A. Can Contractual Rights be Converted?

The basic claim made by proponents of the Inadequacy Thesis is that contractual rights should be protected by the tort of conversion. A necessary presupposition of this argument is that a contractual right is in fact a “thing” capable of being converted.

David Hume famously remarked that, “the principal disturbance in society arises from those goods, which we call external, and from their looseness and easy transition from one person to another”.Footnote ¹⁰ Tangible things are the objects of in rem rights of exclusion actionable in conversion because they are, by virtue of their existence in the world, vulnerable to non-consensual transfer. To borrow Penner's explanation: “[t]he reason why a right to a material object would seem to entail something like a right in rem is simply that a material object, existing as it does in the world, is … in principle accessible by anyone and subject to the depredations of anyone.”Footnote ¹¹

Whilst contractual rights amount to “property” in the broad sense that they are economically valuable assets, many of which can be assigned,Footnote ¹² it does not follow that, like tangible things, they are therefore capable of being converted. This argument has been made by Douglas in an important article on the nature and scope of conversion.Footnote ¹³ His argument is that unlike chattels, which exist independently of any legal system, rights to contractual performance are artefacts of a particular legal system. Consequently, whilst the question of whether an owner has been deprived of his chattel by a thief is simply one of fact, the question of whether a promisee has been deprived of his right to performance is one of law, the answer to which depends on the rules of the legal system in which it is constituted and recognised. He explains that “a contractual right … is a purely legal construct. It is something which depends upon a legal system for its existence. As such, it is for the legal system to determine when such a right comes into existence and …when the holder of such a right is deprived of it”.Footnote ¹⁴

Although subject to important exceptions such as frustrationFootnote ¹⁵ and termination for breachFootnote ¹⁶ or repudiation,Footnote ¹⁷ the general position at common law, which will be referred to as the “Basic Rule”, is that contractual duties can only be discharged by performance rendered by the promisor to the promisee.Footnote ¹⁸ Consequently, it is generally impossible for third parties to “steal” or “convert” a promisee's right to performance because the promisor's duty to perform remains intact.

Douglas's point can be illustrated with the following example. A is an egg wholesaler who, in consideration of £1,000, agrees to sell B, an egg retailer, 1,000 eggs deliverable by a certain date. The contract provides that B need only pay on receipt of the eggs, but that title to the eggs is not to pass to B until A has been paid in full. C is a rival retailer who is aware of A's contract with B but, due to supply constraints, wishes to acquire the eggs for himself. Using his skills of impersonation, C dupes A into delivering the 1,000 eggs to himself instead of to B. In this example, C has plainly interfered with the fact of A's contractual performance. However, due to the operation of the Basic Rule, C's conduct has not discharged A of his duty to B, meaning that B's correlative right to performance survives.Footnote ¹⁹ It thus seems impossible to say that C has “converted” B's right to performance against A.

A potential response to this argument is to point out that liability in the property torts requires only interference with, and not deprivation of, an owner's right. If, for instance, A touches B's chattel or trespasses onto his land, A will have committed an actionable wrong against B even though he has not deprived B of his right to the land or chattels. This objection, however, misses the point. As Douglas explains: “When we talk of the conversion of a chattel … it is not the right, but the thing to which the right relates – the chattel – which the defendant deprives the claimant of.”Footnote ²⁰ For the purposes of an analogy with conversion, the contractual right takes the place of the chattel as the “thing” or object to which the property right relates.Footnote ²¹ Douglas's point is simply that extinguishment of the contractual right is the only jural event that is analogous to an act of physical interference with a chattel. If the jural structure of a purely jural object remains unaltered by the actions of a third party, then it seems impossible to say that it has been converted.

The proceeding argument can be summarised as follows. Scarce, tangible resources are the objects of in rem rights of non-interference because their physical existence in the world makes them vulnerable to the depredations of others. The corollary of this observation is that resources that are not vulnerable to third-party interference need not be objects of such duties. Because, as Douglas observes, “[t]he legal nature of a contractual right makes it generally impossible for the holder of such a right to be deprived of it”,Footnote ²² the attempt to transform contractual rights into objects of property for the purposes of an action in conversion seems redundant.

B. Loss Caused by Interference with Contractual Performance

Douglas's analysis of the nature of contractual rights suggests that the Inadequacy Thesis fails at the first hurdle. Why extend the ambit of conversion to a species of “thing” that is, by its very nature, incapable of being converted?

One answer is that, whilst Douglas's analysis vindicates the traditional view that one “owns” property and is “owed” obligations,Footnote ²³ it does not prove that the law adequately protects holders of valuable contractual rights. More particularly, it does not address the fact that a third party can cause a promisee real loss, even where the latter's contractual right survives. This can be demonstrated by altering the egg merchant example.

Recall that A was fraudulently induced into delivering eggs to C, a roguish third party, instead of B, his promisee. Because A's duty to B remains undischarged, we cannot say that C has converted B's rights. C's duplicity thus appears to be a problem for A, who must bear the costs of performing again, but not for B, who can either insist on performance or sue A for breach of contract. Imagine, however, that before being able to correct his embarrassing error, A became bankrupt. That B's rights remain enforceable against A's personal representatives does not alter the fact that C's interference with A's performance of his contractual duties to B has caused B loss. Assuming that there is no change in the market price of eggs, the quantum of this loss will equal the cost of organising substitute performance less whatever meagre sum B recovers from A's trustee in bankruptcy.

We thus have a situation in which C has caused B loss but, following Douglas's analysis, B has no action in conversion against C because his right to contractual performance remains enforceable against A's trustee in bankruptcy. Does this example demonstrate a lacuna in the law? The answer to this question is “no”, because the law already affords B a means of redress against C. Because C has induced A to breach his contract with B, C will be liable to B in the tort of inducing contract.Footnote ²⁴ Moreover, because C's impersonation of A amounts to an actionable deceit, C will also be liable to B for intentionally causing loss by unlawful means.Footnote ²⁵

As both courts and commentators have observed,Footnote ²⁶ the tort of inducing breach of contract blurs the boundaries between property and obligation. Nevertheless, the limited in rem effect created by the tort is perfectly compatible with the argument advanced above, which is simply that extending the tort of conversion is unnecessary because the law already recognises an action that enables a promisee to recover loss inflicted by a stranger in these circumstances.

Moreover, the quasi-proprietary status bestowed on contractual rights by the inducing tort does not provide a conceptual stepping stone for those who seek to extend conversion to contractual rights. This is so for two reasons. First, consistently with Douglas's argument, liability for inducing breach of contract does not presuppose that the promisee has been deprived of her right to performance by the actions of the inducer. This is illustrated by the locus classicus, Lumley v Gye.Footnote ²⁷ That the promisee, Benjamin Lumley of Her Majesty's Theatre, had an action against the inducer, Frederick Gye of Covent Garden Theatre, did not prevent him from successfully enjoining the promisor, Madame Wagner, from breaking her contract with him by singing at Covent Garden.Footnote ²⁸ Secondly, the liability of a converter of chattels is famously strict.Footnote ²⁹ By contrast, a plaintiff suing for inducing breach of contract must prove that the inducer intended to procure a breach of contract.Footnote ³⁰ The wrong of inducing breach of contract lies not in the inducer's interference with the promisee's right as a “thing”, but in his intention to induce the promisor to commit a breach of duty.Footnote ³¹ Conceptually, then, the tort's nearest relation is not conversion or trespass, but personal liability in equity for knowingly receiving property transferred in breach of trust.Footnote ³²

C. OBG v Allan

D. The Problem of Pure Economic Loss

If the foregoing analysis is correct, it is neither true to say that “the tort of conversion has already jumped the gap between tangibles and intangibles”,Footnote ⁶⁵ nor that it ought to. However, merely demonstrating that OBG does not reveal a lacuna in the legal protection of contractual rights does not prove the broader point that the law is adequate.

Take the following example. B, a large football club, signs A, a star football player, on a multimillion-pound, multi-year contract. To celebrate, A goes out to a bar where he is assaulted by C. A's injuries are sufficiently serious that he will miss the entire season.Footnote ⁶⁶ C's actions have plainly diminished the value of B's rights against A. However, because C neither caused A to breach his contract with B, nor intended to cause B loss, B has no claim in inducing breach of contract or the unlawful means tort, and thus no means of recovering from C.Footnote ⁶⁷

Is the absence of a claim proof of the Inadequacy Thesis, and can we envisage a way in which B can recover from C? According to what Douglas has described as the “reification” theory of contractual rights, the answer may be “yes”. Douglas explains that:

Because it “propertises” the value of a contractual right, the reification theory is a variation on the theme proposed by advocates of the Inadequacy Thesis. On this view, C will be liable to B because he has breached the in rem duty not to interfere with the value of B's contractual rights against A. Likewise, the receivers in OBG would be liable to the insolvent company for diminishing the value of its contractual rights against its customers. Whilst this expansive notion of “contract as property” captures third-party conduct that eludes the economic torts, “reification”, and all cognate theories, suffers from three problems, each more serious than the last.

The first is that, as Douglas has demonstrated,Footnote ⁶⁹ there is no authority for an in rem right to the value of a contractual right. An exclusive licencee of an intellectual property right cannot bring an action against a third party who, by infringing the underlying intellectual property right, diminishes the value of his contractual rights against the licensor.Footnote ⁷⁰ The second problem is that the judicial creation of any such novel property right or “fancy”Footnote ⁷¹ is forbidden by the numerus clausus principle, according to which there is a fixed or “closed list” of property rights recognised at common law.Footnote ⁷² Finally, irrespective of whether it is created by a court or Parliament, the creation of an in rem right to the value of an in personam contractual right would undermine the fundamental principle that third parties are not liable for causing pure economic loss, whether maliciously or otherwise.Footnote ⁷³ As established in Allen v Flood Footnote ⁷⁴ and affirmed in OBG,Footnote ⁷⁵ a plaintiff can only recover pure economic loss from a third party who intentionally induced a breach of contract or inflicted loss by unlawful means.

The law's general prohibition on claims for pure economic loss not only reflects the concern that allowing such claims would create “liability in an indeterminate amount for an indeterminate time to an indeterminate class”.Footnote ⁷⁶ It also reflects the fact that allowing individuals the liberty to inflict pure economic loss on each other is the sine qua non of competition, without which there could be no efficient markets through which to allocate scarce resources.Footnote ⁷⁷ This imperative was expressly recognised by Lord Nicholls, who remarked that “[c]ompetition between businesses regularly involves each business taking steps to promote itself at the expense of the other … Far from prohibiting such conduct; the common law seeks to encourage and protect it”.Footnote ⁷⁸ To recognise any free-standing right to the value of a contractual right, in whatever theoretical form, would be to overthrow this fundamental principle by a side-wind.Footnote ⁷⁹

A. False Analogies

Despite the criticisms made above, the basic truth underpinning the Inadequacy Thesis is that contractual rights can be vulnerable to acts of third-party interference. For instance, if a rogue agent acting outside his actual, but within his ostensible, authority makes a withdrawal from his principal's bank account, the bank will receive a good pro tanto discharge of its duties to the principal as its customer. Nothing in the foregoing denies that, in a colloquial sense, the agent has “stolen” or “converted” his principal's rights against the bank. However, what is denied is that invoking the tort of conversion can solve the problem faced by the contracting parties. To the contrary, it is argued that viewing the problem through the prism of conversion diverts attention from the very real problem that the law must solve in these circumstances; viz. how best to allocate the loss between the innocent promisor and promisee.

Invoking conversion is misleading because, unlike a simple conversion of chattels, attempts by third parties to usurp a promisee's right to contractual performance invariably involve three protagonists: an innocent promisee, an innocent promisor and a wrongdoer. The appropriate analogue of the latter problem is thus not a dispute between an owner and a thief, but that between an owner and a good-faith purchaser who has purchased the former's goods from a thief. Like the problem of the good-faith purchaser, acts of third-party interference with contractual rights create one of the “eternal triangles of the law”.Footnote ⁸⁰

In a legal Utopia in which rogues were identifiable and solvent, and legal proceedings cheaply and expeditiously resolved, triangle disputes would cause the law little anxiety.Footnote ⁸¹ In the case of stolen goods, for instance, either the owner would forfeit the goods to the purchaser and sue the thief in conversion, or the purchaser would forfeit the goods to the owner and sue the thief for breach of his implied warranty to pass good title.Footnote ⁸² In the real world, however, the rogue often cannot be identified or is not worth suing. Consequently, the real significance of the law's solution to any triangle dispute is not that it determines which of the two innocent parties must endure the tedium of suing the rogue, but who must bear the loss that cannot be recovered from him.Footnote ⁸³

B. Bank Money and Triangle Disputes

A particularly good example of a contractual triangle dispute is that caused by a rogue who attempts to steal bank money. The term “bank money” is, of course, something of a misnomer. The relationship between a bank and its customer is fundamentally contractual,Footnote ⁸⁴ and “bank money” is a chose in action; a debt owed by the bank to its customer equal to the sum deposited and payable on demand.Footnote ⁸⁵ If, at the instigation of a rogue, a bank processes an unauthorised transaction on its customer's account, the question that the law must answer is whether the payment constitutes a good pro tanto discharge of the debt that it owes its customer. Because financial fraud forms the core business of many organised crime syndicates, attempts to steal bank money are, unlike the peculiar dispute in OBG, a problem of great practical concern.Footnote ⁸⁶

As is true of all triangle disputes, the problem caused by the fraudulent targeting of bank accounts is not legal inadequacy in the sense that the rogue falls into some interstice between forms of action and cannot be sued. If, for instance, a rogue agent is acting within his ostensible authority, the bank will be entitled to debit the principal's account,Footnote ⁸⁷ leaving the principal to sue the agent for breach of the contract of agency. If, on the other hand, the agent was acting outside both his actual and ostensible authority, the bank will not be entitled to debit the principal's account but can recover the sum from the agent either as compensatory damages for deceit or, by “waiving the tort”, as money had and received.Footnote ⁸⁸ Whichever way the court answers the mandate question, the rogue will be liable for his misdeeds. The problem is instead that an infinity of claims is of no assistance against a rogue who either cannot be found or is so impecunious as to be judgment proof. Consequently, the resolution of the jural question of whether the bank discharged the debt to its customer amounts to the following: which of the two nominally innocent parties should bear the loss caused by the absence of the true wrongdoer? No appeal to the tort of conversion can assist in answering this question.

C. The Symmetry of the Parties

Whilst the apparent moral symmetry of the true parties to a triangle dispute seems to make this “an impossibly difficult question”,Footnote ⁸⁹ the law must answer it if the dispute is to be resolved. Who, then, should draw the short straw?

One possibility, suggested by the symmetry of the parties, is to pick at random. According to one principle of decision-theory, a decision maker should aim to “minimize the sum of the costs of making decisions and the costs of error”.Footnote ⁹⁰ Because all decisions about liability in a generic triangle dispute are equally unfortunate, none can be considered erroneous. Consequently, the only imperative is to minimise the cost of decisions. In resolving such disputes, the law should thus refrain from expending resources in “choosing” a winner and should instead arbitrarily “pick” one.Footnote ⁹¹

Whilst this proposal appears radical, picking strategies are used to resolve other disputes where the merits of the parties appear equal. Take the maxim in pari delicto potior est conditio possidentis.Footnote ⁹² Where both parties to a dispute over the ownership of a chattel are tainted by illegality, the law is content to allow the possessor to prevail, not because he is worthier, but because the alternative is no better. So long as no decision is superior to any other, the decision maker is not justified in incurring costs in altering the status quo ante.

Just as there is nothing to choose between parties in pari delicto because they are equally bad, there is nothing to choose between innocent parties to a triangle dispute because they are equally good. The rules “favour the promisor” and “favour the promisee” are thus equally satisfactory. The only imperative is to make a choice between the alternatives and, to avoid creating prospective uncertainty, stick with it.

Picking rules are both cheap to apply and have the welcome effect of “sanitising” decisions that require discrimination between equals.Footnote ⁹³ Despite these virtues, many will chafe at the idea of harnessing randomness to resolve a question of just distribution.Footnote ⁹⁴ Whilst one might sympathise with this intuition, those who hold this view must find a satisfactory way of breaking the symmetry of the parties to a generic triangle dispute.Footnote ⁹⁵ The problem here is not that arriving at a more just outcome requires a more costly decision process. If a fairer outcome can be reached, then incurring additional decision costs may be justified. It is instead that it is difficult to conceive of decision criteria that a court would find acceptable.

The essential difficulty is that triangle disputes raise questions of just distribution that fit awkwardly, if at all, into the familiar private law paradigm of corrective justice in which reasons for liability are correlatively structured.Footnote ⁹⁶ One need not be a corrective justice evangelist to appreciate that its inherently bilateral structure spares courts from the embarrassment of deciding questions of liability by appeal to unipolar considerations. For instance, whilst a defendant's deep pockets might be a reason to subject her to a higher marginal tax rate,Footnote ⁹⁷ its irrelevance to her bilateral relationship with the plaintiff means that it is an incoherent basis on which to make her liable to the plaintiff as a matter of corrective justice.Footnote ⁹⁸ However, because the innocent parties to a triangle dispute cannot be considered as the “doer” and “sufferer” of the same wrong,Footnote ⁹⁹ some non-correlatively structured basis for liability must be found. What might this be?

A utilitarian, for instance, might solve the problem by assigning liability in the way that minimises the net loss of social utility. If one accepts the law of diminishing marginal returns, a utilitarian decision rule could be to “allocate liability to the richer of the two parties”.Footnote ¹⁰⁰ Someone who sought to allocate liability on the ground of relative “need” may arrive at the same conclusion. Putting aside objections to the philosophical rigour of interpersonal comparisons of welfareFootnote ¹⁰¹ and the equivocal nature of “need”,Footnote ¹⁰² the fatal objection to any such rule is that it would be far too controversial for a court to apply.Footnote ¹⁰³

The criterion of “desert” fares little better. Unlike need, desert is generally conceived of as an entitlement justified by something that one has done or achieved.Footnote ¹⁰⁴ The winner of the 100-metre sprint “deserves” her gold medal because she crossed the finish line first. Whilst desert works well in some applications,Footnote ¹⁰⁵ it is difficult to conceive of what a party might relevantly have done to warrant favourable treatment in a triangle dispute. Whilst we might say that the party who took reasonable precautions is deserving of preferential treatment, this is simply to clothe the least-cost-avoider principle in the language of non-consequentialist morality.Footnote ¹⁰⁶

D. The “Calabresian” Turn

Does the foregoing discussion lead inexorably to the conclusion that the law should resolve triangle disputes by some version of a coin flip? The answer is “no”. Whilst the virtues of a coin flip are not to be underestimated, a better way of breaking the symmetry between the innocent parties is to allocate liability on the grounds of relative cost.

Unlike flipping a coin or any other arbitrary “picking” rule, a decision rule that turns on relative cost pursues a socially worthwhile end, viz. efficiency, and does so without committing any obvious moral faux pas.Footnote ¹⁰⁷ Indeed, as noted above, it is entirely possible that considerations of efficiency and desert will pull in the same direction.Footnote ¹⁰⁸ In any case, the modest claim being advanced here is not that efficiency is the premier value.Footnote ¹⁰⁹ It is instead that, where questions of just distribution are too difficult or controversial for a court to decide, it is desirable, ceteris paribus, to employ rules that minimise the cost of the dispute.

If the foregoing is true, how should disputes between innocent promisees and promisors be resolved? In light of Mautner's suggestion that triangle disputes should be viewed as examples of the accidents that form the core of tort law,Footnote ¹¹⁰ their efficient resolution is best achieved by following Calabresi's rule that liability should be allocated so as to “reduce the sum of the costs of accidents and the costs of avoiding accidents”.Footnote ¹¹¹

Because the Calabresian framework is based on a global assessment of cost, it nominally requires two enquiries.Footnote ¹¹² Ex ante, the law should minimise the cost of prophylaxis by identifying the party who can avoid the accident at least cost. Ex post, after attempts at prevention have failed, the imperative is to minimise the net sunk cost of the accident.Footnote ¹¹³ This second imperative may favour imposing the loss on the party who, though she may not have “caused” the accident in the conventional sense, is the superior loss spreader or has the deepest pockets.Footnote ¹¹⁴

That cost is a defensible basis on which to discriminate between the parties to a triangle dispute does not mean that it is perfect. Calabresi's solution not only requires the legal architect to answer difficult empirical questions about least cost-avoiders and best loss-spreaders, it also requires her to solve a complex optimisation problem where the party who is best able to spread the costs of an accident is not also the least cost-avoider.Footnote ¹¹⁵ In these cases, it must be decided whether the benefits of loss distribution following an accident justify increasing the costs of avoiding accidents in the first place.

A. The Significance of Privity

Mautner describes the paradigmatic triangle dispute as one in which, “parties not in contractual privity themselves assert simultaneous claims of rights over the same asset whose concurrent discharge is legally impossible”.Footnote ¹¹⁶ A crucial point, and one that is completely overlooked in the “contract as property” literature, is that triangle disputes involving contractual rights diverge from the paradigm case of the good-faith purchaser of stolen goods because, unlike an owner and purchaser, promisors and promisees are not strangers.

In principle, then, the law can implement the Calabresian solution by doing nothing. This is because, as the “strong version” of Coase's theorem tells us,Footnote ¹¹⁷ in a world of full information and costless bargaining, the parties will reach the efficient outcome themselves. Returning to the example of bank money, imagine that it costs a customer £10 to avoid losses caused by the misuse of a lost bankcard, but that it costs the bank £20 to avoid the same loss. If, under the default rule, the bank were liable for these losses, then it would be willing to pay, and the customer willing to accept, some amount between £10.01 and £19.99 to transfer liability to the customer. If the roles were reversed and the customer were liable, the outcome would be the same because the bank would demand £20.01 for the reversal of the default rule, and no rational customer would pay a bank £20.01 to assume liability for a risk that, to her, costs £10 to avoid.

Whilst, ex post, each party will disclaim liability for all losses, ex ante, rational parties will agree to an efficient contract.Footnote ¹¹⁸ Any arbitrary allocation of initial entitlements will do because, as the example above demonstrates, any misallocation will be corrected by transactions between the affected parties.Footnote ¹¹⁹

B. Are Contracts Efficient?

The Coasean thought experiment demonstrates why the solution to acts of third-party interference with contractual rights is not to treat contractual rights as property, but to create the conditions that allow the contracting parties to draft more complete contracts. The obvious question is: do parties create tolerably efficient contracts in the real world? There are good reasons to believe that they do. Once again, bank money provides an illuminating example.

When a third party induces a bank to process an unauthorised transaction on the customer's account, the jural question is whether the bank's actions constitute a good discharge of the debt that it owes its customer. As should be obvious from the foregoing discussion, the general answer to this question is “no”.Footnote ¹²⁰ As is true of other contractual relationships, that between a bank and its customer follows the Basic Rule, according to which a promisee's rights can only be discharged by performance rendered to the promisee herself. This is demonstrated by Rogers v Kelly,Footnote ¹²¹ in which money deposited by the plaintiff was mistakenly paid by his bankers to the defendant. The plaintiff's claim in money had and received failed because, in restitutionary terms, the defendant was not enriched at his expense. The enrichment was instead at the expense of the bank, which paid the defendant with its own money and, in the absence of authorisation, had no right to debit the plaintiff's account. Although a transfer of value occurred, the plaintiff's “bank money” remained intact. As Lord Ellenborough explained: “The plaintiff's claim is on the bankers, and they [the bankers] must seek their remedy against the defendant the best way they can. The plaintiff's money must still be considered as in the hands of the bankers. His account with them is the same as if this mistake had not been committed.”Footnote ¹²²

In these circumstances, a customer is entitled to have his account re-credited, and the bank must attempt to recover its losses from the mistaken payee.Footnote ¹²³ Whilst this outcome will certainly be welcomed by the customer, the relevant question is: is it efficient?

Any assessment of the efficiency of contracts between banks and their customers should commence by noting that, whatever default rule is applied, the bank is unlikely to bear the cost of financial fraud in the long term. Whilst the Basic Rule initially places liability on the bank, this arrangement simply transforms the costs of fraud into another operational expense which, in a (reasonably) competitive market,Footnote ¹²⁴ banks will pass on to their customers.Footnote ¹²⁵ In functional terms, the Basic Rule makes banks insurers for their customers who pay what are effectively premiums in the form of slightly higher fees. That the choice of liability will not alter the distribution of income does not, however, mean that there are no benefits, assessed in terms of minimising the ex post costs of financial fraud, of imposing the loss on the bank in the first instance. In particular, there are two benefits of a default rule that makes a bank the de facto compulsory insurer of its customers.

First, as is true of an insurance company, a bank is an effective loss spreader and, ceteris paribus, it is better to distribute the inevitable cost of fraud amongst a bank's many customers than it is to concentrate it on a single, unfortunate victim.Footnote ¹²⁶ Secondly, whilst the amount that the average customer will pay in higher bank fees will equal her expected losses from financial fraud in the long term,Footnote ¹²⁷ most bank customers are likely to be risk averse and thus would prefer to make many small payments to the bank in the form of higher fees than to suffer one very large loss should they fall victim to fraud.Footnote ¹²⁸ Whilst, under a regime of self-insurance, customers could theoretically purchase insurance against financial fraud, the Basic Rule obviates the need for them to do so. If the law wishes to avoid these costs, it is better to have a few large financial institutions bear the initial loss than to have millions of individual customers organise their own insurance.

C. Solving Moral Hazard with Contracts

So far as ex post efficiency is concerned, the Basic Rule appears desirable. However, a global assessment of cost requires one to minimise the sum of the costs of accidents and of avoiding them. This imperative is simple to satisfy either where there is no obvious least cost-avoider, or where the least cost-avoider is also the best loss spreader. However, if the former is not also the latter, a problem arises because parties’ knowledge of ex post liability will alter their ex ante behaviour. As a system of de facto insurance, the Basic Rule suffers from a problem that plagues all arrangements in which one party undertakes to indemnify the other: the inefficient behavioural changes that economists refer to as “moral hazard”.Footnote ¹²⁹

In the absence of insurance, a rational maximiser subject to a particular risk would invest in precautions up to the point at which the cost of risk-elimination equalled the expected cost of the risk. So, for instance, he would spend £4.99, but not £5.01, to eliminate a risk with an expected value of £5.00. However, this calculation ceases to hold when insurance is introduced. Because an insured party will bear none of the loss should the risk materialise, he has no incentive to invest so much as a penny to prevent it. Insurance thus deters the taking of cost-justified precautions and encourages inefficient risk-taking behaviour because the insured party captures all the benefits of the risky behaviour but bears none of its costs, which are imposed on other policy holders.

One might think that because banks will pass on higher ex ante costs to their customers, the moral hazard would be solved by the threat of increased bank fees. Unfortunately, as with insurance more generally, an externality problem remains because of the necessity of pooling people with disparate risk profiles. A risk-prone member of an insurance pool internalises all the benefit of his risky behaviour. However, because his premium is calculated on the average level of risk in his pool, he pays only a fraction of its cost, the balance being borne by the more prudent members. Consequently, although risky behaviour raises premiums for everyone, including the member in question, it is still rewarded.Footnote ¹³⁰ If it were possible to categorise each member of an insurance pool according to his appetite for risk and charge premiums accordingly, there would be no externality problem. However, just as no insurance company can individuate every policy holder based on his appetite for risk, no bank can tailor its fees according to the carelessness of individual customers.

The problem of moral hazard is not confined to attempts to optimise ex ante and ex post costs when those goals are in tension. It also afflicts attempts to minimise ex ante costs where the efficient solution to accident prevention is for both parties to take precautions.Footnote ¹³¹ Take the following example. Imagine that it costs a bank £110 per annum, per customer to prevent sophisticated cyberattacks on its online banking facilities. Imagine also that it would cost an individual customer £200 to avoid the same loss. Because the bank is the least cost-avoider, a contract between a bank and its customer that allocated liability for such losses to the bank is prima facie efficient. However, imagine further that, if customers spent £10 per annum on purchasing updates to their antivirus software, the bank's cost of precautions could be reduced to £90. This is a more efficient solution because the total costs of precautions required to eliminate the risk would be £100, not £110. Unfortunately, as we know from the foregoing discussion, the efficient joint-care solution will not occur because, where the customer is indemnified against losses caused by cyberattack by the terms of her contract with the bank, she has no incentive to take any precautions at all.

Precisely this problem afflicts disputes between owners and good-faith purchasers of stolen goods. The efficient prevention of the dispute requires both the owner to make some efforts to safeguard his property and the purchaser to make some effort to assess the quality of her vendor's title. Because the parties are strangers and cannot bargain prior to the dispute, the law relies on the categorical title rules of property law that favour either the owner or the purchaser.Footnote ¹³² The problem with these all-or-nothing rules is that they cannot simultaneously incentivise each party to invest optimally in precautions.Footnote ¹³³ If the owner invariably succeeds, he will fail to take efficient precautions against theft. If the purchaser invariably succeeds, she will not invest optimally in investigating her vendor's title. Whilst a requirement of good faith on the part of the purchaser may ameliorate the problem, there is no reason to believe that the good-faith standard is also the optimal standard. This is why recent economic analyses of the good-faith purchaser problem have advocated replacing the all-or-nothing title rules of property law with a negligence standard.Footnote ¹³⁴

When the parties are in a contractual privity, these problems of moral hazard can be largely overcome, and without resorting to the law of negligence. In the example given above, that the bank was both the best loss spreader and the least cost-avoider made it the obvious candidate for liability. Nevertheless, a moral hazard remained because fixing the bank with liability destroyed the incentives required to sustain the more efficient joint-care model, under which precautions taken by both contractual parties reduced the net cost of avoidance by £10 per annum. Crucially, a sufficiently detailed contract would eliminate this stubborn moral hazard. This is because the bank's promise to bear the costs of a cyberattack would be made conditional on the customer updating her antivirus software, employing and safeguarding a sufficiently robust online banking password, and so on. Conditional terms such as these allow the parties to minimise the net costs of precaution, and without sacrificing loss spreading and other ex post goals.

Whilst a complete contract, enforceable at zero cost, would eliminate all moral hazard, a frictionless world is not required for parties to substantially ameliorate the problem. Insurance companies, for instance, routinely employ simple contractual mechanisms to prevent the most egregious acts of carelessness on the part of the insured.Footnote ¹³⁵ The most straightforward is the requirement that an insured pay an excess or “deductible” when making a claim, the effect which is to make the insured responsible for some part of the total loss. Likewise, the common strategy of insuring an asset for some fraction of its total value, known as “coinsurance”, ameliorates moral hazard by incentivising the insured to take those precautions whose cost is less than the expected value of the damage to the uninsured portion of the asset. Insurance companies also offer inducements to encourage efficient ex ante conduct. For instance, many insurance companies will offer reduced premiums on home and contents policies to those who install security devices in their homes, and on health and life insurance policies to those who do not smoke or drink to excess. If the reduction in the premium exceeds the cost of the precaution, such terms are advantageous to both the insured, who will enjoy a reduced premium, and to the lower-risk members of his insurance pool who, as explained above, would otherwise subsidise his risky behaviour.Footnote ¹³⁶ Whilst such contractual measures do not make the bargain optimally efficient, they do much to correct the misaligned incentives that arise when one party undertakes to indemnify another.

That we are doomed to inhabit a world of “second best” is only an objection to the strong version of the Coase theorem, according to which the law is indifferent to the content of default rules because the parties can always bargain around them. It is not an objection to the main thesis of this article, which is that the problem of third-party interference with contractual rights raises a problem of loss allocation that is best solved by improving the contract between the affected parties. If the parties suffer from insurmountable information or transaction costs, it is for the law to assist them by creating efficient default rules where they are either known or can be reasonably guessed at.Footnote ¹³⁷

Some might regard the possibility of the state supplying efficient terms as hardly less fantastical than the avowedly hypothetical conditions of the Coase theorem. Yet the reality is that the law has already gone some way to creating default contractual terms that approximate the efficient solution to the problem of financial fraud. Take, for instance, the Payment Services Regulations 2017 (UK) (“PSR”). In conformity with the Basic Rule, the PSR obliges a bankFootnote ¹³⁸ to re-credit its customer'sFootnote ¹³⁹ account if the bank processes a “payment transaction” that is not authorised by the customer.Footnote ¹⁴⁰ By imposing liability on the bank, this baseline rule secures the ex post benefits of loss spreading by dividing the loss amongst the bank's many customers. Significantly, the PSR also creates a series of exceptions that shift liability for certain losses from the bank to the customer. For instance, the customer will be liable for a maximum of £35 if the unauthorised transaction arose from the use of a lost or stolen “payment instrument”.Footnote ¹⁴¹ The customer will be liable for all losses if she has acted fraudulently or has, “with intent or gross negligence”, failed to use the payment instrument in accordance with its terms and conditions, to notify its issuer without “undue delay” if it has been lost or stolen, or to take all reasonable steps to safeguard her personalised security credentials.Footnote ¹⁴² By transferring liability for these losses to customers, these provisions incentivise customers to exercise basic prudence and thus mitigate the worst of the moral hazard created by the application of the Basic Rule, or some version of it.

In Australia, similar efficiency-promoting provisions can be found in the “ePayments code” (“Code”),Footnote ¹⁴³ a voluntary code of conduct covering electronic payments that is administered by the Australian Securities and Investments Commission, and to which most major banks and financial institutions subscribe.Footnote ¹⁴⁴ As with the PSR, the Code preserves a version of the Basic Rule by providing that a customer “is not liable for loss arising from an unauthorised transaction where it is clear that [the customer] has not contributed to the loss”,Footnote ¹⁴⁵ and specifies when she will be taken to have so contributed.Footnote ¹⁴⁶ The Code also lists other circumstances in which the customer will be liable for unauthorised transactions on her account. Significantly, and consistently with the PSR, those circumstances include, inter alia, customer fraud and the failure to satisfy “pass code security requirements,”Footnote ¹⁴⁷ leaving a card in an Automatic Teller Machine,Footnote ¹⁴⁸ or unreasonably delaying reporting the misuse, loss or theft of a “device” to its issuer.Footnote ¹⁴⁹

As is true of the insurance contracts discussed above, the liability-shifting provisions in these regulatory codes allow for the creation of a contract between a bank and its customer that preserves the loss-spreading benefits of the Basic Rule whilst avoiding the worst of the moral hazard created by it. This is not to say that they perfectly replicate the frictionless universe, or even that this is their primary objective. Like every legislative or regulatory instrument, the PSR and the Code are the product of lobby-group pressure and political expediency.Footnote ¹⁵⁰ The important point is that they demonstrate that, even in our non-Coasean world of expensive information and costly bargains, the solution to the problem caused by acts of third-party interference with contractual rights is not more property rights, but better contracts.