. See Dietrich, above n 9, pp 132-1 33. For cases in quasi-contract/restitution making reference to parties’ agreements to determine reasonable value, see eg Scarisbrick 1.1Parkinson ( 1869) 20 LTR 175 and Ward \i Grifiths (1928) 28 SR (NSW) 325. 49. Cf McKendrick, above n 10, p 18 1. 50. Birks, above n 48, p 265. Cf Bronner v First Artists’ Munczgernenr Pt?; Lrd [ 19931 2 VR 221 at 260. 51. Or else it adds a further, unnecessary stage of reasoning. Consider eg McKendrick, above n 10, p 181, who suggests that free acceptance justifies restitution where there is some ‘initial unconscionability’, such a5 where a defendant had no actual intention to pay. If, however, a defendant requests work intending to pay for it, but subsequently changes his or her mind, then according to McKendrick, ‘it is not easy to see why the terms of the initial request should give rise to a claim for payment. The request, when made, did not generate a restitutionary obligation to pay for the work done because the parties at this stage believed that payment would be made through the completion of the contract.’ Surely, the answer is obvious: the defendant should pay because he or she requested work intending that it be remunerated in some way.

Classifying precontractual liability: a comparative analysis 163 The second suggested unjust factor is that of a ‘failure of consideration’ or condition. This unjust factor faces a significant initial hurdle to its application to recovery for precontractual services, namely that ‘failure of consideration’ has only been utilised by the common law courts in relation to the recovery of money.s2Putting this difficulty aside, however, the concept also suffers from considerable conceptual problems. These can be simply illustrated. A failure of consideration is said to arise upon the failure of the expected completion of the main contract and consequently, failure of the expected remuneration of services rendered. Yet many cases of work done precontractually do not give rise to any claim at all. How, then, is one to distinguish those cases in which there is a failure of condition and those in which there is not? McKendrick suggests that this depends on the ‘basis’ upon which, and at whose risk, the work was carried out. McKendrick accepts that this, in turn, depends upon the construction of the ‘understanding’ of the parties.53For example, if both parties appreciate that they are free to withdraw from negotiations at any time, then they generally may do so without legal liability arising.s4 The issue thus again turns on contractual reasoning, namely the intentions and expectations of the parties as to whether the work was to be at the risk of one or the other should the anticipated contract not materialise. McKendrick acknowledges as much when he states: ‘One objection to the failure of consideration analysis is that it has contractual overtones and so should be analysed in contractual terms. . . In other words, there is in these cases an implied agreement or understanding between the parties that work done will be paid for if a contract fails to materialise.’55 McKendrick’s criticism that such an approach is often ‘implausible’ because the parties are ‘unlikely to have considered expressly what is to happen if a contract fails to m a t e r i a l i ~ e can ’ ~ ~ be rebutted by restating Hedley’s comments quoted above5’ and adding that contractual intentions need not be express intentions.

(d) Summary To summarise, the meanings of the concepts of both ‘enrichment’ and ‘unjust’, which are critical to an unjust enrichment analysis of cases of precontractual liability, are the subject of considerable debate and raise considerable complexities. As Hedley has said: ‘we simply don’t know what “enrichment” or “unjust” mean in practical terms. I don’t need to belabour the point, because McKendrick’s paper makes it for me. The issue of “enrichment”, he says, is shrouded in uncertainty, and it is a very difficult question to answer . . . There is no single “unjust 52. See McKendrick, above n 10,p 182. The other hurdle to the application of ‘failure of consideration’ as an unjust factor is the ‘traditional’ requirement that such failure needs to be total; there are strong indications, however, that this requirement will be abrogated by the courts. 53. McKendrick, above n 10, p 183. 54. Eg Attorney General 11 Humphreys Estate (Queens Gardens) Ltd [ 19871 1 AC 114. 55. McKendrick, above n 10, p 182. 56. Ibid. 57. See text t o n 39 above.

164 Legal Studies enrichment” view here, only a sequence of suggestions - and no one of these suggestions has convinced very many people.’58 My argument goes one step further in suggesting that ‘enrichment’ and ‘unjust’ in this context are given operative meaning by resort to (and would be largely empty concepts unless defined by) ideas and principles at the core of contract law. Liability can be said to arise because of the parties’ intention that the plaintiff‘s work is not performed gratuitously and that the risk of the contract negotiations failing lies with the defendant. There is no real competition between contract and unjust enrichment here. Unjust enrichment analysis and reasoning merely cloaks in a different guise essentially contractual tests relevant to establishing liability, tests focusing on the intentions and expectations of the parties that services are to be paid for. The point is all the more difficult to refute when one considers that the judgments of the courts almost without exception do not address questions of ‘enrichment’ and ‘unjustness’ at all.s9 Instead, decisions are based on a determination of, amongst other factors, the express and implied intentions and expectations of the parties, their intended allocation of risks and the like. One advantage of recognising the relevance of contractual reasoning in determining issues of precontractual liability is that the scope of any such liability should reflect as much as possible the mutual intentions and expectations of the parties which are discernible from their relationship and transaction, such as to the quality and timing of the work performed. In other words, the work which the defendant is to pay for should fulfil all other requirements of the parties’ incomplete agreement.h” McKendrick has suggested that similar results can be achieved by resort to unjust enrichment, that it is possible -

‘to scale down [a defendant’s] enrichment in accordance with the established tests to reflect the fact that the defendant did not get what he asked for. A defendant who does not get what he asked for should be entitled to devalue subjectively, and, to that extent. the restitutionary claim should be reduced.’6’ Since, however, the extent of the ‘restitutionary’ obligation imposed varies according to the parties’ intentions as to the quality and timing of the work, then reference to the principle of unjust enrichment again merely disguises that effect is being given to the parties’ intentions. In essence, the parties’ underlying agreement is given effect on both sides, and it is a fiction to suggest that the basis of recovery is the reversal of a defendant’s enrichment. In suggesting that unjust enrichment has little explanatory value and draws largely on principles of contract, it is not intended to suggest, however, that precontractual liability can be explained solely in terms of contractual reasoning. Clearly, the very formulation of the issue in terms of precontractual liability suggests otherwise. It is necessary, therefore, to consider the common elements of all cases imposing precontractual liability, elements which, it is suggested, are present irrespective of whether the courts have utilised contract law (such as by enforcing an agreement ancillary to that anticipated), the law of quasicontract, restitution or unjust enrichment, estoppel, or tort law. 58. Hedley, above n 31, p 197.

59. Hedley, above n 3 I , p 197, and cf Hedley, above n 40, pp 271-272. 60. See Ball, above n 34, p 577, Farnsworth, above n 1 1, p 220. 61. McKendrick, above n 10, p 173.

Classifying precontractual liability: a comparative analysis 165 4 THE COMMON ELEMENTS OF CASES IMPOSING PRECONTRACTUALLIABILITY

(a) A belief (at least on the part of the plaintiff) that a contract exists, or an expectation of a future contract Although the parties may not have agreed on all the terms of the contract, the plaintiff none the less legitimately expects the agreement and contract to be finalised in due course. At other times, the parties may have agreed on all essential terms of the contract and only some final, necessary step to completion of a contract has not been carried out, but is considered a mere formality.62At times, a request by a defendant that substantial preliminary work be carried out by the plaintiff may raise, or contribute to, an expectation that the main contract will be c~ncluded.‘~ For example, in Brewer v Chrysler Canada Ltd@ the plaintiff, whose application for a car dealership had been recommended and required only head office approval, was encouraged to prepare himself to commence business by buying stock, hiring staff, renting storage space and the like. The plaintiff was held to be entitled to a ‘quantum meruit’ award for the costs incurred as well as to two months’ salary, despite the absence of a binding contract for the dealership. If, however, both parties appreciate that they are free to withdraw at any time,6sor if the plaintiff deliberately gambles as to the future outcome of negotiations for his or her own purposes,66then failure to complete the contract will generally not lead to legal liabilit~.~’

(b) A plaintiff has relied detrimentally on that belief or expectation A plaintiff‘s detrimental reliance can take a variety of forms, such as by abandoning existing legal rights68or otherwise forgoing o p p ~ r t u n i t i e s by ;~~ 62. See eg Waltons Stores (Interstate)Ltd v Maher (1988) 164 CLR 387. 63. Eg William Lacey (Hounslow)Ltd v Davis [ 19571 2 All ER 7 12; and Sabemo P o Ltd v North Sydney M C [1977] 2 NSWLR 880. 64. [ 19771 3 WWR 69. 65. Eg Attorney General v Humphreys Estate (Queens Gardens) Ltd [ 19871 1 AC 1 14. 66. Cf Austotel Pry Ltd v Franklin Selfserve Pty Ltd (1989) 16 NSWLR 582. 67. It is suggested, however, that an agreement to pay for preliminary work will be enforced if such an ancillary agreement is intended to operate irrespectiveof whether a final contract is ultimately completed. Cf Brewer Street Investments Ltd v Barclays Woollen Co Ltd [ 19541 1 QB 428 at 433, per Somervell LJ, in which case both parties knew that either could resile from the contract negotiations, but the defendant’s express undertaking of responsibility for preliminary work carried out by the plaintiff was held to be enforceable. 68. See Avondale Printers Stationers Ltd v Haggie [1979] 2 NZLR 124. 69. See eg Banner Homes Group plc v Luff Developments Ltd [2000] 2 All ER 117, in which the appellant, on the faith of aprecontractual, ‘informal’ arrangement with the first respondent to purchase a valuable commercial site as joint venturers, had treated the site as ‘out of play’, that is, did not consider the site as a potential acquisition for its own portfolio. Further, on the assurances of the first respondent that the parties would enter a contract governing theirjoint venture as soon as the fmt respondent’ssolicitorreturned from holidays, the appellant allowed the second respondent (a wholly owned subsidiary of the first respondent) to exchange contracts with the vendors of the site even though no formal written agreement between the appellant and first respondent had been entered into. See also below n 76 and text to nn 101-103.

166 Legal Studies carrying out work that is preliminary to the performance of the expected main contract;”’ or else by carrying out work which is unrelated to the performance of the expected contract, such as improvements upon a defendant’s land which the plaintiff expects to inure to his or her benefit upon completion of a binding contract for sale of the land.” At times, the plaintiff may have begun or completed the very performance which forms the subject matter of the incompletely negotiated contract, such as where a manufacturer was requested to, and did, commence manufacturing certain goods despite essential terms of the parties’ proposed contract not having been agreed.72 ‘In all these cases, a plaintiffs reliance proves detrimental when the defendant refuses to complete the agreement or accept any responsibility to the plaintiff who has changed his or her position on the faith of the incomplete contract. This element of detriment is highlighted by the belief of the plaintiff that the reliant action was not, ultimately, intended to be gratuitous. Even where the plaintiff does not expect the work to be paid for separately, such a plaintiff must have expected remuneration in some form. . .’73 It is not necessary that a defendant has benefited in order for precontractual liability to arise, even in cases imposing liability in restitution. Perhaps one of the best examples is Sabemo PQ Ltd v North S.ydney IVC,’~in which Sheppard J expressly found that the defendant had not been enriched by the plaintiff‘s wasted expenditure. Nor ought there to be a need for such a benefit, for the liability of defendants is established by showing some fault o n their part (see (c) below), and consequently, such defendants ought at the very least to make good plaintiffs’ losses and not be limited to disgorging benefits (if any) 70. Eg Brewer Street Investments Ltd v Barclays Woollen Co Ltd [ 19541 1 QB 428 at 429; Angelopoulos v Sabitino (1995) 65 SASR 1. 71. See eg Estok v Heguy (1963) 40 DLR (2d) 88; T & E Developments v Hoornaert (1977) 78 DLR (3d) 606; Lexane Pry Ltd v Highfern Pry Ltd [ 19851 1 Qd R 446. 72. Eg British Steel COT v Cleveland Bridge & Engineering Co Ltd [ I9841 1 All ER 504. 73. Dietrich, above n 9, p 126. 74. [I9771 2 NSWLR 880. Other examples can be found, especially from Canadian jurisdictions. See Dietrich, above n 9, pp 64-66. Two types of cases are illustrative:(1) those in which a plaintiff incurs expenses in preparation to performing an anticipated contract, where such work has not been requested (at best only encouraged or approved) by the defendant, or may even have been only reasonably foreseeable; and (2) cases in which the plaintiff performs work which would have benefited the plaintiff, had the contract proceeded, but which benefits do not flow to the plaintiff or the defendant because the contract does not eventuate. An example of type ( 1 ) is Brewer v Chrysler Canada Ltd [1977] 3 WWR 69, in which the plaintiff incurred expenses in preparation for commencing business, in expectation of being awarded a car dealership; the ‘benefit’ of this work was ultimately either wasted or went to a third party (the party eventually granted the dealership). Importantly, the plaintiff, succeeding in a quantum meruit claim, was awarded not only the full expenses he had incurred, but also successfully recovered two months’ salary for the time spent organising the preparatory work (and see also Maclver v American Motors (1976) 70 DLR (3d) 606). Cases of type ( 2 ) often involve work on the defendant’s land (the subject of the incomplete contract) and are particularly interesting, because in some examples the work will not have resulted in any objectively measurable benefit (see eg Estok v Heguy (1963) 40 DLR (2d) 88, in which the plaintiff fertilised land of the defendant which he was not intending to use for agriculture. and cf Preeper v Preeper (1978) 84 DLR (3d) 74, and cases above n 71).

Classifying precontractual liability: a comparative analysis 167 received. Even if a defendant has obtained a benefit as a result of the plaintiff‘s efforts, unjust enrichment theory requires that such benefit be ‘at the expense of‘ the plaintiff, that is, that there must also be a corresponding loss to the plaintiff. Consequently, any ‘restitution’ of the defendant’s benefit automatically also compensates such (equal) loss. Hence, it is suggested, rather than divide the cases according to a largely artificial determination as to whether the defendant has or has not been enriched, it is possible to identify the common features of all cases of precontractual liability, namely that the plaintiff has suffered some detriment and can establish the first and third elements of liabilit~.’~ In exceptional circumstances, a plaintiff need not have suffered financial detriment in reliance upon a belief or expectation. According to a body of equitable case law, where a defendant acquires property and, in breach of a precontractual, informal arrangement, denies the plaintiff a beneficial interest in the property, the plaintiff may be entitled to equitable relief in the form of a constructive trust over part or all of the property. As it is not necessary that the plaintiff has suffered financial detriment as a result of relying on the arrangement in order for equitable relief to be granted, it has been suggested that it suffices if the defendant has obtained an advantage from the precontractual a~~angernent.’~ In such cases, however, plaintiffs who have not suffered financial detriment (when compared with their status quo before reliance on the informal arrangements) will at the very least have forgone opportunities they may otherwise have pursued, such as bidding for the desired property themselves.

(c) The defendant’sconduct Liability will be imposed where the defendant’s conduct is such: (i) that an obligation to bear the risk of an incomplete agreement including, specifically, to pay for the plaintiff‘s work or make good his or her reliance loss, has been or will be assumed by the defendant; or (ii) that the risk of an incomplete agreement 75. The reasons for the remedy in both reliance-based cases and in restitution (even if such reasons are disguised by the label ‘unjust factor’) are the same: see (c) below. Since restitution must of necessity also make good a plaintiff‘s corresponding loss, there is thus no reason to continue with a separate unjust enrichment analysis. Contrast McKendrick, above n 10,pp 180-181. 76. See eg Banner Homes Group plc v Luff Developments Ltd [2000] 2 All ER 1 17 at 140-141 and below, text to nn 101-103. The constructive trust in cases such as this is based on what has been called the Pallant v Morgan equity, which arises where it would be unconscionable for a defendant to deny the plaintiff a beneficial interest in property in breach of an informal (that is, not contractually enforceable) arrangement or understanding. The elements which need to be made out in order to establish the equity (see at 138-139) are very similar to the elements of precontractual liability set out in this article. In Pullunt v Morgan [ 19531Ch 43, [ 195212 All ER 95 1 itself, the plaintiff had failed to bid for property on the basis of an agreement with the defendant that he should bid for it and reconvey part of it to the plaintiff. On the facts, the plaintiff had not suffered any financial detriment as a result of the defendant’s refusal to comply with the agreement, since any bid of the plaintiff would in any case have been exceeded by that of the defendant. The defendant, however, obtained an advantage as a result of the agreement in obtaining the property more cheaply than would otherwise have been the case. See discussion of the case in Banner Homes Group plc at 129-1 30.

168 Legal Studies (and usually of thus having to bear any reliance loss) is imposed, in the circumstances, upon the defendant. 77 This element encompasses two quite different grounds for imposing liability. The starting premise for both is that the risk of negotiations failing (and of any losses consequential upon such failure) is on both parties to negotiations and that it is thus necessary to identify some reason for shifting such risk. The first of these grounds of liability is founded upon a defendant’s intention, as either expressed or (objectively) discernible from his or her conduct, to assume an obligation in relation to the work performed or expenses incurred by the plaintiff. In other words, despite the incomplete agreement, it is none the less possible to discern some ‘underlying measure of agreement’7Rbetween the parties. The defendant has assumed the risk of the consequences of the anticipated contract not eventuating and, in particular, though not limited to this, that certain non-gratuitously performed services may thus need to be paid for separately. Should the defendant refuse to pay for the plaintiff‘s work, he or she will be in breach of such an assumed obligation. Hence, such precontractual liability can be described as contract-like, as it ‘is imposed on defendants who either ( I ) had the actual intention to assume an obligation or risk, or (2) who can be taken to have assumed an obligation or risk because they have conducted themselves in a way that can be interpreted by a reasonable person as suggesting that such obligation or risk has been or will be assumed. In either case, the plaintiffs are entitled as a consequence of such conduct to order their affairs in reliance thereon. Although in the circumstances, the rules of contract say that no enforceable contract exists . . . nonetheless liability may attach to a defendant’s conduct, provided the plaintiff has detrimentally relied on it. Such liability is very similar to contractual liability and is established by proving, in essence though not in form, similar elements. In effect, though there may not have been an assumption of a contractual obligation, nonetheless there was an assumption of some other, collateral obligation (or contract-like duty) on which reliance is j ~ s t i f i e d . ’ ~ ~ The second broad ground of precontractual liability involves a shifting of risk, not because of an intention to assume it, but rather because of the

77. R Grantham ‘Securityof Contract: The Challenge from Restitution’ (2000) 16 JCL 102, 1 15, n 77, has suggested that a similarly formulated third element or factor of precontractual liability ‘canbe understood as merely a conclusion legitimately to be drawn from the presence of the first two factors’. Such a suggestion does not correspond with the case law, however. Whichever doctrinal mechanisms are used to impose precontractual liability, detrimental reliance (element (b)above) on a belief that a contract exists or will exist (element (a) above) alone does not suffice for liability to be imposed. For example, if estoppel is relied on to impose liability,a plaintiff would need to show that the defendant both was responsible for (or at least was aware of) the plaintiff‘s belief and either encouragedor acquiesced in the plaintiffs detrimental reliance (such that the defendant’sconduct is unconscionable). That people make detrimental choices on the basis of beliefs which turn out to be false or which are falsified is not reason enough to shift the consequences of their choices to others. There must exist some reasons for shifting the losses to others and, it is suggested, such reasons are encapsulatedby the third element of liability as set out herein. 78. G H L Fridman Restitution (Ontario: Carswell, 2nd edn, 1992)p 301. 79. Dietrich, above n 9, p 109.

Classifying precontractual liability: a comparative analysis 169 imposition by law of such risk upon a defendant. Of course, sufficiently serious reasons must exist for such an imposition. The circumstances must justify a legal determination that a defendant is the more appropriate party to bear the particular loss, for example, because the defendant is in a better position to be aware of the impediments to a complete agreement,s0 and is aware of the plaintiff‘s vulnerability to incur foreseeable detriment in reliance upon an expectation that a contract will be finalised.” In effect, the law seems to require the defendant to fulfil a ‘positive’ duty of care to warn or safeguard the plaintiff against foreseeable pure economic loss. This duty is ‘higher’ than that which is ordinarily owed to one’s neighbours and stems from the parties’ close, near-contractual relationship. Examples are provided by cases of proprietary estoppel by acquiescence, such as Denny v Jensen.82In that case, the plaintiff had improved the defendant’s land in the belief that he had a contract to purchase the land, in circumstances in which the defendant ‘must have known’ of the plaintiff‘s detrimental conduct and thus shared ‘the responsibility’ for the plaintiff‘s detrimental reliance because of his failure to alert the plaintiff as to the true position.R3In acquiescence cases such as these, I have suggested p r e v i o ~ s l y , ~ ~ ‘the parties are invariably in relationships of close proximity, where a defendant will have or ought to have known that the plaintiff was “likely” or “might well” undertake detrimental actionsR5which the defendant could readily avert. As a result, the defendant is under a duty to take reasonable steps to apprise the plaintiff of the true facts before any detrimental reliance occurs. In short, it appears in such cases that the “neighbourhood idea [is] most clearly at work”.86The significant difference in these acquiescence cases is that the courts are prepared to extend notions of duty of care to require positive actions on the part of a defendant, so that in effect, they are imposing a duty to re~cue.’~’ On whichever of these two broad grounds it is based, precontractual liability can therefore be described as being fault-based, with the obligation upon a defendant arising, in essence, for a breach of a contract-like obligation or a breach of a tort-like duty of care. (Such descriptive labels are not intended to suggest, however, that there exist two distinct sub-categories of liability. Instead, as will be argued below, they represent merely two ends of a spectrum of possible reasons for imposing liability.) Is it therefore relevant which party (and in which circumstances) breaks off contractual negotiations? Clearly it must be, not because breaking off negotiationsper s e is blameworthy, but because the reason for the failure to complete the contract must be within the scope of the risk either 80. Eg Box v Midland Bank [ 19791 2 Lloyd’s Rep 39 1 ; perhaps cases cited above n 7 1. 81. Eg Van den Berg v Giles [ 19791 1 NZLR 1 1 1. If a plainti’is better equipped to be aware of or safeguard against impediments to a final contract, then such a plaintiff ought not to succeed in an action for losses arising from a failure to complete the contract. Cf Jennings & Chapman Ltd v Woodman, Matthews & Cu [ 19521 2 TLR 409. 82. [I9771 1 NZLR 635. 83. Ibid, 637-638. 84. Dietrich, above n 9, p 145. 85. See K Grey Elements of Land Law (London: Butterworths, 2nd edn, 1993) p 33 1 , nn 13-1 4, and cases cited therein. 86. P D Finn ‘Unconscionable Conduct’ (1994) 8 JCL 37,49. 87. Ibid, p 42.

170 Legal Studies assumed by, or imposed upon, the defendant.” In either case, the determinative question ultimately is: ‘which party bears the risk of those losses occurring in those particular circumstances?’ For example, if a defendant has assumed a risk, whether liability in fact arises will depend on the exact scope of that assumed risk. Presumably, the risk assumed will not extend to cover all possible contingencies (for example frustrating events) which might prevent the contract eventuating. As I have suggested previously:R9 ‘A defendant will generally not be taken to have assumed in absolute terms the risk of an incomplete contract, so as to include the possibility of the plaintiff breaking off negotiations. In cases in which a plaintiff was at fault in the breakdown of the relationship and thus the failure to complete the contract, recovery has generally been denied.’”’

Similarly, I would suggest, the courts will only in exceptional cases impose the risk of a particular loss upon a defendant where such loss has arisen because of the plaintiff‘s decision to withdraw from negotiations. Consequently, it is always necessary to consider the nature of the risks assumed by, or which ought to be assumed by, each party to a precontractual arrangement. For example, in the recent decision of Easat Antennas Ltd v Racal Defence Electronics Ltd,” the claimant had incurred considerable costs in anticipation of being awarded a sub-contract with the defendant, should the defendant’s tender for a defence project contract prove successful. The defendant’s bid was successful, but the sub-contract was awarded to someone else. It was held that the claimant was entitled to a quantum meruit for the costs it had incurred. Although the claimant did not expect to be paid for such work separatelyy’ and took the risk that, if the defendant’s bid would fail, its costs would not be remunerated, it was not, however, ‘prepared to run the risk that, if the defendant’s bid succeeded, it would not be rewarded’.y3Even though the agreement by which the defendant had agreed to award the sub-contract to the claimant was uncertain (and thus not enforceable), none the less it was held that the ‘whole purpose and underlying a s ~ u m p t i o n of ’ ~ that ~ agreement was that the claimant did not accept (and the defendant did accept) the risk of any costs incurred by the claimant, should the sub-contract not be awarded despite a successful bid by the defendant.

88. See also generally G R Shell ‘Opportunismand Trust in the Negotiation of Commercial Contracts: Towards a New Cause of Action’ (1991)44 Vanderbilt LR 221. 89. See Dietrich, above n 9, p 127. 90. E g Construction Design & Munagernenr Ltd 11 New Brunswick Housing Corp (1973) 36 DLR (3d) 458. 91. 7 September 2000, unreported, HC. Hart J (www.law.cam.ac.ukstitution/archive/ englcases/easat.htm). 92. ‘The claimant’s costs were incurred in the belief that it would get them (and more) back under the sub-contract if the defendant’s bid succeeded.’ Ibid, para [70]. 93. Ibid. 94. Ibid.

Classifying precontractual liability: a comparative analysis 171 5 CATEGORISINGPRECONTRACTUAL. LIABILITY The above description of the elements necessary for establishing a precontractual claim has highlighted the relevance of principles of liability from within both contract and tort law. Of course, in many of the cases under consideration, a claim brought within either contract or tort ‘proper’ would fail, since not all the elements necessary for establishing such a claim are necessarily present. If it were otherwise, such cases would present few difficulties. None the less, many of the core concepts relevant to establishing a breach of contract or a tort are in evidence, to varying degrees, in precontractual liability cases. If the basis for establishing precontractual liability can thus be described in terms of principles operating in contract and tort, it is worthwhile to consider the nature of the interests protected by a finding of precontractual liability; in other words: what are the remedial consequences of liability, and how do the interests protected by such remedies relate to the interests characteristically protected by contract and tort? The second element of precontractual liability set out above, that a plaintiff has suffered loss as a result of relying on his or her (reasonable) expectations, is suggestive of tort law. Unlike the position where a valid contract exists, a plaintiffs unfulfilled expectations themselves do not generally give rise to legal consequences in the absence of such detrimental reliance. One might expect that this tort-like requirement of loss will lead to the compensation of such reliance loss as the appropriate remedial response to a finding of precontractual liability. Interestingly, however, such a remedial response does not always follow and it is worthwhile briefly to consider the different remedial consequences which flow from a determination of precontractual liability. Perhaps the most common remedial response to a finding of precontractual liability is that of a quantum meruit award for the reasonable value of services rendered. This remedy may be granted irrespective of whether the courts resort to contract law (by filling gaps as to the price payable either in the anticipated main contract or in an ancillary contract) or quasi-contracthestitution. Arguably, such a remedy can be described as making good the service performer’s reliance but two points ought to be noted. First, the quantum meruit award typically includes an element of reasonable profit.96Secondly, in determining what is, indeed, a reasonable value of services, the courts consider the intentions of the parties in determining such value.97Hence, a plaintiff is awarded what he or she

95. Eg where the award is limited to the actual expenses incurred by the plaintiff without any profit component. See further Dietrich, above n 9, pp 7 1-72. 96. See McKendrick, above n 47, p 218. 97. This will be the case, again irrespectiveof whether the claim is in contract (see eg Wuy v Latilla [ 19371 3 All ER 759) or restitution. The clearest illustration of the latter is provided by cases involving defective or unenforceable contracts, where the existence of a complete agreement makes it easier, as a matter of fact, to discern the parties’ intentions. Given the willingnessof the courts to considerthe parties’ intentions in such cases, even whilst insisting upon the restitutionary rather than the contractualbasis of such claims (see eg Puvey & Murrhaos Pry Ltd v Paul (1987) 162 CLR 22 l), there is no reason why parties’ intentionsas to the price payable for services will not also be considered in the context of incomplete contracts where there is some evidence of such intention.

172 Legal Studies might reasonably be entitled to expect, that is, his or her reasonable expectations.yx At other times, however, the courts will protect a plaintiff‘s uctuul expectations in relation to the anticipated contract, usually either by an order of specific performance or damages measured by the plaintiff‘s (positive) expectation interest. Apart from resort to contract law proper, this may be achieved by resort to estoppel, such as where parties have agreed on all essential terms and only some final, formal step is required for a complete contract.99In such cases, the parties are as near as it is possible to being in a contractual relationship without formally being so. It must be stressed, however, that fulfilment of expectations is not the only remedy flowing from estoppel. Estoppel is a doctrine which exhibits remedial flexibility and compensation of a plaintiff‘s reliance loss is another means by which the equities in the plaintiff‘s favour may be given Alternatively, courts have also utilised the constructive trust in order to protect a plaintiff‘s expectation interest under a precontractual arrangement. A recent illustrative example is provided by the case of Banner Homes Group plc v Luff Developments Ltd.Io’ In that case, the appellant and first respondent had agreed to purchase a commercial property as joint venturers, by means of a subsidiary company shares in which were to be held equally by the two parties. The first respondent had second thoughts about the appellant as a joint venture partner, but continued to assure the appellant that the joint venture would proceed because it feared that the appellant would itself bid for the property. The appellant was led to believe that a formal written agreement would be concluded, there being no outstanding differences i n principle to such agreement and that the only reason for delay was the absence of the first respondent’s solicitor on holiday.lO? The first respondent acquired the property through the second respondent, its wholly owned subsidiary. The appellant successfully claimed that half the shares in the second respondent be held on constructive trust for it. In effect, therefore, the appellant obtained half the shares in the subsidiary company owning the commercial site, the very interest which the in-principle agreement between the parties had promised. We could therefore

98. Cf Mason, above n 21, p 72, who emphasises contract law’s ‘new focus on the reasonable expectations of the parties’, though his comment is not necessarily primarily directed at the remedial consequences of liability, but to the broad protective purpose of rules imposing such liability. Although I am using the term ‘reasonable expectations’ to mean the same as what a plaintiff ‘might reasonably be entitled to expect’ (as seemingly does J Stapleton ‘Good Faith in Private Law’ (1999)52 Current Legal Problems 1,17), the two formulations are not coterminous; the term ‘reasonable expectations’ may also encompass a plaintiffs actual expectations which are reasonably held in the circumstances. 99. Eg Walrons Stores (Interstute) Ltd v Maher (1988) 164 CLR 387. 100. Some cases of proprietaryestoppel are illustrative,eg Uniq Joint StockMutual Banking Association v King (1858) 25 Beav 72,53 ER 563. 101. [2000] 2 All ER 1 17. The constructive trust in such cases arises where it would be unconscionablefor a defendant to deny the plaintiff a beneficial interest in property in breach of an informal (that is, not contractually enforceable) arrangement or understanding. See also above n 76. 102. [2000] 2 All ER 117 at 122.

Classifying precontractual liability: a comparative analysis 173 say that the appellant’s actual expectations under the ‘unenforceable’ arrangement with the first respondent were fulfilled.’”7 As this brief summary shows, the imposition of precontractual obligations may result in the protection of a range of different interests; but, despite this variance and irrespective of the separate doctrines utilised to impose liability, cases of precontractual liability are unified by the existence of the three common elements described above. Further, in describing these elements, principles from within both contract law and tort law were drawn upon. Arguably, the elements of precontractual liability could be said to constitute a sufficiently coherent set of principles justifying the recognition of one cause of action for precontractual liability.lW The alternative view, consistent with the present approach of the courts, is that there are merely several separate doctrines which operate in similar factual circumstances raising similar problems; none the less, such doctrines, it is suggested, are unified by the

103. The English Court of Appeal stressed, however, that it was not enforcing the parties’ unenforceable bargain (nor making for them some bargain which they had not themselves made: at 140), but rather that ‘the equity is invoked where the respondent has acquired property in circumstances where it would be inequitable to allow him to treat it as his own; . . . It is invoked because there is no bargain which is capable of being enforced’ (at 140). Consistent with this view is the fact that although the appellant did get the share of the property it had bargained for under the unenforceable agreement, it did so without the associated rights and obligations attaching to the expected joint venture agreement. Consequently, even if we cannot say that the appellant’s actual expectations were fulfilled, at the very least its reasonable expectations were. 104. One objection to recognising a single cause of action for precontractual liability (eg by recognising a general duty to negotiate in ‘good faith’ - cf German law discussed below) is that such acause of action must of necessity be stated in general terms and thus requires the application of a range of different standards of conduct in order to solve the range of possible problems. For example, as is evident from German law considered below, precontractual duties giving rise to liability culpa in contrahendo encompass at least two different standards of conduct, namely good faith or honesty, and reasonableness or due care. Another objection which may be made to not maintaining a distinction between the various doctrinal mechanisms or causes of action which are utilised in the precontractual context is that each individual action gives rise to specific and diflerent remedial outcomes. For example, McKendrick has suggested that contract law is often an inappropriate mechanism because of the ‘relative generosity’ of the expectation based-remedy (above n 10, p 17 1). This is misleading, however, for as already noted above, contract law nced not be used necessarily to enforce the anticipated contract itself; it can also be used to enforce an ancillary contract in relation to preliminary services performed. And in either type of contract, gaps as to price may be filled by resort to reasonable terms; consequently, the remedy granted is a quantum meruit for the reasonable value of the services,the same remedy which usually follows from liability imposed in restitution. Further, contractual relief can also be limited to compensation of reliance losses. As Famsworth (above n 20, p 25) has pointed out, if parties agree to negotiate in good faith, than the parties will generally perceive such an agreement precisely as protecting such reliance interest should one party pull out of negotiations, rather than as taking away a party’s right to withdraw. It should be added that even if separate doctrines do give rise to different remedial responses, they ought not to be given disparate treatment if certain common elements necessary for relief can be identified, allowing for recovery in one or other of the doctrines, that is, if the different remedial responses are triggered by the same causative events or facts and are thus explicable on the same basis.

174 Legal Studies existence of the common elements of liability considered above. In either case, the question needs to be posed: where does the cause (or do the causes) of action of precontractual liability fit? How do we categorise cases of precontractual liability within our law of obligations? Of course, some of the cases considered fall within contract law p r ~ p e r , " and ' ~ tort law has also been utilised to resolve precontractual problems. Liability in the case law goes beyond such examples, however, and as useful and important as classification may be, we must avoid trying to fit all cases of liability within one of these two recognised categories.'" Instead, it will be suggested that cases of precontractual liability draw on principles of liability of both categories of contract and tort, and lie somewhere between the two. This is not as radical a suggestion as it may at first appear to be, for two reasons: first, as is evident from the above discussion, the judgments in the cases themselves draw on ideas both from contract and tort law; and, secondly, categorising precontractual liability as a cause or causes of action between contract and tort, lying within neither category, but drawing on ideas from both, is a notion that has been argued to apply in the context of civil law. It is thus proposed to explore this idea further by reference to civilian legal systems, specifically German law. The purpose of this discussion is not to suggest that merely because German law recognises precontractual liability rules as lying between contract and tort, that the common law for that reason alone should also do so. Rather, the discussion will demonstrate the considerable advantages of such a conceptualisation for solving problems of precontractual liability; advantages which flow from a recognition that drawing rigid lines between the categories of contract and tort does not reflect the way concepts from both categories are utilised in the case law.

6 PRECONTRACTUAL LIABILITY IN GERMAN LAW:'"' CULPA IN CONTRAHENDO (a) The operation of the doctrine The notion or doctrine of liability for culpa in contruhendo, or breach of duty prior to contract, is of considerable importance in imposing liability upon parties in a range of different circumstances. The doctrine pre-dates the German Civil Code (Biirgerliches Gesetzbuch or BGB) and it has been developed largely independently of Code provisions by the German courts, in part relying on the 105. Eg ButlerMachhe Tool Co Ltd 1' Ex-cell-0 Cotprution (England)Ltd [ 19791 1 WLR 401, [ 19791 1 All ER 965. A less formal, more inclusive view of contract law based on a common core concern with the assumption of obligations - consider, eg, S J Stoljar's discussion of 'loose agreements' in The Law ofQuasi-Contract (Sydney: Law Book Co, 2nd edn, 1989) pp 192-196, 239-245, and 'Estoppel and Contract Theory' (1990-1) 3 JCL 1 -could solve some, but clearly not all, problems of 'precontractual' liability. To the extent that the courts enforce ancillary or 'more modest' contracts than those anticipated, we can still refer to such cases as imposing precontractual liability. See also Dietrich, above n 9. pp 136-137. 106. The third possibility, that of unjust enrichment, it has been argued, does not have any useful explanatory function in this context. 107. The doctrine of culpa in contruhendo also exists in Austrian and Swiss law, though there are, of course, differences in its operation in each legal system.

Classifying precontractual liability: a comparative analysis 175 considerable academic development of the concept. Although there is debate as to where cases of liability for culpa in contrahendo appropriately belong,Iox it is clear that some cases of liability for culpa in contrahendo are closer to liability in mainstream tort, and others closer to liability for breach of contract; and it has thus been argued that culpa in contrahendo is a ‘general ground of liability which opens a third lane (dritte Spur) between contract and tort’.l”’ The relationship between parties who have entered into contractual negotiations”” (or who merely prospectively may enter into contractual negotiations)”’ gives rise to comprehensive duties of care and of good faith, breach of which duties gives rise to liability to compensate for loss (including pure economic loss) incurred by parties to the negotiations. The nature of such duties will be considered in more detail below in the limited context of failed precontractual negotiations; it must be noted for now, however, that since these special duties are not owed to the public in general, they go beyond ordinary tort liability; and since they are not dependent on successful contract formation, they are not considered to be based on contract proper either,”’ at least in the view of many scholars. It should be noted, however, that although such a view has considerable support and has been persuasively argued by Canaris, for example,’” it has not gained universal acceptance amongst German scholars and courts, and some seek to describe all cases of culpa in contrahendo in terms of modified contract principles.Il4

108. See eg the summary of the state of affairs in German and Swiss law in D Medicus ‘Die culpa in contrahendozwischen Vetrag und Delict’ in P Forstmoser, H Giger, A Heini

and WR Schluep (eds) Festschriftfur Max Keller (Zurich: Schulthess, 1989). As P Schlechtriem Schuldrecht, Allgemeines Teil (Tubingen:Mohr Siebeck,3rd edn, 1997)p 17, has said, a number of ‘differingtheories as to the doctrinal classification of such liability have been put forward’. 109. B S Markesinis, W Lorenz and G Dannemann The German Law ofobligation (Vol I The Law of Contracts and Restitution: A Comparative Introduction) (Oxford: Clarendon Press, 1997) p 64. One of the foremost advocates of such a view i n German law is C- W Canaris ‘Schutgesetze- Verkehrspflichten- Schutzplichten’ in C-W Canaris and U Diederichsen (eds) Festschrififur Karl Lurenz (Munich: C H Beck, 1983) p 27. 110. More broadly, the parties must have entered some relationship with a view to a legal transaction (‘Aufnahme von rechtsgeschaftlichemKontact’: H Koziol dsterreichisches Hufpjlichtrecht Vol II (Vienna: Manz, 2nd edn, 1984)p 7 1. 111. Where a potential customer,perhaps even an unwilling purchaser, enters a shop, the precontractual duty may arise. See Medicus, above n 108,p 21 1. Note, however, that cases such as these are not of particular interest here, since they are best seen as ‘rectifyingthe deficienciesof the German law of torts’ (Markesiniset al, above n 109,p 66) in particular, by expanding the limited form of vicarious liability applicable to torts ($831 BGB) and replacing it with the broader vicarious liability provisions applicable to contracts. 112. Markesinis et al, above n 109, p 64. 113. See above n 109. 114. For references to the literature adopting differing classificatory approaches, see egschlechtriem, above n 108, p 13. See also S Schmidt Der Abbruch von Vertragsverhandlungen im deutsch-schweizerischen Handels-und Wirtschaftsverkehr (Konstanz:Hartung-Gorre, 1994)p 88, suggesting that the decision of the Bundesgerichtshof (BGH)of 22 February 1989, JZ 1991 199 does not support culpa in contrahendo liability on the basis of a ‘thirdlane’ between contract and tort.

176 Legal Studies Culpa in contrahendo has been utilised in a range of different situations. As Schmidt has said, culpa in contrahendo ‘applies to a very heterogeneous group of cases, and it thus deals with “problem areas of considerably varying origins”’.’15Many of the problems addressed by culpa in contrahendo are not of direct relevance to this article. For example, one stream of cases imposes liability for personal injury and property damage which, at common law, fall within tort law proper. German law, however, has utilised precontractual special duties in these cases in order to remedy perceived defects of tort provisions in the BGB, imposing only a limited vicarious liability upon employers.”6 Thus, culpa in contrahendo allows for the application of the broader vicarious liability provisions governing contractual obligations’” to precontractual situations, such as where a client has entered a shop and is injured by the negligence of the defendant’s employee. Other groups of cases of culpa in conrrahendo arise where a contract is concluded, but in unfavourable or burdensome terms.”* In such cases, precontractual duties to provide adequate information compliment the contractual obligations which may arise. Again, such cases are not of relevance to the subject matter of this article.’” Instead, of particular interest here are cases of liability for culpa in contrahendo arising as a result of incomplete or defective contracts, such as where a contract is defective for a failure to satisfy legal requirements as to form (‘nicht beachtung einer gesetzlichen Formvorschrift’);’?”or where a contract is void as being contrary to public policy and a defendant bears some responsibility for the circumstances invalidating it;’” or where contractual negotiations have failed. A brief outline of the operation of culpa in contrahendo in this third context, the focus of this article, will be given in order to illustrate the essential similarity of the outcomes of civil law decisions to those of the common law,

115. Schmidt, ibid, p 59, quoting U Huber GufcrchtenI, p 647 (my own translation). 116. See $831 BGB. In essence, employers can escape vicarious liability under this provision if they can show that they have hired competent staff and carefully supervised them. The position is slightly different under Austrian law: see $ 1315ABGB (Allgemeines Burgerliches Gesefzhuch),which provision is both narrower and broader than $831 BGB; broader in that there is strict liability of employers if they employ ‘unfit’ or knowingly employ ‘dangerous’ persons (ie liability cannot be avoided in such cases even if employers have taken all reasonable steps to safeguard others);narrower in that should employees not fall within these two types (and the onus of proof of such remains with plaintiffs), there will be no vicarious liability merely for failure on the part of employersto supervisetheir employees carefully. 117. See $278 BGB, and $1313(a) ABGB. Under these provisions, vicarious liability extends to liability for any acts committed by both employees and independent contractors and employers can not exculpate themselves by demonstrating their employees’ competence. 118. For a summary of such cases, see Medicus, above n 108, p 215. 119. At common law, such situationswould be covered by the law of misrepresentation, duties of disclosure and related concepts. 120. Medicus, above n 108, p 214. 121. Cf Schlechtriem,above n 108, p 15, and see Medicus, above n 108,pp 213-214 and U Eisenhardt ‘Anspriicheaus culpa in contrahendo wegen Verletzung der Verpflichtung uber erkennbare Unwirksamkeitsgriinde aufzuklaren’ in H G Leser and T Isomura (eds) Weg zum Japanischeiz Rechf: Fesfschr@fur Zenraro Kitagawa (Berlin: Duncker & Humblot, 1992)p 297.

Classifying precontractual liability: a comparative analysis 177 despite the seemingly different starting premises of each legal system. This outline of liability for failed contractual negotiations will lead into a discussion of the categorisation of such claims as lying between contract and tort, and the interesting consequences which flow from such a categorisation, particularly for our understanding of precontractual liability at common law. Important to note at this point is that, despite the well-developed German law of unjust enrichment, an unjust enrichment claim will rarely be available in the context of failed contractual negotiations, since a plaintiff will usually not have carried out a Leistung, that is, a conscious transfer of something intended to, and which actually does, enlarge the defendant’s estate; and other possible grounds for an unjust enrichment claim will generally not be present.

(b) Failed contractual negotiations Where one party negotiating a contract with another breaks off negotiations, such a party may be liable to the other party, usually for reliance losses suffered and even though no final agreement has been reached on the terms of the contract, if certain requirements of a claim for culpa in contrahendo can be satisfied. Speaking very generally, these requirements appear to be as follows.

1. Parties have entered into a ‘precontractual’ relationship, having entered into negotiations or dealings with a view to a legal transaction. Consequently, the parties are in a closer relationship (than before negotiations commenced), in which each is more vulnerable to the other, but also has a correspondingly greater capacity to effect the (economic) interests of the other (detrimentally) (‘gesteigerte Einwirkungsmoglichkeit’).

2. One party negligently or intentionally creates the expectation in the other that a contract will be (and not merely might be)”4finalised in due course. Such intention or negligence will be established in circumstances in which that party knows or ought to have known that that expectation could not be realised.lZ5 There is some debate as to whether the defendant’s culpability lies in awakening the expectation where there is no sound basis for a final agreement being concluded, or else whether the defendant’s culpability lies in the subsequent breaking off of the negotiations without good cause126(see (4)below) once such a culpable creation of the expectation has been awoken. The issue seems a moot point, however, since both the creation of an expectation and its subsequent

122. Apart from unjust enrichment arising from a plaintiffs Leisrung, 98 12 BGB provides for a claim for unjust enrichment where someone has received something ‘in any other manner’; this has generally been interpreted to include three possible sub-categories: the Eingriflskondiktion,RuckgrifJkondiktionand Verwendungskondiktion.For an explanation of these, see eg R Zimmermann and J du Plessis ‘Basic Features of the German Law of Unjustified Enrichment’ [1994] RLR 14. 123. Cf H Koziol ‘Delikt, Verletzung von Schuldverhaltnissen und Zwischenbereich’ JB1 1994 209, p 213. 124. Cf decision of BGH of 22 February 1989, JZ 1991 199. 125. WF Ebke ‘Report of the Federal Republic of Germany’ in Formation ofconfructs undpreconrractual liability (International Chamber of Commerce, 1990)p 42. 126. See Schmidt, above n 114, p 83ff, and references therein.

178 Legal Studies falsification need to be established in order for liability to arise, and arguably, culpable conduct may occur in the combination of these two factors.”’ German law12xmay go even further. It has been suggested that a breakdown in contractual negotiations may give rise to liability culpa in contrahendo even in the absence of prior culpable conduct. This view has gained the support of the Bundesgerichtshof (BGH) in its decision of 22 February 1989 (the ‘newspaper case’): liability can arise as a result of the falsification of expectations where negotiations are broken off without good cause. It is not necessary, therefore, to show any culpable creation of the expectation. The decision is considered further below, where it will be seen that the outcome of the case is hardly controversial. None the less, the reasoning of the BGH has been criticised. For example, Gunst has argued that the decision goes further than the previous legal position and amounts to a considerable broadening of the risk of incumng legal liability in the precontractual realm.”’ It should be noted, however, that the BGH’s formulation of liability in the newspaper case, without any requirement of fault in the creation of expectations, may not be as significantly different from previous formulations as at first seems. For it is still necessary to show a legitimate or justifiable expectation (‘berechtigte Vertrauen’) on the part of the plaintiff that a contract will be finalised. It is difficult to conceive of fact situations in which such an expectation can be established without words or conduct on the part of a defendant which, objectively seen, can reasonably give rise to such an expectation; and further, engaging in such conduct where the conclusion of a contract is not a certainty (bamng some good cause justifying withdrawal), must thus surely be, at the very least, unreasonable (that is, negligent). This requirement for a claim for culpa in contrahendo may also be made out if a party creates merely an expectation in another that potentially worthwhile negotiations will take place and that such party subsequently does not inform the other when the negotiations cannot take place or would be pointless.130

3. The plaintiff has been induced to rely on the expectation by incurring expenses or perhaps foregoing opportunities in a way which proves detrimental once no contract is finalised.I3’ 127. A comparison can be made with the requirements for establishing liability on the basis of estoppel i n common law systems: a defendant’s conduct must, in all the circumstances,be unconscionable, and such unconscionability can be established both by reference to the manncr in which an expectation was created and the circumstances of and reasons for its subsequent falsification. See Thompson I’ Palmer ( 1 9 3 ) 49 CLR 507 at 547, and cf Mason, above n 2 1, p 92. 128. It is not clear whether Austrian law goes so far: cf Koziol, above n 1 10, pp 76-78. 129. JZ 1991 202, p 203. Schmidt has pointed out that if liability can be established without any fault on the part of the party breaking off negotiations, then it would be more accurate to speak of contruhendo sine culpa, rather than culpa in contrahendo (above n 1 14, p 91, and see references therein for other criticisms). 130. For example, because the object of the intended negotiationshas already been sold. This covers the type of problem arising in the example considered below text to nn 184-1 88. The example is similar to one given by K Larenz in Lehrbuch Des Scliuldrechrs (Munich:C H Beck, 12th edn, 1979)p 92 and earlier editions,though Larenz does not give any authority for his inclusion of this example as a typical case of liability for culpa in contrahendo. 131. Or no negotiations take place (to cover the possibility considered in the paragraph immediately above).

Classifying precontractual liability: a comparative analysis 179 4. The defendant breaks off contract negotiations without good cause or justifiable reason (‘ohne triftigen Grund’ or ‘rechtfertigenden Grund’), having failed to warn the plaintiff of the possibility of no contract being completed in sufficient time to avoid any detrimental reliance on the plaintiff‘s part.I3*It is not entirely clear as to what exactly will constitute a good cause for breaking off negotiations, though it seems that any grounds which would justify rescission of a completed contract will also justify refusal to conclude a ~ 0 n t r a c t . IIt~is~ probably fair to surmise that a good cause for breaking off negotiations will generally be made out where a contract fails due to any impediment originating from the plaintiff‘s sphere.”? This requirement, at first blush, seems most at odds with the common law position stressing a right to break off negotiations for any reason whatsoever. As I concluded above, however, even at common law, the reason for the failure to complete a contract is of significance in determining whether a plaintiff‘s detrimental reliance in the circumstances which have in fact occurred comes within the scope of the risk which has been assumed by or which ought to be imposed upon the defendant.13sIt will be a rare case indeed, where a defendant’s conduct is such as to amount to a guarantee to meet the plaintiff‘s costs in all eventualities including, for example, the plaintiff‘s own (unreasonable) refusal to complete the contract. The converse of this, even at common law, is that a defendant can generally be said only to bear the risk of a contract failing without good reason.13h One criticism that has been made of the requirement that negotiations not be broken off without good cause is that a party is thus obligated to complete a contract he or she does not wish to enter.I3’ It must be pointed out, however, that breach of this duty generally does not lead to the fulfilment of a plaintiff‘s expectation or positive interest, that is, completion of the anticipated contract; instead, compensation of reliance loss is the usual remedy and thus the duty is satisfied by compensating the plaintiff for his or her foreseeable and unnecessary expenses incurred on the basis of the expectation of a final contract. A defendant can be required to meet this more limited duty, and still be otherwise free to break off negotiations.

(c) The scope of precontractual liability under German law When set out as above, the requirements for establishing a successful claim for culpa in contruhendo under civil law do not appear to differ significantly from 132. Cf Schmidt, above n 1 14, p 84. Similarly, this will be satisfied if aplaintiff has not been given a timely warning that no negotiations will take place. 133. See Koziol, above n 110, p 78. 134. Cf Markesinis et al, above n 109, p 70. 135. See above text to nn 88-94. 136. Cf also Mason, above n 21, p 83: a duty to negotiate in good faith will need to be confined to situations ‘in which the relationship of the parties is such as to generate a reasonable expectation that a party will not withdraw for capricious, arbitrary or bad faith reasons’. 137. Eg Gunst, above n 129, pp 204-205, and note the BGH decision, NJW 1996,1884, 1885,that such an obligation ‘to compensate reliance loss nevertheless signifies an indirect compulsion to conclude a contract’.

180 Legal Studies the requirements for establishing a claim for precontractual liability at common law. This is despite the seemingly very different starting positions from which each legal system’s rules have developed: the common law is generally said to start from the premise of no duty to negotiate in good faith and a freedomfrom contract, that is, that one has the right to break off negotiations at any time and without justifying reasons; whereas the civil law commences with duties of good faith and care owed to negotiating parties arising from the precontractual relationship.’’xYet the detailed rules defining a breach of duty under the civil law have a very similar content as the exceptions to the ‘no duty’ rule at common law. As Kessler and Fine have argued in comparing culpa in contrahendo (in its broader operation) with the common law: ‘Although the cases do not use the term culpa in contrahendo, its underlying philosophy of responsibility for “blameworthy” conduct has found expression in numerous ways.’l’’ Reflecting these similarities in the formulation of precontractual liability rules, it has also been pointed out that the results achieved in the German courts are not at great variance to the results of common law cases.14”A brief consideration of civil law decisions will bear out this conclusion. Perhaps a good starting point is the 1989 newspaper case,lJ1an important decision precisely because its reasoning appears to extend liability to the boundaries of accepted principle under German law. None the less, the result reached in the case is uncontroversial. The facts were as follows. The plaintiff was negotiating with the defendants for the sale of two newspapers published by the defendants. Approval of the purchase by the plaintiff‘s parent company was necessary before a contract could be finalised and, pursuant to gaining such approval, a ‘detailed offer’ from the defendant seller was sought by the plaintiff. The defendants provided such an ‘offer’ on 3 March 1986. On 7 April 1986, the parent company gave its approval to the plaintiff to go-ahead with the purchase. The plaintiff informed the defendants of this approval on 10 March 1986. Consequently, the plaintiff’s lawyer drafted a notice of intended purchase required by a regulatory body, and prepared a draft agreement. After a further draft agreement was prepared by the plaintiff‘s lawyers, incorporating amendments made by the defendants, a meeting was held on 18 April 1986 at which the parties reached agreement on the outstanding essential terms of the contract. Consequent upon this meeting, a third draft of agreement was prepared by the plaintiff‘s lawyers. On 28 April 1986, the plaintiff informed the defendants in writing of its intention to accept their ‘offer’ of 3 March 1986. On the same day, the defendants replied by telephone that they were no longer prepared to sell. The plaintiff proceeded to claim nearly DM 106,000 in damages, consisting mostly of legal fees and accountants’ fees. 138. Cf Cohen, above n 2 1, pp 30-3 1. 139. F Kessler and E Fine ‘Culpuin Conrruhendo,Bargaining in Good Faith, and Freedom of Contract: A Comparative Study’ ( I 964) 77 Harv L Rev 401, p 448. 140. As discussed above, considerable inroads have been made into the ‘no duty’ rule, by

application of rules from contract, estoppel,restitution and tort. Cf Markesinis et al, above n 109, p 70. See also G Kiihne ‘Reliance, Promissory Estoppel and Culpa in Contrahendo: A Comparative Analysis’ ( I 990) 10 Tel Aviv University Studies in Law 279, 294-295. 141. BGH JZ 1991 199.

Classifying precontractual liability: a comparative analysis 181 The BGH held for the defendants. According to the court the document of 3 March 1986 did not constitute an offer capable of grounding a complete agreement upon its acceptance. Further, the BGH denied the plaintiff's claim on the basis of culpa in contrahendo. The court noted the principle that 'generally, each party has to carry its own expenses incurred in the course of negotiating a prospective contract. The risk that the contract does not later ensue and that the expenses incurred thus prove to be wasted, lies with each negotiating party.''42 In order for the risk to be shifted, the plaintiff needed to show more than merely an expectation that the contract might be concluded, given that agreement had been reached on outstanding matters. Instead, the plaintiff had to show that the defendants had induced the belief the contract would, with certainty, be finalised. Since the costs had not been incurred on the basis of such a belief, they were not recoverable.'" It should be noted that this requirement of a belief that a contract will be concluded, with certainty, may go even further than the common law cases, which appear merely to require that there be a legitimate expectation that a contract will be concluded in due course.lWAlternatively, it may be that the required expectation or belief varies according to the nature of the reliance of the plaintiff. Perhaps a 'lesser' expectation may suffice, for example, if a defendant has requested specific, onerous tasks14sor if the plaintiff has acted in part performance of the anticipated contract (with the defendant's k n ~ w l e d g e ) ;whereas '~~ the expectation induced by the defendant may need to be one as to a certain contract where the plaintiff has incurred (albeit foreseeable) preliminary expenses of his or her own accord. The circumstances in German law in which a sufficient basis exists for a plaintiff to rely on an expectation induced by the defendant are not dissimilar to the typical cases of precontractual liability at common law.I4' For example, a plaintiff may have been encouraged to perform (or commence to perform) the very work required by the anticipated contract;'" or the parties have agreed on all essential terms and only some formality requires to be completed;149or a plaintiff has made improvements to the defendant's property in expectation

b i d , 200 (my own translation). See also BGH NJW 1996, 1884. Ibid, 202. See above nn 3-5 and text thereto. Perhaps BGHZ 92, 164, and cf Sabemo Pry Ltd v North Sydney MC [ 19771 2 NSWLR 880. One can thus imply an assumption on the part of the defendant to pay for the services should the contract not be finalised. 146. Cf BGH MDR 1961,49. 147. For an extensive outline of relevant decisions of the German and Austrian courts, see R Ostheim 'Zur Haftung fur culpa in contrahendo bei grundloser Ablehnung' JB1 1990, 522, 570. 148. See eg BGH MDR 1954, 345 (discussed in Ostheim, ibid, p 571). 149. See eg BGH NJW 1975, 1774; cf Wulrons Stores (Interstate) Lrd v Muher (1988) 164 CLR 387. 142. 143. 144. 145.

182 Legal Studies

of some interest therein,I5" in cases similar to those covered by proprietary estoppel at common 1aw.l5' Perhaps one difference between the civil and common law approaches to precontractual liability can be seen in relation to the remedial response which follows such liability. The civil law generally protects against reliance losses, aiming to restore plaintiffs to their status quo before the detrimental conduct. The remedy often awarded at common law, however, a quantum meruit for services rendered, not only protects a plaintiff's reliance loss but may also give effect to the reasonable expectations of the plaintiff (or even both parties) since the parties' intentions may be taken into account in determining the 'reasonable' value of the services rendered.Is2 Further, a plaintiffs actual expectation under the contract anticipated at times may be enforced, often by resort to the estoppel doctrine. Even these differences, however, may not be as significant as they at first seem. It is accepted in German law, for example, that in exceptional cases involving contracts failing to comply with required formalities, at least, specific performance may be awarded,Is7or that damages may be calculated on the basis of the plaintiff's expectations, so that his or her contractual or 'positive' interest is thereby protected.'5JSome legal writers also support the possible award of a

150. See eg BGH NJW 1996,1884, and the BGH decision of 19 October 1960,in Ostheim, above n 147, p 571. 151. See above text to nn 82-87. Similarly, Austrian case law does not appear to extend precontractual liability beyond that which exists at common law and the facts of cases granting relief are often comparable to those at common law. A detailed summary of Austrian (and German) decisions is provided by Ostheim, above n 147. These include SZ 7/66, in which a plaintiff gave up his existing job upon repeated assurances of obtaining a position with the defendant. Cf Brewer v Chrysler Ccinada [ 19771 3 WWR 69; and H u m a n v Red Owl Stores 133 NW 2d 267 (1965). See also SZ 27/120, in which the facts seem even closer to those of Brewer 1' Clirysler Curiada and yet recovery was denied. The Austrian courts also consider the question of who has taken or assumed the risk of a particularexpense or loss in determining whether liability arises. For example, in the Oberstes Gerichtshofs (OGH) decision JB1 1977,315, the plaintiff, a prospective purchaser of a house, indicated to the vendors his intention to sell gold coins in order to finance the purchase of the house. He proceeded to sell the coins the following day, despite the fact that the parties had not agreed on a purchase price and that the plaintiff's latest offer fell ATS 20,000 short of the previously rejected offer. The defendants later decided to withdraw the house from sale. The plaintiff unsuccessfully sued for losses incurred in the sale of the gold. The Court stressed that he had made the sale entirely at his own risk. There was no duty on the part of the defendantsto warn the plaintiff against his intended actions as there existed no expectation of a contract being finalised. In any case, the plaintiff had merely indicated the possibility of his selling gold coins as a means of financing the as-yet-unagreed purchase price. See Ostheim, above n 147, p 530. One much criticised case which seems to go further than the common law is OGH JBI 1990,127, in which the OGH allowed an unsuccessfuljob applicant to recover the costs incurred in attending an interview which he had been asked to attend. 152. Even if this is not the case, quantum meruit may incorporate a reasonable profit component. 153. See Markesinis et al, above n 109, p 68, citing the 1967 decision, BGHZ 48,396. 154. See eg BGH NJW 1965,812.

Classifying precontractual liability: a comparative analysis 183 plaintiff‘s expectation interest even in cases of failed contractual negotiations, but the case law does not appear to have gone this far.155

7 CLASSIFYING CULPA IN CONTRAHENDO: A COMPARATIVE ANALYSIS Having set out briefly the requirements of liability arising in culpa in contrahendo and suggesting that the scope of precontractual liability is not significantly different in civil or at common law, I want to turn to one of the significant issue engaging civil law jurists, namely the question of the categorisation or classification of culpa in conrruhendo within the law of obligations. It has already been noted that culpa in contrahendo is sometimes described as a third lane between tort and contract; this idea has been persuasively and influentially argued by C a n a r i ~and ’ ~ ~adopted and developed by other academic writers. Koziol, for example, has put the argument in the following terms: ‘Liability based on tort and liability based on breach of contract usually are taken as clearly separated contrasts. But I think that they are the two ends of liability based on fault and that between them there is a connecting chain of intermediate stages. This understanding is important because one has not to sort liability in one of these two categories and, therefore, is able to avoid abruptly different treatment of rather similar cases. In the result, this has already been done to some extent, eg, by accepting special duties of care prior to conclusion of contract [culpa in ~ontrahendo].”~’ This encapsulates the essence of Koziol’s argument, developed in his article ‘Delikt, Verletzung von Schuldverhaltnissen und Zwischenbereich’, in which he states: ‘If one considers the distinction between tort and (breach of) contract, then it is recognisable that the basic principles of each only fully apply at the core of each category. In a relatively wide intermediate area, only some of the factors determinative of liability [in each category] may be present, and sometimes present fully or only to a lesser extent. Consequently, rather than seeing (liability in) tort and contract as sharply delineated, incompatible opposing categories, they should instead be seen as legally regulated core areas between which there is a fluid transition.’I5* 155. See references in Koziol, above n 110, p 79, who points out that in Austrian law such a remedy would run counter to 4861 ABGB. It may also be difficult to establish that the loss of expectation was caused by a defendant’s breach of duty. In other words, if the defendant had made the plaintiff aware of the impediments to a contract being concluded, then in most cases of failed negotiations at least (though not necessarily in cases of contracts unenforceable for failure to comply with formalities),the plaintiff would have been able to avoid the reliance losses, but would have been unlikely to have concluded the contract successfully. 156. Above n 109. 157. See H Koziol ‘The Borderline Between Tort Liability and Contract’ in H Koziol (ed) The Unijkution ofTort Law: WronRfulness (The Hague: Kluwer Law International, 1998), p 25 (footnotesomitted). 158. Above n 123, p 214 (my own translation).

184

Legal Studies

The point is not, it must be stressed, that in a given case ‘concurrent liability’15”may arise both for a breach of contract and in tort- an idea which is, of course, accepted at common law. Rather, the point is that liability may arise in a given case as a result of the operation of rules (or causes of action) which cannot be classified as lying in either category (and an action brought for breach of contract or in tort would therefore fail)L60but as lying in between these categories and drawing on principles of liability from both. Sometimes liability may be based more on fundamental principles at the core of contract law (and can thus be described as more contract-like or ‘vertragsahnlich’),161at other times, liability may be based more on fundamental principles of liability at the core of tort law (and can thus be described as more tort-like). Importantly, it is the existence of liability producing factors of both categories which in combination are sufficient to substantiate liability in some form. It is not intended to suggest, therefore, that these are two distinct sub-groups of liability. In this intermediate area between tort and contract (‘Zwischenbereich’), there are numerous gradations or intermediate stages of liability (‘Zwischenstufen’), and any attempt to classify these intermediate stages themselves into rigid, conclusive categories, according to Koziol, should be avoided.lb2None the less, it may be useful to describe the differing stundurds which underlie and justify a determination of liability at the ‘contract’ and ‘tort’ ends of the spectrum respectively, as I have sought to d o in setting out the third element of precontractual liability at common law, above. That liability rules and doctrines may draw upon principles from both contract and tort law is not a new concept to the common law, and in particular, to equity. As I have argued previously: ‘Where a defendant’s conduct appears to have resulted in harm to another, causes of action in contract or tort will often be available to remedy the consequences of such conduct; but clearly, this will not always be the case, even where there is a perceived need for the imposition of liability in some form. Contract and tort are categories self-limited by certain historical and conceptual criteria. Gaps consequently arise and gap-filling rules and doctrines develop . . . The point is perhaps best illustrated by equity as a whole, in its metamorphosis and development of ideas already familiar to the common law as part of its well-recognised ameliorative function. For example, one quintessential and wide-ranging equitable gap-filling technique is estoppel, which operates at the periphery of both contract and tort and covers a diverse range of subject-matter . . . 159. S A Smith ‘ConcurrentLiability in Contract and Unjust Enrichment: The Fundamental Breach Requirement’ (1999) 115 LQR 245,248, points out that the term ‘concurrentliability’ is used in two different senses, namely: ( 1 ) that in respect of a particular set of facts, a plaintiff may satisfy the elements of two (or more) causes of action; and ( 2 ) that a breach of contract may also, as a possibility, ‘support’ or give rise to an action in tort, or vice versa. 160. It will be difficult to establish d l the elements necessary for aclaim in either category. 161. Cf Gunst JZ 1991 203. 162. Koziol, above n 114, p 217. Cf also Kessler and Fine, above n 139, p 449, writing in the context of ‘good faith’ and culpa in contrahendo as ‘residual’concepts encapsulating broad standards of fairness and justice: ‘The law confronts the task, in the interest of certainty, of identifying and categorising these “residual” concepts, only to be faced with the realization that this process is never-ending.’

Classifying precontractual liability: a comparative analysis 185 The point, very simply, is that estoppel does not operate in a doctrinal vacuum -it draws many of its ideas from the familiar categories of contract and tort . . . [Conduct which can give rise to an estoppel] can range from that which has a strongly contractual flavour, such as where an assumption “formed the conventional basis upon which the parties entered into contractual or other mutual relations”; to tort-like conduct, such as where imprudence or carelessness are the proximate cause of the plaintiff‘s reliance on an Other conduct may not be so readily classified as contract-like or tort-like (exhibiting elements of both, for example).’Ia The recognition of the existence of liability rules which are classified as being neither in tort nor in contract, but as drawing on ideas from both categories, may be particularly challenging to those who stress the need for classifying claims within clearly delineated categories. McKendrick, for example, quotes Zimmerman as concluding that culpa in contrahendo ‘falls squarely into the grey area between the law of contract and the law of delict, and there is much to be said for the proposition that it does not fit neatly into either of these, but rather forms an integral part of a third ‘track’ of liability. ’Ih5 McKendrick considers this to be to the doctrine’s discredit: that it is a ‘doctrine which lacks stability’ and that the ‘importation of a doctrine which straddles the contract/tort divide in such an uncertain fashion would do little to improve the current state of English law. Better to search out a doctrine which has secure foundations’.lMContrary to McKendrick, I would see these qualities of culpa in contrahendo a virtue. I would consider it a positive step if we openly recognise the existence of many ‘shades of grey’ of liability in our law, and if, in some cases at least, liability was recognised to exist in some intermediate area or ‘no-man’s land’’67partaking of different elements of tort and contract. It would be to recognise, in the words of Stapleton, that ‘our rules are not “ideal types” based on a single pure idea’.I6*If a more accurate description and explanation of the operation of liability rules can thereby be attained, we should not shut our eyes to the reality of such shades of grey.’” It is worthwhile to consider the words of Kessler and Fine, arguing in support of a notion of good faith in negotiating contracts: ‘In the language of Corbin, “certainty in the law is largely an illusion at best, and altogether too high a price may be paid in the effort to attain it.”170Indeed, we may argue that a more satisfactory degree of “certainty” is attained with

163. See Thompson v Palmer (1933) CLR 507 at 547, per Dixon J (as he then was). 164. Dietrich, above n 9, p 105. 165. McKendrick,above n 10,p 186, from The Law of Obligations-Roman Foundations of the Civilian Tradition (Oxford: Clarendon Press, 1996) p 245. 166. Ibid, pp 186187. 167. The slightly pejorative term is McKendrick’s,above n 10, p 163. 168. Stapleton, above n 98, p 27. 169. Cf Koziol, above n 123, p 215. 170. 3 Corbin $609, at 689 (revised edn 1960).

186 Legal Studies flexible notions of fairness eliminating the need to circumvent the inflexible application of mechanical rules.’ ‘’I The question that therefore needs to be asked is: Can a greater understanding of the common law of precontractual liability be gained from a conception of such liability as lying between the established and well-recognised categories of contract and tort? I would suggest that the answer to this question is yes.I7? Perhaps the most significant insight to be gained from a conception of precontractual liability as lying between contract and tort is that it explains why, though neither a cause of action in tort nor for breach of contract (nor in unjust enrichment) is made out, none the less liability may arise. This liability is established because some of the factors determinative of liability in each category are present, such that broader than usual contractual liability may be justified by further factors normally associated with establishing a claim in tort, or broader than usual tort liability may be established by further factors normally associated with establishing a claim in contract. It is important to remind ourselves of the fact that the concern is with work or services carried out precontractually; in other words, financial detriment incurred by a plaintiff. Tort law protects against pure economic loss only within narrow constraints: even foreseeable economic loss is not recoverable unless some further, closer relationship exists between plaintiff and defendant. Parties to a contract, however, are required to make good economic loss arising from a breach of the obligations assumed. In those cases of precontractual liability in which a tort-like duty to warn a plaintiff against foreseeable economic loss is imposed upon a defendant, the basis for such a ‘higher’ than ordinary duty is found in the near-contractual, close relationship of the parties, in which the defendant creates an expectation that a contract will be completed and allows the plaintiff to rely on such an expectation in a way which proves detrimental when it is not fulfilled. The closer, contract-like relationship existing between the parties gives rise to a corresponding increased vulnerability of each, in opening themselves up to the other in the course of negotiations. As has been pointed out, ‘commercial transactions can only function smoothly . . . if the increased risks [which contractual negotiations entail] are compensated with higher duties of care”77than that owed to the ordinary person. In other cases of precontractual liability at common law, liability stems from an intention (albeit one not ripening into a fully-formed contract) to assume the risk of the contract negotiations failing and, in particular, to assume an obligation to pay for particular work. Although contract law is the usual mechanism for assuming particular obligations, the law places fairly stringent conditions upon the assumption of contractual obligations, partly because

171. Above n 139, p 449. See also Mason, above n 21, p 70: ‘with many legal concepts rooted in formalism, [the] element of certainty [is] illusory.’ See also ibid, p 89. Contrast Cohen, above n 21, pp 52-53. 172. The appropriate classification of precontractual liability may also be significant for the purposes of determining questions of jurisdiction and choice of law in private international law, particularly given that many contractual negotiations take place in an international context, and different legal systems may approach questions of liability in such cases in different ways. This issue, however, is outside the scope of this article. 173. Koziol, above n 123, p 213, citing Welser (my own translation).

Classifying precontractual liability: a comparative analysis 187 parties to a contract ‘are one another bound to very extensive and severe duties”74 and, further, are generally entitled to protection of their full contractual expectations. However, even though parties may fail to complete their agreement or otherwise meet the requirements of contract law, they may none the less have conducted themselves in a way which evidences a sufficient intention to assume an obligation in some form. If a plaintiff detrimentally relies on the reasonable expectations created by such conduct, this further element justifies some measure (if not necessarily the full contractual expectation measure) of protection being granted to such a plaintiff. It should be noted at this point that, in civil law, the preferred view is that precontractual liability is an imposed obligation not dependent on any contractual intention, that is, culpa in contrahendo duties arise irrespective of or independently of any intent of the parties.’7sThis view may stem, however, from the no doubt correct observation that no final contract need be formed for a duty to arise and that a party’s expectations under the contract anticipated will not generally be enforced. None the less, it has been argued above that the parties may have formed a sufficient intention in relation to a lesser, preliminary contract (to pay for certain services or work) which realistically can be discerned from their conduct. Further, the civil law view of precontractual liability as an entirely imposed duty may also stem from that system’s emphasis upon a subjective theory of contract, requiring a meeting of the parties’ minds, whereas under the common law’s emphasis upon an objective approach to contract, liability can be founded on the intentions of a defendant as objectively evidenced by his or her c ~nduc t.‘~‘ I would suggest that a number of other advantages also flow from a recognition of precontractual liability as best classified as lying between contract and tort. Such a classification may explain the range of remedial rights which follow from a finding of precontractual liability at common law; and further, such recognition may give us useful insights into what the appropriate remedy in a given case should be. If rigid categorisation is undesirable in this context, so, it would seem, is an inflexible approach to remedy. For example, if the reasons for liability in a given case lie closer to the core of contract, indeed such that there very nearly is a contract, then a remedy protecting a plaintiffs positive or contractual expectations may be appropriate, as Banner Homes Group plc v Luff Developments Ltd177(discussed above) and Waltons Stores (Interstate) Ltd v ~ V f a h e r so ~ ’ ~aptly demonstrates. At the other end of the spectrum, where liability appears to be based on an expansion of core tort concepts, such as a negligent infringement of a duty of care to warn against certain risks, then only the compensation of (reliance) losses caused thereby would seem to be appropriate. If a given case falls somewhere in between these two extremes, perhaps the most appropriate remedy is the protection of what a plaintiff may reasonably be entitled to expect, consistently with many cases at common law. 174. Koziol, above n 157, p 25. 175. Koziol above n 110, p 75, and see also Medicus, above n 108, p 207, and Canaris, above n 109, pp 93-94. 176. See generally Kessler & Fine, above n 139. 177. [2000] 2 All ER 117. See text to nn 101-103 above. 178. (1988) 164 CLR 387.

188 Legal Studies A similar point has been made by civilian lawyers. It has been argued that the question of the scope of any liability for culpa in contrahendo depends on the harm caused by the failure to fulfil the particular duty required to be met in the circumstances. Schlechtriem, for example, has said: ‘Decisive should be, which interests the breached duty was intended to protect; the basic principle, that the scope of liability is limited by the protective purpose of the breached duty, also applies to the duties to advise and of disclosure in the context of precontractual obligations. The injured party is principally to be put in the position as if the duty had been correctly fulfilled. Defining these duties and stipulating in more concrete terms the protective reach of their content must . . . remain the responsibility of the judiciary.’ A number of factors may prove to be critical in determining liability in a given case, and the scope of that liability. Of primary relevance, it is suggested, will be the particular factual circumstances and relationship of the parties. For example, in some relationships, certain assumptions of risk may be implied in fact. Different relationships and situations may generate ‘circumstance-specific’ duties’*Oto conduct negotiations in particular ways; in other words, the content of precontractual duties will vary according to the circumstances. Although facts may well be ‘messy things’, we cannot avoid delving deeply into the facts of a given case in order to resolve issues such as the parties’ intentions and expectations which can be implied from their conduct and situation.’*’It should be added that particular relationships of negotiating parties, such as that existing between a public body or agency of government and a tenderer bidding for a government contract, may give rise to more onerous duties to engage in conduct of a high standard, implied by law as a legal incidence of the particular class of contract or of pre-contractual relationship.’** Also of relevance, in particular in determining the scope of any liability, may be matters of causation. For example, had a plaintiff been warned of the impossibility of a complete contract, or of its voidness as contrary to public policy, then such a plaintiff may well have avoided incurring reliance losses, but would not have been able to ensure that a contract would be concluded. An expectation-based remedy would in such a case be inappropriateLx3 because the defendant’s breach of duty did not cause the loss of the plaintiff‘s positive contract interest. 179. Schlechtriem, above n 108, p 20 (my own translation). 180. Cf ibid, p 18. 181. Cf Hedley, above n 3 1, p 196. Identifying and articulating the particular standards of conduct demanded of negotiating parties in order to solve particular factual problems should rebut any concerns, I would suggest, that broad principles such as a duty to negotiate in good faith and with care are a carte bkinche for judicial discretion. See also, generally, Stapleton, above n 98. 182. See eg Hughes Aircraft Systems International v Airservices Australia (1997) 146ALR 1 particularly at 38-42; and under German law, $97 Gesetz gegen Wettbewerbsbeschrunkungen (GWB, ‘Act Against Restraints of Competition’), which sets out general principles governing the procedures for the award of public contracts. Note the wide definition of ‘contracting entities’ to which the provisions apply ($98). 183. Cf Schlechtriem, above n 108, p 19. See also above n 155.

Classifying precontractual liability: a comparative analysis 189 Accepting that precontractual liability may arise as a result of the interaction of principles from both contract and tort does not mean that such liability rules cannot be described with sufficient certainty (as the elements of liability both at common law and in civil law set out above show), nor does it mean that such rules cannot be applied in a predictable way in solving new problems. This can be illustrated by reference to an example, taken from Markesinis, Lorenz and Dannemann: ‘Assume the defendant has invited the plaintiff to come, say, from London to Munich for contract negotiations on a certain date, the outcome of these negotiations being open. The plaintiff takes a plane, books a hotel and rents a car. It is only after arrival that the defendant notifies the plaintiff that he changed his mind and gave the contract to a third party two weeks before.’184 Markesinis et al, consider that German law would allow recovery for the plaintiff‘s wasted outlay in this case, on the basis of culpa in contrahendo, for the defendant’s negligent failure to notify the plaintiff to cancel the trip. They doubt, however, whether English law would allow recovery in this example ‘as there is no misstatement, no contract, and no enrichment’.lg5 Arguably, however, precontractual liability would here lie, even at common law, though the example is obviously a borderline case. Let us consider the principles from contract and tort law which might support such a conclusion. It is difficult to argue that in this situation, the defendant has manifested an intention to assume the risk of the plaintiff‘s loss, or expressly or impliedly agreed to pay for the trip, since the defendant has merely invited the plaintiff to the negotiating table. Such preliminary costs, like tendering costs, are ones which a plaintiff can expect to have to carry in most cases. The conditions for establishing a contract-like assumption of obligation would not appear to be satisfied. Nor does the plaintiff have any expectation that a contract will be concluded, though she does have a reasonable expectation of an opportunity to negotiate and that such negotiation may, in due course, lead to the formation of a contract. As noted above, tort law does not generally protect a plaintiff against even foreseeable pure economic loss: some further, closer relationship must exist between plaintiff and defendant justifying a duty of care. Here, we could argue that such a relationship is established by the invitation to negotiate. The plaintiff is known to the defendant as an individual and is particularly vulnerable to 184. Above n 109, p 71. 185. Above n 109, p 7 I , and see above n 130. Although the BGH decision JZ 1991 199 (the newspaper case) makes it clear that a belief that a contract will be finalised is usually required, and not merely ar. expectation that it might be, it seems clear that an expectation that potentially fruitful negotiations will, with certainty, take place will also suffice if foreseeable reliance losses are incurred in preparation for such negotiations (against which losses a defendant could simply have safeguarded the plaintiff by advising him or her that there is no longer any point to the negotiations). And see, under French law, the case cited in Farnsworth, above n 20, p 21. Even under German law, however, a conclusion of liability would be doubtful if the facts were slightly different and the plaintiff merely relied on an enticing advertisement for the sale of certain goods which, prior to the plaintiff incurring travel costs to attend the sale, were sold out. See Medicus, above n 108, p 213. It seems clear that in such a situation, a plaintiff who travels to the sale without first checking the availability of stock is taking the risk.

190 Legal Studies harm arising from the defendant’s conduct.Ix6Indeed, that vulnerability was created by the defendant’s invitation and the defendant has been careless, perhaps even grossly so,”’ in failing to inform the plaintiff during the two weeks as to the futility of her coming to Munich. Given the absence at common law of a tort-like duty of care akin to culpu in contruhendo, liability would have to be established within mainstream tort. But although it is possible to argue for a duty of care in tort, it should be noted that such a duty is greater than that ordinarily owed to one’s neighbours, namely, it is a duty to protect the plaintiff against incurring pure economic loss. And the reason for the imposition of such an enhanced duty of care stems from the precontractual relationship the plaintiff was invited to enter into by the defendant. The absence of any expectations of a contract being completed does not preclude an expectation that negotiations will not be futile and this gives rise to an obligation of due care from a defendant who has created and grossly disregarded the plaintiffs vulnerability to suffering unnecessary economic loss.’x8Consequently, the appropriate remedy will be limited to compensation of the plaintiff‘s reliance loss.

8 CONCLUSION From the above discussion, the following conclusions can be drawn. 1. Unjust enrichment theory has little to offer by way of explanation of cases imposing precontractual liability for services performed in anticipation of contracts which do not materialise. 2. Although the outcomes of the cases at common law are for the most part appropriate, a variety of doctrinal mechanisms have been utilised to impose precontractual liability in a ‘piecemeal’ fashion; none the less, such cases of liability exhibit certain common features or elements. It is desirable that the common elements unifying all such cases of precontractual liability be recognised. 3. These elements draw on ideas and principles from within both contract and tort in defining the basis and the scope of any liability. 4. The civil law doctrine of culpu in contmhendo, utilised to solve problems of precontractual liability on the premise of duties of care and good faith arising

186. Cf Perre v Aparrd Pty Ltd (1999) 198 CLR 180. 187. As to the relevance of the degree of negligence as a factor in determining whether liability for pure economic loss arises, see Perre v Apand P f y Ltd (1999) 198 CLR 180 at 328-329, per Callinan J; cf at 231, per McHugh J. 188. One of the reasons underlying the law’s reluctance to impose wide-ranging liability in tort for pure economic loss is the capacity for such liability to excessively infringe upon parties’ freedom to engage in commercial activities which, of necessity, have economic consequences. Contract law, however, protects against pure economic harms when a defendant has assumed an obligation in relation thereto. Hence, if the protection of pure economic loss is generally within the realm of contract law and not tort law, it is suggested that a duty in tort may arise if aspects of contract law are present, aspects which justify wider, more far-reaching obligations. Eg if A owes a contractual obligation to B not to do a certain act in a way which harms B’s economic interests, then there is no further infringement upon A’s freedom of conduct, if A does that something and is thereby held liable in tort to C as well. Cf Perre v Apnnd Pry Lrd ( 1999) 198 CLR 180.

Classifying precontractual liability: a comparative analysis 191 during contractual negotiations, gives rise to similar outcomes in the determination of cases as occur at common law. 5. Culpa in contrahendo is said to be classified as lying between contract and tort and drawing on ideas and principles from both. 6 . As with culpa in contrahendo, I would argue that at common law, a better understanding of the basis and scope of precontractual liability can be obtained if such liability rules are classified as lying between contract and tort and drawing on ideas and principles from both categories. Such better understanding will also allow us to more accurately predict future outcomes of disputes arising out of precontractual negotiations.