Implied terms

Implied terms are commonly divided into terms implied in fact and terms implied in law. What are they and when will they arise?

Terms implied in fact

The original test for when the courts would imply a term in fact into a contract was laid down in 1889 in The Moorcock (where it was argued, and accepted, that there was a term implied in a contract for mooring a ship at the jetty under which the defendant owners of the jetty guaranteed that they had taken reasonable steps to see that the depth of the water around the jetty was such that it was safe for the claimant shipowners to moor their ship at the jetty). This was the business efficacy test: ‘what the law desires to effect by the implication is to give such business efficacy to the transaction as must have been intended at all events by both parties who are business men’ (per Bowen LJ). So under the business efficacy test the courts will imply a term in fact into a contract where it was necessary to do so to make the contract work in the way that was intended by the parties to the contract.

A different test was suggested by MacKinnon LJ in the 1939 case of Shirlaw v Southern Foundries (1926) Ltd (where it was argued, and accepted, that there was an implied term in a contract appointing a managing director of a company for ten years that the company would not remove the director from his position for the period of his appointment, and it would not alter its articles of association so as to enable the director to be removed). This was the officious bystander test: ‘Prima facie that which in any contract is left to be implied and need not be expressed is something so obvious that it goes without saying; so that, if, while the parties were making their bargain, an officious bystander were to suggest some express provision for it in their agreement, they would testily suppress him with a common “Oh, of course!”’

There are problems with both these tests. Consider:

Dodgy: A Finance company sells an investment product to Mug knowing that the product is worthless and that Mug will probably lose a lot of money on the deal. However, selling the product to Mug (and thousands like him) will help Finance avoid the consequences of some terrible investment decisions that it made in the past. Mug has now lost a lot of money as a result of owning this investment product and wishes to sue Finance, arguing that there was an implied term in the contract of sale that Finance would warn the claimant if it knew that the product it was selling him was worthless.

Secretary: Randy appoints Beauty to work for him as a secretary. Randy has always fancied Beauty; in fact, that is the main reason why Beauty got the job. Two days after starting work, Randy tells Beauty he can’t stop thinking about her and would like them both to go away for a naughty weekend in Blackpool. Beauty does not welcome Randy’s advances and wants to argue that there is an implied term in her contract with Randy that Randy will not ask her out or in any way indicate that he is sexually attracted to her.

Neither of these terms need to be implied into the contracts in Dodgy and Secretary respectively in ordert to make the contracts work. Implying a term that Finance will warn Mug if it knows the product it is selling him is worthless will actually negate the contract, as Mug will not contract with Finance if he is given such a warning. And Beauty does not really need Randy to shut up about how much he fancies her to get on with her job of working for him as a secretary. So neither of these terms could be implied under the business efficacy test. And if we apply the officious bystander test, it is obvious that had an officious bystander asked Mug and Finance, ‘Is Finance undertaking to warn Mug if it knows this product is worthless?’ then Mug would probably have said ‘Of course!’ but Finance would have said ‘No, we’re not – it’s for Mug to decide whether or not to buy this product.’ And if an officious bystander had asked Randy and Beauty, ‘Is Randy undertaking not to sexually harass Beauty?’ then Beauty would probably have said, ‘Of course!’ but Randy would have said, ‘I think that’s a bit extreme – if I want to tell Beauty how beautiful she is, I think I should be allowed to do that.’

Maybe we should rest content with that, and say that Mug and Beauty will lose their cases, insofar as they are based on the implication of a term in fact into their contracts with Finance and Randy respectively. However, Mug and Beauty may still have two arguments left to them when arguing that there was an implied term in fact in their contracts:

(1) The Belize argument

In Attorney General of Belize v Belize Telecom Ltd (2009) (where it was argued, and accepted, that there was an implied term in a company’s articles of association that a director who had been appointed under a special procedure that could only be invoked by a shareholder holding a ‘golden’ share in the company and over 37.5% of ‘C’ shares in the company should step down if the shareholder who had appointed him subsequently sold some of its ‘C’ shares so that it no longer held 37.5% of the ‘C’ shares in the company), Lord Hoffmann suggested that both the ‘business efficacy’ and ‘officious bystander’ tests for implying a term in fact into a contract should be superseded by a new test under which a term will be implied into a contract if such a term ‘would spell out in express words what the [contract], read against the relevant background, would reasonably be understood to mean’ (at [21]).

As John McCaughran (‘Implied terms: the journey of the man on the Clapham omnibus’ (2011) 70 CLJ 607, at 614) has acutely pointed out, this test for implying a term in fact into a contract involves applying a reverse ‘officious bystander’ test – instead of the officious bystander asking the parties what they have meant to agree, the parties ask the bystander ‘You know the business background against which we are dealing – what do you think we have agreed in entering into this contract?’ Taking the power to determine the terms of the contract out of the hands of the parties and into the hands of a reasonable bystander might make it easier for Mug and Beauty to win their cases.

However, Hoffmann’s test for implying a term in fact into a contract can be criticised on a number of grounds. (i) He uses the exact same test for interpreting the terms of a contract (see his decision in Investors Compensation Scheme Ltd v West Bromwich Building Society (1997)) as he does for implying terms into a contract, but as Bingham MR pointed out in Phillips Electronique Grand Public SA v British Sky Broadcasting (1995), the two things are very different exercises – implying a term in a contract that the parties did not expressly say they wanted to be part of the contract is a lot more of an intrusion on the parties’ freedom of contract than trying to determine what the parties meant by a term that they did expressly say that they wanted to be part of the contract. (ii) Hoffmann’s test makes it too easy for the courts to impose on the parties to a contract terms that they did not want to be part of the contract, on the basis that a reasonable person with a similar business background to the parties would have thought that that term was part of the contract. (iii) Hoffmann’s test is also very vague and creates a great deal of commercial uncertainty for parties entering into contracts, as they would find it hard to predict what terms a court employing Hoffmann’s test might end up implying into the contract.

Perhaps significantly, the caselaw since the Belize case has not expressed much enthusiasm for Hoffmann’s test, with (1) the CA in Mediterranean Salvage & Towage v Seamar Trading & Commerce Inc (2009) endorsing the ‘business efficacy’ test, (2) the CA in Groveholt Ltd v Alan Hughes (2010) endorsing the ‘officious bystander’ test, and (3) a majority of the UKSC effectively disowning the test in Marks & Spencer plc BNP Paribas Securities Services (2015) (at paras [22]-[31], per Lord Neuberger, with Lord Carnwath supporting the Hoffmann test in his judgment) in favour of implying a term when the ‘business efficacy’ test or the ‘officious bystander’ test indicates such a term should be implied (para [21], per Lord Neuberger). (See elsewhere on this website for a detailed casenote on the Marks & Spencer case.) So it is doubtful whether we should seek to rely on Hoffmann’s test in Mug and Beauty’s cases.

(2) The officious bystander argument revisited

Both Mug and Beauty might be able to argue that the ‘officious bystander’ test is satisfied in their cases. Both would rely on the fact that had an officious bystander asked whether Finance was promising to warn Mug if it knew that it was selling him a worthless product, or whether Randy was guaranteeing that he would not sexually harass Beauty, while Finance and Randy might say now that they would not have gone along with the officious bystander’s suggestion, at the time the contract was entered into things would have been very different. The officious bystander’s question would have put Finance and Randy in a very difficult position: they would either have had to confess that, no, they wanted the option of selling Mug or worthless product or sexually harassing Beauty and endanger the deal, or they would have had to act as though they were decent people and say to the officious bystander, ‘Of course we would never sell Mug a product that we knew to be worthless’ or ‘Of course I would never dream of making sexual advances to Beauty.’ Faced with this choice, they probably would have done the latter and suppressed the officious bystander with a testy, though grudging, ‘Oh, of course!’. Given this, Mug and Beauty might be able to argue that the officious bystander test is satisfied in their cases, or at least that Finance and Randy should not be allowed now to argue that the officious bystander test is not satisfied, given that at the time they would have probably been embarrassed (for fear of looking like a really dodgy person) or forced (for fear of endangering the deal) into giving a positive response to the officious bystander’s question.

I think argument (2) would probably work (see, in particular, para [21] of Lord Neuberger’s judgment in the Marks & Spencer case: ‘If one [asks] what the parties would have agreed, one is not strictly concerned with the hypothetical answer of the actual parties, but with that of notional reasonable people in the position of the parties at the time at which they were contracting’), and indicates that the officious bystander test, if sensitively applied, copes well with difficult cases such as Dodgy and Secretary. But less can be said in favour of the business efficacy test – if it would rule out an implied term in cases like Dodgy and Secretary on the ground that the terms were not needed to make the contract work, when such terms would be implied on the officious bystander test, then that seems to indicate that officious bystander test provides a more comprehensive test for implying a term in fact into a contract than the business efficacy test does.

Terms implied in law

It follows from the above that terms implied in fact into a contract are designed to give effect to the intentions of the parties – or what the parties reasonably indicated that they intended – in entering into the contract. Terms implied in law are quite different. Terms implied in law act as default rules for particular, well-established types of contractual relationship, such as those that exist between a business seller of goods and a purchaser of those goods, an employer and an employee, an insurance company and someone purchasing insurance from that company, a shipowner and an owner of goods placed aboard the ship, a builder and a client, a landlord and a tenant, and so on and so on. A default rule is a rule that will apply to a particular relationship unless the parties to that relationship indicate that they don’t want it to apply. Default rules are a huge aid to efficiency in contracting: they allow the parties to a well-established type of contractual relationship to contract with each other safe in the knowledge that a large number of terms will automatically be implied into the contract to safeguard their respective interests and that all they need to focus on in their negotiations are a few terms which need to be fiddled with to fit the circumstances of their case. (In the same way, the fact that when you open a new document on your computer, the fact that document is preformatted according to certain default rules for page size, font, margins, line spacing and so on, means they you can get on with writing what you want to write much more quickly: you don’t specify every aspect of what your document should look like – you can just focus on adjusting particular default rules that you don’t want to apply.)

Once we understand the status of terms implied in law as default rules, two things follow:

(1) We can make sense of the emphasis that the House of Lords placed in Liverpool City Council v Irwin (1977) on the courts’ only implying a term by law into a contract when it is necessary to do so. ‘Necessary’ here does not mean the same as ‘necessary to give business efficacy’ to the contract. If it did, then the House of Lords’ decision in Irwin would have effectively abolished terms implied in law. But it did not: the term that the House of Lords ended up implying in Irwin into the contract in that case between the landlord (the city council) and its tenants occupying a block of flats known as ‘the Piggeries’ was a term requiring the landlord to take reasonable steps to keep the common parts of the let premises in good repair. (In this case, that would have been the lifts and the stairs.) The House of Lords did not need to imply that term into the contract between the landlord and the council tenants in order to make the contract work properly – so ‘necessary’ in Irwin could not have meant ‘necessary to give business efficacy’ to the contract. What the House of Lords actually meant (see, on this, Jane Stapleton, ‘Duty of care and economic loss: a wider agenda’ (1991) 107 LQR 249, 290-1) by ‘necessary’ in this context was ‘The suggested term has to be such that we think it would be reasonable to imply this into any contract of the general type with which this case is concerned, so that from now on that term would become (unless the parties indicated otherwise) a necessary incident of a contract of that type.’ So before the House of Lords could imply a term by law into the landlord-tenant contract in Irwin, requiring the landlord in that case to take reasonable care of the common parts of the let premises, the House of Lords had to be convinced that it would be reasonable to imply such a term into any landlord-tenant contract no matter who the landlord might be and who the tenant might be. And plainly it was reasonable: when you are letting premises to a large number of people, it seems obvious that the responsibility for looking after the common parts of the premises should fall on the landlord and it would be very difficult for the tenants to co-ordinate efforts themselves to keep those common parts in good repair.

(2) When you are doing a problem question which turns on whether you can imply a term into a particular contract, and you have found that no such term can be implied in fact into the contract, it would be unwise to confidently assert that the courts will imply such a term in law into the contract unless you have statute or caselaw to support that assertion. If you do not, and it is a novel question whether the courts would imply a term in law of the type you are considering into this kind of contract, you cannot have any confidence that the courts will imply such a term into a contract, given the huge number of factors the courts would have to take into account in determining whether it would be reasonable to imply such a term into all contracts of that type. (For an account of these considerations see Peden, ‘Policy concerns behind implications of terms in law’ (2001) 117 LQR 459.) So when you are considering whether the courts will find that a term was implied by law into a particular type of contract, stick closely to the existing caselaw and statute law on this issue. And when it comes to the existing law, the two most important terms that will be implied by law into a contract (unless the parties indicate that they do not want the term to be implied – and even then their freedom to avoid that term may be limited by the Unfair Contract Terms Act 1977) that you have to know about and carry around with you in your mental knapsack are: (i) the terms implied into a contract between a business seller of goods and someone buying those goods that the goods will be of satisfactory quality and reasonably fit for the purpose for which the buyer let the seller know he wanted the goods (under s 14 of the Sale of Goods Act 1979); and (ii) the term implied into a contract for the supply of services, supplied in the course of business, that those services will be performed with a reasonable degree of care and skill (under s 13 of the Sale and Supply of Goods and Services Act 1982). Implied term (ii) is particularly useful as it applies in a wide range of circumstances – builder-client contracts; hotel-guest contracts, cinema-customer contracts, restaurant-diner contracts, and so on and so on.

Consumer surplus

What is it?

‘Consumer surplus’ is an economic term, referring to the difference between what someone would have been ready and willing to pay for a particular item and what they actually ended up paying for it.

The irrelevance of consumer surplus to contract damages

As we will see, it has been suggested by some academics and judges that in a breach of contract case, the damages payable to the victim of the breach may in some way reflect or protect the victim’s consumer surplus. This is untrue.

Suppose that you contract to sell me a ticket to a Bruce Springsteen concert for £100. I would have been ready and willing to pay £300 for a ticket (I’m relatively well off, and love The Boss), so my consumer surplus on this transaction is £200. You then tell me that you won’t be able to get me a ticket for the concert after all – your supplier of Springsteen tickets has let you down.

If I can get a ticket from another source for, say, £150, then you will be liable to pay me £50 in damages: the difference between what I would have paid you had you properly performed, and what I’ve ended up having to pay someone else for what I entitled to from you.

If I can’t get a ticket from another source to the concert, then – as this was a contract for fun – I’ll be entitled to sue you for damages for the loss of enjoyment of attending the concert. Awarding such damages requires the court to put a monetary value on how much I would have enjoyed attending the concert. Some might argue that this should be assessed according to how much I would have been ready and willing to pay for a ticket (£300) – but this is a mistake. How much I would have been ready and willing to pay for a ticket not only reflects my assessment of how much I would enjoy going to a Bruce Springsteen concert, but also how much money I have got in my pocket. So I might value the enjoyment of attending a Bruce Springsteen concert at £1,000, but only have been ready and willing to pay £300 for a ticket because I simply don’t have enough money to be able to afford to splash out £1,000 on a Bruce Springsteen concert. One way of putting a monetary value on how much I would have enjoyed attending the concert is to ask – Once I had the ticket, how much would I have been willing to sell it for? It makes sense that I would not have sold the ticket for less than the monetary value that I put on the enjoyment of attending the concert. It’s highly unlikely that once I got the ticket to go to the Bruce Springsteen concert that I would have been willing to sell it just for £300. It’s likely that I would only have been ready and willing to sell the ticket for a lot more money than I would have been ready and willing to buy the ticket for. Let’s say that had I got the ticket, I would have been ready and willing to sell it to someone else for £750. In that case, we should value my loss of enjoyment from not attending the concert at £750 and award me £650 in damages – £750 minus the £100 I have saved as a result of your breach because I no longer have to pay you for the ticket.

So – if you breach your contract with me to supply me with a ticket to the Bruce Springsteen concert, I will either be able to sue you for £50 (if I can get a ticket from another source) or £650 (if I can’t). In neither case can I sue you for £200 – which is my consumer surplus on my transaction with you. My consumer surplus is completely irrelevant to how much I might be able to sue you for.

Consumer surplus in the cases

Despite this, there are a number of cases which say that damages awards in breach of contract cases can be designed to protect the victim of breach’s ‘consumer surplus’.

In Ruxley Electronics and Construction Ltd v Forsyth (1996), Lord Mustill said that ‘the law must cater for those occasions where the value of the promise to the promisee exceeds the financial enhancement of his position which full performance will secure. This excess, often referred to in the literature as the “consumer surplus”…is usually incapable of precise valuation in terms of money, exactly because it represents a personal, subjective and non-monetary gain. Nevertheless where it exists the law should recognise it and compensate the promisee if the misperformance takes it away.’

In Farley v Skinner (2002), Lord Steyn observed of Lord Mustill’s judgment in Ruxley that ‘for Lord Mustill…the principle of pacta sunt servanda [contracts must be kept] would be eroded if the law did not take account of the fact that the consumer often demands specifications which, although not of economic value, have value to him. This is sometimes called the “consumer surplus”… Lord Mustill rejected the idea that “the promisor can please himself whether or not to comply with the wishes of the promise[e] which, as embodied in the contract, formed partof the consideration for the price”.’

Lord Steyn did go on to say that ‘Labels sometimes obscure rather than illuminate. I do not therefore set much store by the description “consumer surplus”.’ Quite right too. It is obvious that in these passages the phrase ‘consumer surplus’ is being misused to act as synonym for ‘enjoyment’ or ‘pleasure’. So when we say (1) ‘The law can award damages to compensate the victim of a breach of contract for the loss of his consumer surplus’ all we are actually saying is (2) ‘The law can award damages to compensate the victim of a breach of contract for the loss of the pleasure or enjoyment he would have obtained from proper performance of the contract’. But if that’s the case, then why not say (2) rather than (1)? We should prefer simple words whose meaning is plainly obvious over complex words whose meaning has to be explained. And we should definitely not borrow concepts from another discipline where they have a particular, technical meaning and then use those concepts to say something completely different. Doing so can only make it even more difficult for other people to understand what we are saying.

The source of the problem

The blame for the fact that the language of ‘consumer surplus’ has now crept into judgments on assessing damages for breach of contract in cases where the victim of the breach has suffered a loss of enjoyment (and from there has gotten into the textbooks – where even more confusingly, they sometimes distinguish between a claim in contract for distress and loss of enjoyment, and a claim for loss of one’s ‘consumer surplus’) must be laid squarely at the foot of an article published in the Law Quarterly Review in 1979: Harris, Ogus and Phillips, ‘Contract remedies and the consumer surplus’ (1979) 95 Law Quarterly Review 581. The article is not available online, so I will quote in full from the crucial passage in the article where the authors introduce the concept of ‘consumer surplus’ (while discussing the decision of Oliver J in Radford v De Froberville (1977)). The passage is at pp 582-83 of the article:

‘The existence of the wall on the new boundary of his land [in Radford v De Froberville] was worth more to the plaintiff than the increase which the wall would make to the market value of the land; it was argued that he valued the particular architectural style and privacy it afforded more highly than the average purchaser of the land. This is an example of what economists refer to as “consumer surplus”, the excess utility or subjective value obtained from a “good” over and above the utility associated with its market price. (As explained below, the consumer surplus expected by a person who intends to use a good is equivalent to the profit which a businessman expects to make from the contract.) The concept of consumer surplus is important in any attempt to measure consumer losses because, unlike firms, consumers make purchases for the pleasure or utility they confer; this utility has no necessary relationship with the price paid and is of a quite different order from market prices or business profits. It is, of course, difficult to measure utility, but generally economists avoid the conceptual problem by measuring utility in terms of the maximum amount a consumer would pay for a particular purchase. For instance, if a purchaser can buy a plot of land for £1,000, when he would be prepared to pay up to £1,500 for it, the extra £500 represents his “consumer surplus”.’

Notice how the authors end up correctly defining how a purchaser’s ‘consumer surplus’ on the purchase of land is assessed. But for the whole of the rest of the passage, ‘consumer surplus’ is simply identified with the difference between how much pleasure or enjoyment (or, in the authors’ language, ‘utility’) one will obtain from a good, and how much you have paid for that good. That is not how economists define the term ‘consumer surplus’, but unfortunately that does seem to be the definition that has crept into the cases (and from there the textbooks), via the Harris, Ogus and Phillips article.

It is time to stop the rot. Lawyers should stop talking about ‘consumer surplus’ and stick to their knitting. When they are actually talking about damages being awarded for loss of enjoyment or pleasure then they should actually talk in terms of damages being awarded for loss of enjoyment or pleasure. And students should do the same in writing essays and problem answers and avoid at all costs using the phrase ‘consumer surplus’ (unless positively required to do so by the terms of the essay or problem question).

The rule in Hadley v Baxendale

What is it?

The rule in Hadley v Baxendale basically says that if A has committed a breach of a contract that he has with B by doing x, and B has suffered a loss as a result, that loss will count as too remote a consequence of A’s breach to be actionable unless at the time the contract between A and B was entered into, A could have been reasonably been expected to foresee that his doing x was likely to result in B suffering that type of loss, because either: (1) it would have been quite normal or natural for B to suffer that type of loss as a result of A’s doing x; or (2) A was informed before he entered into the contract between him and B of any special circumstances which meant it was likely that A’s doing x would result in B suffering that type of loss.

Why do we have it?

The rule in Hadley v Baxendale is basically a rule of fairness; one of about ten different features of the English contract law that can be seen as requiring the parties to a contract to deal fairly with each other. The reason why we have the rule in Hadley v Baxendale is to give each contracting party a fair chance to decide whether or not they want to enter into the contract, and if so on what terms. A would be deprived of that chance if he were held liable for a loss suffered by B as a result of A’s breach of contract when he had no way of knowing at the time the contract was entered into that B stood to suffer that type of loss if he breached. Had A known this, and known that he might be held liable for that loss, he might have refused to enter into the contract with B, or bargained for an alteration in the terms of the A-B contract to protect himself against being held liable for that loss, or to reward him for running the risk of suffering that loss. As Alderson B remarked in Hadley v Baxendale (1854) itself, of the case where B suffers a loss as a result of A’s breach due to special circumstances that A was unaware of at the time he entered into his contract with B,

‘…had the special circumstances been known, the parties might have specially provided for the breach of contract by special terms as to the damages in that case, and of this advantage it would be very unjust to deprive them.’

A concrete example can make this point clear. Suppose that Loaded enters into a contract with Builder for the construction of a swimming pool in Loaded’s country house. The contract specifies that the swimming pool must be ready for use by June 1st. A few days after the contract is entered into, Loaded tells Builder that he needs the swimming pool to be installed by June 1st because a movie company is taking over his estate during the summer to do some filming in and around it, but they have made it a condition of their contract with him that he have a swimming pool installed as a number of crucial scenes take place in and around a swimming pool. Loaded also tells Builder that the movie company are paying him ‘crazy money’ – £5m – to hire his estate from June 15th – September 15th. Builder does not complete the swimming pool until July 1st, and Loaded’s deal with the movie company falls through.

At the time Builder breached his contract by failing to complete the swimming pool on time, it was perfectly foreseeable that his doing so would result in Loaded suffering a loss of £5m. However – unlike the position in tort law (most of the time) – we do not determine the remoteness or otherwise of a loss suffered as a result of a breach of contract by looking at what was foreseeable at the time the breach occurred. Instead, we look at what was foreseeable at the time the contract was entered into. The reason is that not to do so would be unfair on Builder. Had Builder known that his failing to complete the swimming pool on time might result in his being held liable for the loss of Loaded’s  deal with the movie company, he might have refused the job completely, or negotiated a much higher price for getting the work done on time, or insisted that there be a clause in the contract limiting the scope of his liability. So fairness demands that Builder only be held liable for the losses that he could have contemplated that Loaded might have suffered as a result of his failing to build the pool on time at the time Builder entered into his contract with Loaded, as those were the only losses Builder could have taken the risk of being held liable for when he decided to enter into a contract with Loaded, and on what terms.

The letter and the spirit of the rule

For the most part, giving effect to the letter of the rule in Hadley v Baxendale will also give effect to the spirit of fair dealing that underlies the rule. Where the two diverge, though, is where A enters into a contract with B, knowing that B is likely to suffer a particular kind of loss if A breaches that contract, but A does not factor that knowledge into his decision as to whether or not to enter into his contract with B.

If A’s breach does result in B suffering that kind of loss, the letter of the rule in Hadley v Baxendale indicates that A should be held liable for that loss: at the time A entered into his contract with B, it was reasonably foreseeable that if A breached his contract with B, then B would suffer that type of loss.

However, the spirit of the rule indicates that A should not be held liable for B’s loss – at least where he wasn’t at fault for not factoring in the prospect of B’s suffering that type of loss into his decision as to whether or not to contract with B, and if so on what terms. The reason is that holding A liable for that kind of loss would mean that he wasn’t given a fair chance to consider whether or not he should contract with B, and if so on what terms. Had A taken seriously the prospect that he might be held liable for the sort of loss that B has suffered as a result of A’s breach, he might have refused to contract with B, or have contracted on different terms. As he did not take that prospect seriously – and, apparently, acted reasonably in failing to take that prospect seriously – he should not be held liable for the loss that B has suffered.

So the letter and the spirit of the rule in Hadley v Baxendale will go in different directions in the situation where a contracting party foresees that the other party might suffer a particular type of loss if the contract was breached, but does not factor in the prospect of being held liable for that loss in deciding whether or not to enter into the contract, and if so on what terms. But when would such a situation arise?

The Achilleas (2008)– otherwise known as Transfield Shipping v Mercator Shipping – apparently presented one such situation. The defendants hired a ship from the claimants. The ship was due to be given back on May 2 2004. Expecting to get the ship back by May 2 at the latest, the claimants agreed on April 21 to hire out the ship for 191 days to Cargill International SA for $39,500 a day, with the period of hire to start once the claimants got their ship back from the defendants. Under the agreement, Cargill had the option of cancelling it if they had not received the ship by May 8. By May 5, the defendants still hadn’t handed the ship back to the claimants, and there was no prospect of the claimants getting it back by May 8. Fearful that Cargill would cancel the contract to hire the ship, the claimants renegotiated its terms, extending the cancellation date, but at the same time agreeing to a substantial reduction in the rate of hire for the ship – $31,500 a day rather than $39,500 – to reflect the fact that there had been a sharp fall in the general market rates for hiring ships like the claimants’. The claimants finally got their ship back from the defendants on May 11, nine days late.

The claimants sued the defendants for damages, arguing that ‘Had you not been late redelivering the ship, we would now be hiring out the ship for 191 days to Cargill at $39,500 a day rather than $31,500 a day. So we have suffered a loss of about $1.5m (191 x $8,000) as a result of your breach of contract, and you are liable to us for that loss.’ However, the defendants argued, ‘The custom in the industry is that when a ship is delivered back late, all the owner can sue for is the difference between what he could have earned hiring out the ship during the period the ship was wrongfully retained, and what is due under the hire contract for retaining the ship for that period of that time. So we are only liable for $158,000: the extra amount you could have made during the nine days we retained the ship compared with what we have to pay under our contract of hire for retaining the ship for those nine days.’ If the defendants were right – and this is something that is disputed – that the custom in the shipping industry on late return of a ship was simply to sue for the loss suffered as a result of not being able to hire out the ship to someone else during the period it was detained, then it would have been unfair on the defendants to hold them liable for the $1.5m loss that the claimants suffered because the defendants’ hanging on to the claimants’ ship for nine extra days resulted in the claimants losing out on the chance of hiring the ship out to Cargill for $39,500 a day, as opposed to $31,500 a day. This is because the defendants never seriously contemplated that they might be held liable for that kind of loss when they agreed to hire the claimants’ ship, and had they known that they might be held so liable, they would almost certainly have renegotiated the terms of the contract under which they hired out the claimants’ ship.

So if the defendants were right about the custom in the shipping industry, then The Achilleas was a case where the letter and the spirit of the rule in Hadley v Baxendale might have produced divergent results. According to the letter, whether or not the defendants should have been held liable for the claimants’ $1.5m loss depended on whether the defendants contemplated when they entered into the contract with the claimants that their hanging on to the claimants’ ship beyond the hire period would result in the claimants suffering the kind of loss on the follow-on contract of hire that they suffered here.

On the other hand, if we followed the spirit of the rule in Hadley v Baxendale, then we would regard it as irrelevant whether or not the defendants knew what losses might be suffered by the claimants if the defendants held on too long to the claimants’ ship. If the custom in the industry was such as the defendants described, then the scope of the defendants’ liability should have been determined by that custom. To hold the defendants liable on any other basis, and hold them liable for losses they contemplated the claimants might suffer as a result of breach when they hired the claimants’ ship, would be unfair on the defendants as they never seriously contemplated that they might be held liable for those losses, and did not factor in the possibility that they might be held liable for those losses when they decided to hire the claimants’ ship on the terms they did.

As it happens, the House of Lords did not think that the letter and the spirit of Hadley v Baxendale did diverge in The Achilleas as to what the result of the case should be. Lord Rodger applied the letter of the rule in Hadley v Baxendale and found that at the time the defendants hired the claimants’ ship, there was no reason for them to contemplate that a delay in returning the ship would result in the claimants suffering the type of loss that they had suffered on the follow-on contract as the loss was purely due to ‘unusual’ (at [53]) movements in the market rates for hiring ships. Lord Hoffmann’s approach to the case (which I will discuss in more detail below) was much more consistent with the spirit of Hadley v Baxendale and was – for the reasons explained above – consequently not in favour of holding the defendants liable for the losses suffered by the claimants on their follow-on contract. Baroness Hale agreed with Lord Rodger’s approach, though with some doubts about how it applied in this case. The fact that both approaches resulted in the same outcome allowed Lords Hope and Walker to agree with Lord Rodger and Lord Hoffmann, thus resulting in The Achilleas producing no overall majority in favour of whether the letter or the spirit of Hadley v Baxendale should be followed when they diverge.

And there will be cases where they will diverge in terms of the result they reach. For example: Executive hails a taxi driven by Driver. Executive tells Driver, ‘Take me to the airport. I have a plane to catch in two hours. A huge deal is riding on my making the plane.’ Shortly afterwards, Driver carelessly crashes the taxi. Executive is unharmed but is unable to make his plane, and fails to close the business deal that he was flying out to negotiate; the deal was worth $5m to Executive. Can Executive sue Driver for this loss? According to the letter of the rule in Hadley v Baxendale, yes he can. At the time Driver let Executive into his car, he knew that if he screwed up driving Executive to the airport, that Executive would suffer this kind of loss. According to the spirit of the rule in Hadley v Baxendale, Executive shouldn’t be able to sue Driver for the loss of his deal. At the time Driver let Executive into his car, Driver wasn’t factoring the possibility that he might be sued for that kind of loss into his decision as to whether or not to take Executive to the airport. In fact, Driver wasn’t making any kind of conscious decision as to whether or not to accept Executive as a passenger – as soon as Executive hailed his cab, he was going to take Executive as a passenger, whatever Executive said. Given this, it would be unfair to hold Driver liable for the loss of Executive’s deal. If Executive wanted to be able to sue Driver for that kind of loss, he should have been much more explicit with Driver: ‘Take me to the airport. I have a plane to catch in two hours. There is a huge deal riding on this, and I will sue you for millions if something happens to stop me catching the plane. Are you happy to take me on that basis?’ But if he said something like that, then Driver’s reaction would almost certainly have been: ‘No, I’m not – hop off mate and find someone else to take you.’ Holding Driver liable for the loss of Executive’s deal when that would have been Driver’s reaction had the possibility of his being held liable for that loss been brought home to him illustrates just why the spirit of the rule in Hadley v Baxendale stands opposed to holding Driver liable for that loss.

Where the letter and spirit of the rule in Hadley v Baxendale diverge, which should we prefer? In favour of following the spirit is simple common sense – the letter is supposed to serve the spirit, and must give way when it fails to do this. But in favour of following the letter in all cases is the desire for commercial certainty. Applying the letter of the rule in Hadley v Baxendale across the board may do injustice in individual cases such as Executive v Driver, but it does at least allow litigants in breach of contract cases to know where they stand so far as their potential liabilities are concerned. In contrast, applying the spirit of the rule in Hadley v Baxendale to determine the scope of a contract breaker’s liabilities requires the court to make difficult inquiries into the contract breaker’s expectations when he entered into the contract as to what he might be held liable for if he breached the contract.

Where Lord Hoffmann went wrong in The Achilleas

I said in the preceding section that Lord Hoffmann’s judgment in The Achilleas was ‘consistent’ with the spirit of the rule in Hadley v Baxendale, as I have explained it above. However, this is not quite true. According to the spirit of the rule in Hadley v Baxendale, a defendant should not be held liable for a loss that he did not take the risk of being held liable for when he entered into his contract with the claimant. (Compare the argument of counsel for the defendants in Hadley v Baxendale: ‘Where the contracting party is shewn to be acquainted with all the consequences that must of necessity follow from a breach on his part of the contract, it may be reasonable to say he takes the risk of such consequences.’) Lord Hoffmann said something different in The Achilleas. He said that a defendant should not be held liable for a loss that he did not agree to be held liable for when he entered into the contract (all emphases added):

[12] It seems to me logical to found liability for damages upon the intention of the parties (objectively ascertained) because all contractual liability is voluntarily undertaken. It must be in principle wrong to hold someone liable for risks for which the people entering into such a contract in their particular market, would not reasonably be considered to have undertaken.

[15] …one must first decide whether the loss for which compensation is sought is of a ‘kind’ or ‘type’ for which the contract-breaker ought fairly be taken to have accepted responsibility.

[26] …[in this type of case] the court is engaged in construing the agreement to reflect the liabilities which the parties may reasonably be expected to have assumed and paid for.

The idea that Lord Hoffman seems to be advancing here – that a contract-breaker’s liabilities to compensate the victim of his breach are attributable to the fact that he has agreed to assume those liabilities in entering into the contract – should be rejected as heretical. As Kyle Lawson has brilliantly pointed out (in ‘The remoteness rules in contract: Holmes, Hoffmann, and ships that pass in the night’ (2012) 23 King’s Law Journal 1), Lord Hoffmann’s language in The Achilleas echoes that of Holmes J in deciding the case of Globe Refining v Landa Cotton Oil (1903) in the US Supreme Court. In that case, Holmes J said that:

‘When a man commits a tort he incurs by force of the law a liability to damages, measured by certain rules. When a man makes a contract he incurs by force of the law a liability to damages, unless a certain promised event comes to pass. But unlike the case of torts, as the contract is by mutual consent, the parties themselves, expressly or by implication, fix the rule by which the damages are to be measured… [In considering what the plaintiff is entitled to recover in this case we] have to consider…what liability the defendant fairly may be supposed to have assumed consciously, or to have warranted the plaintiff reasonably to suppose that it assumed, when the contract was made.’

Holmes’ theory of the basis of a contract breaker’s liability followed from his view – famously expressed in ‘The path of the law’ (1896-7) 10 Harvard Law Review 457 – that someone who commits a breach of contract does not actually do anything legally wrong: ‘The duty to keep a contract at common law means a prediction that you must pay damages if you do not keep it – and nothing else. If you commit a tort, you are liable to pay a compensatory sum. If you commit a contract, you are liable to pay a compensatory sum unless the promised event comes to pass, and that is all the difference’ (ibid, 462). So the only obligation that a contracting party undertakes is either to ensure that something happens or to pay damages instead. It follows that the contract breaker’s obligation to pay damages is traceable to the fact that the contract breaker undertook to pay such damages if he failed to perform.

All this is nonsense. It is not possible to argue that a contract breaker’s liability to pay damages to the victim of his breach is attributable to the fact that the contract breaker agreed to pay those damages if he did not perform. A couple of examples show this:

(1) A enters into a contract with B. Under the contract, A undertakes to pay B the penal sum of £1m if he breaches his contract in any way at all. On breach, B will not be allowed to sue A for £1m but will instead be confined to suing for the actual loss he has suffered as a result of breach. This is so even though A undertook to pay B £1m if he breached his contract with B, and not to compensate B for his actual loss.

(2) A enters into a contract with B. The contract specifies that A’s liability on breach will be capped at £5,000, whatever the nature of A’s breach, and whatever the extent of B’s losses as a result of A’s breach. It is held that the clause limiting A’s liability is invalid under some statutory provision. A is then held liable for the actual loss suffered by B as a result of A’s breach, even though A never agreed to be held liable for that loss, but only £5,000.

Examples like these show that a contract breaker’s liability to pay damages to the victim of his breach is imposed on him by the law, and is not assumed by him under the contract. So the rule in Hadley v Baxendale cannot be explained as existing to give effect to a defendant’s intentions at the time he entered into a contract as to what liabilities he was agreeing to assume under that contract. Instead, the rule operates to prevent the law imposing on the defendant a liability to compensate the claimant for a loss that the defendant did not take the risk that the law might hold him liable to compensate the claimant for that loss when he contracted with the claimant. Any suggestion to the contrary in Lord Hoffmann’s judgment in The Achilleas is to be regretted.

(Though see para [13] of his judgment, which is entirely consistent with the analysis of the basis of the rule in Hadley v Baxendale advanced here: ‘The view which the parties take of the…risks they are undertaking will determine the other terms of the contract and in particular the price paid. Anyone asked to assume a large and unpredictable risk will require some premium paid in exchange. A rule of law which imposes liability upon a party for a risk which he reasonably thought was excluded gives the other party something for nothing.’)

The fact that Lord Hoffmann adopted an unworkable theory of the basis of contractual liability in The Achilleas should not lead us to think that his refusal simply to give effect to the letter of the rule in Hadley v Baxendale in The Achilleas was a mistake. As we have seen, the letter and the spirit of the rule in Hadley v Baxendale can diverge and in such a situation, a case can be made for adhering to the spirit rather than the letter of the rule.

Intention to create legal relations

The fictional nature of this area of law

An agreement will not be legally binding unless it was intended to be legally binding. However, the only cases where there will be an issue as to whether or not the parties to an agreement intended it to be legally binding are cases where it is impossible to tell whether or not they intended it to be legally binding.

For example, in Granatino v Radmacher (2010) – where the UK Supreme Court had to decide whether or not to give effect to the parties’ pre-nuptial agreement that neither party would seek to make a claim to property belonging to the other party if they got divorced – there was no issue as to whether that agreement was intended to be legally binding. It was obvious that it was intended to be binding. The only issue the UK Supreme Court had to decide was whether the public interest demanded that they should refuse to give effect to it. (On which issue, they split 8:1 in favour of giving effect to the pre-nuptial agreement so long as it was freely entered into, and did not prejudice the interests of the children to the marriage.)

Again, in Cobbe v Yeoman’s Row (2008) – where the claimant’s work in obtaining planning permission to develop the defendant’s land was wasted because the defendant pulled out of a provisional agreement to sell his land to the claimant – there was no issue as to whether or not the claimant and the defendant intended their provisional agreement to be legally binding. It was obvious that they did not. The only issue the House of Lords had to decide was whether the claimant was entitled to a remedy even though he had not yet entered into any kind of contract with the defendant. (On which issue, the House of Lords ruled that the claimant was entitled to a reasonable sum – a quantum meruit – for the work he had done in trying to obtain the planning permission.)

So the issue as to whether or not the parties to an agreement intended it to be legally binding only ever comes up in cases where it is impossible to resolve that issue because what the parties intended is shrouded in obscurity. It follows that in a case – like Balfour v Balfour (1919) and Jones v Padavatton (1969) – where there is an issue as to whether the parties to an agreement intended it to be legally binding, when the court tries to determine whether or not that agreement was intended to be legally binding, that is not, and cannot be, what it is actually doing. This is because it is impossible to determine whether or not the parties intended that their agreement be legally binding in these kinds of cases.

So what were the courts doing in cases like Balfour v Balfour or Jones v Padavatton? The answer is pretty simple – They were trying to decide whether or not to find that the agreement reached in those cases was legally binding. In other words, in a case like Balfour v Balfour or Jones v Padavatton, the court was trying to decide whether the reasons for finding that the agreement was legally binding outweighed the reasons against.

So when the courts say that they will ‘presume’ that domestic agreements are not intended to be legally binding, and that they will ‘presume’ that commercial agreements are intended to be legally binding, what they are really saying is that in a case where it is impossible to tell whether or not an agreement was intended to be legally binding, the balance of reasons applying to the case will usually come down in favour of refusing to enforce a domestic agreement and will usually come down in favour of enforcing a commercial agreement.

I’ll now substantiate this argument by looking at Balfour v Balfour and Jones v Padavatton in more detail.

Balfour v Balfour (1919)

In Balfour v Balfour, a husband and wife were based in Ceylon, where the husband had a job. They came back to England when the husband was on leave. When his period of leave expired, he went back to Ceylon; on the other hand, his wife stayed in England as she was too ill to travel. The husband promised to send his wife £30 a month while she had to stay in England. After he got back to Ceylon, he wrote to his wife saying that he thought it would be better if they remained apart. He subsequently failed to keep up the promised maintenance payments of £30 a month, and so his wife sued him for the money.

It seems not to have been noticed that the wife had supplied no consideration for the husband’s promise (though Warrington LJ observed that ‘The wife…made no bargain at all’). Instead, the case was regarded as turning on the issue of whether the husband’s promise had been intended to be legally binding. This was obviously impossible to resolve (otherwise there would have been no issue as to whether the promise had been intended to be legally binding) and so, in ‘resolving’ this issue, the Court of Appeal was really determining whether or not they ought to find that this agreement was binding.

In favour of finding that the agreement was binding was the fact that the husband had a moral responsibility not to leave his wife destitute and that he had acknowledged that responsibility in promising to send her £30 a month for her maintenance. (This was the reason why the first instance judge had found in favour of the wife.) However, the Court of Appeal thought that this reason was outweighed by two reasons against finding that the agreement was binding.

The first was that finding that this agreement was binding would fix the husband’s obligation to pay maintenance at £30 a month when a change of circumstances might mean that it was important that he pay his wife more, or might mean that he could only afford to pay his wife less than this sum. For Warrington LJ this made it ‘quite impossible’ to find that the husband’s promise to pay his wife was legally binding: ‘If we were to imply…a contract in this case we should be implying on the part of the wife that whatever happened and whatever might be the change of circumstances while the husband was away she should be content with this £30 a month, and bind herself by an obligation in law not to require him to pay anything more; and on the other hand we should be implying on the part of the husband a bargain to pay £30 a month for some indefinite period whatever might be his circumstances… it seems to me that it would be impossible to make any such implication.’ So finding that the husband’s promise here was legally binding would freeze an element of the husband’s relationship with his wife – how much he would give her by way of support – that needed to remain flexible if the husband and wife’s relationship were to be able to cope with unexpected changes of circumstances

Secondly, the Court of Appeal was concerned that finding that a promise like this was legally binding would be ‘a possible fruitful source of dissension and quarrelling…it would lead to unlimited litigation in a relationship which should be obviously as far as possible protected from possibilities of that kind’ (per Duke LJ). This is because turning a married couple into contracting parties might have a damaging effect on the nature of their relationship. Marriage is supposed to be a partnership focussed on achieving shared goals together, while people in a contractual relationship use each other in order to achieve their own goals. So finding that a married couple are in a contractual relationship might encourage them to stop thinking of themselves as being in a partnership and start thinking of each other in more antagonistic terms. Their relationship would become less ‘What can I do for you?’ and more ‘What have you done for me lately?’ It was because of these kinds of concerns that Atkin LJ famously observed that: ‘Agreements such as these are outside the realm of contracts altogether. The common law does not regulate the form of agreements between spouses. Their promises are not sealed with seals and sealing wax. The consideration that really obtains for them is that natural love and affection which counts for so little in these cold Courts.’

On balance, then, the Court of Appeal thought it would be undesirable to find that the promise in this case was legally binding. Of course, they dressed up their conclusion in terms of a finding that the promise was not ‘intended’ to be legally binding – but that was not what they were really deciding, as it was completely unclear whether or not the promise was intended to be legally binding.

Jones v Padavatton (1969)

In Jones v Padavatton, a mother was suing to recover possession of a house from her daughter. The mother lived in Trinidad. She had bought the house for her daughter to live in while her daughter studied for her Bar exams. In 1962, the mother had persuaded the daughter to give up a good job working in the Indian Embassy in Washington DC, and travel to England with her son in order to qualify as a barrister. (The mother’s plan was that once her daughter became a barrister, she would come back to Trinidad and practise there.) The mother had promised to give the daughter an allowance of (what turned out to be – there was some misunderstanding between the mother and daughter as to how much she would get) £42 a month to help cover her expenses while she studied for the Bar.

The mother subsequently suggested – and the daughter agreed – that instead of paying her £42 a month, she would instead buy a house for the daughter to live in with her son, and the daughter could let out rooms in the house and use the rent money to cover her expenses. The house was purchased in 1964. In 1967, the mother – unhappy that her daughter had provided her with no information as to how much money she was making by letting out rooms in the house – travelled to England to find out what was going on, and commenced proceedings to recover possession of the house from her daughter. The daughter argued that the mother could not do this, as she was bound by her agreement to allow her daughter to live in the house so long as she was studying for the Bar, and the daughter had still not passed all her Bar exams. (In fact, even in 1969, when the case was considered by the Court of Appeal, the daughter had still to do some of her Part I exams, and had passed none of the Part II exams.)

The Court of Appeal disagreed over whether there was a binding contract between the mother and daughter in this case. Salmon LJ thought that the Court should find that the mother’s promise to support her daughter was legally binding. The fact that the mother had induced her daughter to give up a very good job in Washington and travel to England by making that promise was a very weighty consideration in favour of finding that the mother’s promise of support was legally binding on her: ‘I cannot think that either [the mother or the daughter] intended that if, after the daughter had been in London, say, for six months, the mother dishonoured her promise and left her daughter destitute, the daughter would have no legal redress.’ However, he thought that the mother’s promise of support was not unlimited:

‘The parties cannot have contemplated that the daughter should go on studying for the Bar and draw the allowance until she was seventy, nor on the other hand that the mother could have discontinued the allowance if the daughter did not pass her examinations within, say, eighteen months. The promise was to pay the allowance until the daughter’s studies were completed, and to my mind there was a clear implication that they were to be completed within a reasonable time.’

Salmon LJ thought that five years was a ‘reasonable time’ for passing the Bar exams, and so by the time the mother came to England, the five years was up. The mother was therefore no longer bound by then by her promise of support, and could recover possession of the house.

Danckwerts and Fenton Atkinson LJJ thought that the mother’s promise of support was never legally binding. Fenton Atkinson LJ seemed to think that there was actually no issue over whether the promise of support was intended to be legally binding: he thought it was obvious from the evidence (in particular, the open-ended and imprecise nature of the arrangements between the mother and the daughter, and the daughter’s being aggrieved at the idea of her mother taking legal action against her) that the mother’s promise of support was not intended to be legally binding and that the daughter was simply trusting her mother to come through with the promised support, and the mother was trusting her daughter to get through the Bar exams reasonably quickly. Danckwerts LJ thought that the need for flexibility in how the mother would support her daughter counted against the Court’s finding that the mother’s promises to her daughter were legally binding: ‘What was required was an arrangement which was to be financed by the mother, and was such as would be adaptable to circumstances, as it in fact was… It was not a stiff contractual operation…’

The law on estoppel

Introduction

The law on estoppel can be quite intimidating to students. The biggest problem they have with it is the language – what are we talking about when we talk about estoppel? Lord Denning MR’s judgment in McIlkenny v Chief Constable of the West Midlands (1980) (quoted by John Cartwright in his excellent article on estoppel, ‘Protecting legitimate expectations and estoppel in English law’ 10.3 Electronic Journal of Comparative Law (December 2006)) provides some useful guidance:

‘The word “estoppel” only means stopped… It was brought over by the Normans. They used the old French “estoupail.” That meant a bung or cork by which you stopped something from coming out. It was in common use in our courts when they carried on all their proceedings in Norman-French. Littleton writes in the law-French of his day (15th century) using the words “pur ceo que le baron est estoppe a dire,” meaning simply that the husband is stopped from saying something.’

So the law on estoppel is essentially concerned with situations where the law will stop or prevent someone from doing something that they would otherwise be entitled to do. The law on estoppel has many different branches, as Lord Denning went on to explain:

‘…there has been built up over the centuries in our law a big house with many rooms. It is the house called Estoppel. In Coke’s time it was a small house with only three rooms, namely, estoppel by matter of record, by matter in writing, and by matter in pais. But by our time we have so many rooms that we are apt to get confused between them. Estoppel per rem judicatam, issue estoppel, estoppel by deed, estoppel by representation, estoppel by conduct, estoppel by acquiescence, estoppel by election or waiver, estoppel by negligence, promissory estoppel, proprietary estoppel, and goodness knows what else. These several rooms have this much in common: They are all under one roof. Someone is stopped from saying something or other, or doing something or other, or contesting something or other. But each room is used differently from the others. If you go into one room, you will find a notice saying, “Estoppel is only a rule of evidence.” If you go into another room you will find a different notice, “Estoppel can give rise to a cause of action.” Each room has its own separate notices. It is a mistake to suppose that what you find in one room, you will also find in the others.’

In this note, I want to explore three of the rooms that make up the Estoppel house.

Estoppel by representation

This branch of the law on estoppel really provided the springboard for all future developments in this area of the law. The law on estoppel by representation basically says this: If A and B are litigating a case in court, and A wants to deny that x is true, A will be prevented (or estopped) from doing so if (1) A had previously represented to B that x was true, (2) A intended that B should rely on that representation, and (3) B did rely on that representation. The rule is a rule of evidence – it is a rule about what A can and cannot say in court. Under this rule, A is prevented from playing fast and loose with the truth. He cannot outside court say one thing to B, with the intended effect that B relied on what A said, and then go into court and deny that what he said to B was true.

As a mere rule of evidence, the law on estoppel by representation does not give B a ground for suing A, but as Brandon LJ observed in Amalgamated Investment and Property Co Ltd (1982), B ‘may, as a result of being able to rely on an estoppel, succeed on a cause of action on which, without being able to rely on that estoppel, he would necessarily have failed.’ Here are two examples of how relying on an estoppel by representation could allow B to win a case against A that would otherwise fail.

A owns a large area of land. A tells B that a cottage located on A’s land now belongs to B. B relies on that representation in various ways. In fact, A has done nothing to transfer title to the cottage into B’s hands. A and B later fall out, and while B is away on holiday, A has the locks changed on the cottage and B’s belongings thrown out of the cottage. B wants to sue A for trespass to land. A wants to deny trespass: he wants to argue he was justified in acting as he did as the cottage belongs to him. But he is not allowed to plead this in court, as it would involve going back on a representation that he made to B (‘this cottage now belongs to you’) which B relied on, as A intended he should. So an estoppel by representation arises which prevents A justifying his behaviour, and as a result B is able to sue A for trespass to land.

B takes a dress to be cleaned in A’s dry cleaners. A has B sign a form, setting out the terms on which A is prepared to dry clean B’s dress. B queries one of the terms under which B undertakes not to sue A if the dress is returned to her in a damaged condition. A tells B, ‘Don’t worry about that – that only applies if we are cleaning dresses that have lots of beads and sequins on them: we can’t guarantee our cleaning process won’t damage that sort of thing. But your little black dress will be no problem.’ When B picks up the dress, it is stained. B sues A for breach of contract. A will want to defend the claim by arguing that under the standard terms on which B and A were dealing, B agreed not to sue if her dress was damaged. But A will be estopped from raising such a defence: A told B that the term in question did not apply to the job of cleaning B’s dress, and B relied on that representation by agreeing to allow B to clean the dress. A cannot now go back on that representation and argue that the term did after all apply in B’s case. As a result, B will be able to sue A for breach of contract (assuming that the damage to the dress was the result of a lack of skill and care in the way A cleaned the dress, which presumably it was).

So the law on estoppel by representation can help a claimant sue a defendant by providing the claimant with one of the building blocks that she needs to make her case. But all the other building blocks need to be in place if the claimant is going to winher case. For example, suppose that by making various cryptic remarks and suggestions, A leads B to believe that A is promising to leave B all his property in his will. B relies on that belief – as A intended that she should – by looking after A in his old age. After A dies, it turns out that A has left B nothing in his will. B now wants to claim all of A’s property. The law on estoppel will kick in to prevent A’s representatives from denying that A promised to leave B all his property in his will. But that won’t get B over the winning line. B still has to show that A’s promise to give her all his property is legally binding, or has some legal effect, and the law on estoppel by representation has nothing to say about that. But as we’ll see, other areas of the law on estoppel do have something to say about that.

Promissory estoppel

At some point, the idea underlying the law on estoppel by representation – that you could stop someone going back on a representation if they had intentionally induced someone to rely on that representation – gave rise to the idea that the law on estoppel could also be used to prevent someone from going back on a promise if they had intentionally induced someone to rely on that promise. Thus was born the law on promissory estoppel.

In England, the law on promissory estoppel only applies to prevent someone from going back on a promise that they have made not to enforce some legal right that they have against someone else. So just as the law on estoppel by representation is based on the idea that you can’t play fast and loose with the truth, the law on promissory estoppel – in England at any rate – is based on the idea that you can’t play fast and loose with your rights. If you have a right against someone, you can’t tell them that you won’t enforce it, with the effect that they rely in some way on your promise, and then turn round and say that you’ve changed your mind and that you will be standing on your strict legal rights after all. This principle was first articulated by Lord Cairns LC in Hughes v Metropolitan Railway (1877):

‘[if the parties to a contract] enter upon a course of negotiation which has the effect of leading one of the parties to suppose that the strict rights arising under the contract will not be enforced, or will be kept in suspense, or held in abeyance, the person who otherwise might have enforced those rights will not be allowed to enforce them where it would be inequitable having regard to the dealings between those parties.’

In that case, the defendants leased some properties from the claimants. The claimants served a notice of repair on the defendants, which basically gave the defendants six months to carry out certain repairs to the let properties, or face forfeiting their lease. The defendants weren’t enthusiastic about carrying out the repairs and suggested that the claimants buy out their lease instead. The claimants asked the defendants how much they wanted to give up their lease, and the defendants replied ‘£3,000.’ The claimants said, ‘That’s too much – come back to us with a more reasonable offer.’ The defendants never replied. The six months that the defendants had been given to repair the premises expired without the repairs having been carried out, and the claimants sued to eject the defendants from the premises, claiming that they had forfeited the lease by not carrying out the repairs. The House of Lords held that the claimants could not do this. They had impliedly promised that they wouldn’t enforce the defendants’ obligation to repair the premises while negotiations for the defendants to surrender the premises were still going on, and those negotiations had never been definitively broken off.

For one reason or another, the law on promissory estoppel – as it was set out in Hughes – slumbered for about 70 years until it was revived by a first instance judge called Alfred Denning in the case of Central London Property Trust Ltd v High Trees House Ltd (1946) (popularly known as the ‘High Trees case’). That was another case about a lease. High Trees leased a block of flats in London from Central London Property Trust (‘CLPT’) in 1937, on a nine year lease, at a rent of £2,500 a year. High Trees then let out the individual flats to tenants. When World War II broke out, and London came under heavy bombing, High Trees found it harder and harder to find tenants who were willing to rent their flats. Accordingly, CLPT agreed in 1940 to halve the rent payable by High Trees, to £1,250 a year. And that is what High Trees paid for the rest of the war. By the time the war ended in June 1945, CLPT was in receivership (which meant its affairs were being administered by a receiver, appointed by the company’s creditors to get as much of their money back) and the receiver went to court to find out: (1) whether High Trees were now liable to pay rent at £2,500 a year; and (2) whether High Trees had been liable during the course of the war to pay rent at £2,500 a year and were therefore liable for about £8,000, which was the difference between the rent that High Trees should have paid if they had been liable to pay £2,500 a year during the course of the war, and what High Trees did pay as they only thought that they had to pay £1,250 a year.

Denning J ruled that CLPT could now put the rent back up to £2,500 a year, but that it would not be allowed to go back on its promise that it would not sue High Trees for the full amount of the rent during the course of the war. He argued that a promise not to sue for the full amount of a debt that was owed to the promisor would be binding if ‘it was intended to be binding, intended to be acted on, and in fact acted on… The courts have not gone so far as to give a cause of action in damages for breach of such a promise, but they have refused to allow the party making it to act inconsistently with it. It is in that sense, and that sense only, that such a promise gives rise to an estoppel… The logical consequence, no doubt, is that a promise to accept a smaller sum in discharge of a larger sum, if acted upon, is binding notwithstanding the absence of consideration…’. Denning J’s judgment in High Trees has given rise to three issues.

(1) Reliance

Denning J made it clear that CLPT’s promise to accept half the rent owing to it under the rental agreement was binding on it because their promise had been relied on by High Trees. But how? What did High Trees do differently that they would not otherwise have done had CLPT not promised to halve their rent?

It was suggested by Arden LJ in Collier v Wright (2008) that in a case where B owes A £1,000 and A tells B that if B pays him £500, he will forget the rest, B’s act of paying A £500 can count as reliance on A’s promise that he will not pursue B for the full amount of the debt. This is difficult to accept. In order for B to argue that when she paid A £500, she was relying on A’s promise not to sue her for the full amount, she would essentially have to argue that, ‘Had A not promised not to sue me for the full amount, I wouldn’t have paid him anything. So A’s promise induced me to do something – pay him half the amount I owed him – that I would not otherwise have done.’ It is not clear that a court would, or should, allow B to say something like that.

In suggesting that part payment of a debt could count as reliance on a promise not to sue for the full amount, Arden LJ was simply following the judgment of Lord Denning MR in the case of D & C Builders v Rees (1965) (builders who were owed £480 for work they had done on the defendants’ property were told that the defendants could only afford to give them £300 in full satisfaction of the defendants’ debt; they accepted the £300 on that basis and then tried to sue for the balance). In that case, Lord Denning said that ‘Where there has been a true accord, under which the creditor voluntarily agrees to accept a lesser sum in satisfaction, and the debtor acts upon that accord by paying the lesser sum and the creditor accepts it, then it is inequitable for the creditor afterwards to insist on the balance. But he is not so bound unless there has truly been an accord between them’ (emphasis in original). (In D & C Builders itself, there was no such ‘true accord’ as the promise not to sue for the full amount had been extorted under pressure.)

There seems no doubt – on the basis of this dictum – that Lord Denning thought that High Trees had relied on CLPT’s promise not to sue for the full amount of the rent by paying half the rent that they owed. But the problem remains: someone who pays half of what they owe can only be said to be relying on a promise that the balance will not be sued for if, in the absence of that promise, they would have paid less than half, or even nothing at all. It is difficult to see how High Trees could have argued that in the High Trees case.

(2) Suspensory, not extinctive

The fact that in High Trees, CLPT was held to be entitled to go back to charging High Trees the full rent of £2,500 a year once the war was over has given rise to the impression that promissory estoppel only operates to suspend, and not extinguish someone’s strict legal rights. There are two reasons why it is misleading to read High Trees as saying that promissory estoppel has a ‘suspensory, not extinctive’ effect on people’s rights.

First, CLPT’s rights to the full amount of the rent that was payable during the war were definitively extinguished under the law on promissory estoppel. They simply could not get that rent back, ever.

Secondly, Denning J’s judgment in High Trees makes it abundantly clear that the reason why CLPT could go back to charging High Trees rent of 2,500 a year once the war was over had nothing to do with any feature of the law on promissory estoppel, but was simply because CLPT’s agreement to accept £1,250 a year in rent from High Trees was only meant to apply while High Trees found it difficult to find tenants for their flats. As Denning J observed: ‘I am satisfied that [CLPT’s] promise was understood by all parties only to apply under the conditions prevailing at the time when it was made, namely, when the flats were only partially let, and that it did not extend any further than that. When the flats became fully let, early in 1945, the reduction ceased to apply.’

However, there is one sense in which the law on promissory estoppel can apply simply to suspend someone’s rights, rather than extinguish them. This sense was identified by the Privy Council in Ajayi v R T Briscoe (1964), in which case the Privy Council ruled that if A promises not to enforce his strict contractual rights against B, with the result that B relies on that promise, ‘[A] can resile from his promise on giving reasonable notice, which need not be a formal notice, giving [B] a reasonable opportunity of resuming his position’. So if B can, with sufficient notice, undo his act of reliance on A’s promise that he would not enforce his strict legal rights against B, then A will only be prevented from enforcing his strict legal rights for as long as is needed to give B a chance to undo whatever it is she did in reliance on A’s promise. It is only in the case where B cannot ‘resume’ her original ‘position’ that A’s promise not to enforce his strict legal rights against her will become ‘final and irrevocable’.

These dicta from Ajayi make the position adopted by Arden LJ in Collier v Wright (and, before her, Lord Denning MR in D & C Builders v Rees) that part payment of a debt can count as a sufficient act of reliance on a promise not to sue for the full amount even more difficult to sustain. If undoing the promisee’s act of reliance allows the promisor once again to enforce his strict legal rights against the promisor, then if the only thing B has done to rely on A’s promise that he would not sue her for the full £1,000 that she owes him is to pay A half of that amount, then it seems that A could escape being estopped from suing B for the money that she owes him by the simple expedient of paying B back her £500. If he does that, they are back where they started and A can then sue B for the £1,000 that she owed him in the first place. But instead of making A repay B the £500 she paid him, and then allowing him to sue her for the full £1,000 that she owed him in the first place, it would seem simpler to allow A to keep the £500 and hold that the law on promissory estoppel does not prevent him suing B for the remaining £500. This was the position taken by Viscount Simonds in Tool Metal Manufacturing Co Ltd v Tungsten Electric Co Ltd (1955) where he observed that: ‘the gist of the equity [that arises in promissory estoppel cases] lies in the fact that one party has by his conduct led the other to alter his position. I lay stress on this because I would not have it supposed, particularly in commercial transactions, that mere acts of indulgence are apt to create rights.’

(3) A shield, not a sword

As I have already said, the law on promissory estoppel in England only works to prevent a promisor going back on a promise that they would not enforce their strict legal rights against the promisee. As such, the law on promissory estoppel in England operates as a ‘shield’, protecting defendants from being sued by claimants who have previously promised not to sue the defendant, and does not operate as a ‘sword’, allowing claimants to sue defendants for breaking a promise that the defendant made to them. Denning J acknowledged as much in High Trees when he said of a promise that was ‘intended to be binding, intended to be acted on, and in fact was acted on’ that ‘The courts have not gone so far as to give a cause of action in damages for breach of such a promise, but they have refused to allow the party making it to act inconsistently with it.’ However, the end of his judgment could have been read as leaving the door open to claimants to sue for breach of such a promise: ‘a promise intended to be binding, intended to be acted on and in fact acted on, is binding so far as it terms properly apply.’

However, Denning slammed that door shut in the subsequent case of Combe v Combe (1951). In that case, a husband and wife were divorcing, and the husband promised to pay the wife £100 a year in maintenance. The promise was not made with the object of persuading the wife not to seek an official order for maintenance from the Divorce Court, so when the wife failed to seek such an order, her failure did not provide consideration for the husband’s promise. When the now ex-husband failed to come through with the promised maintenance payments, his ex-wife took him to court. As she could not argue that his promise to pay her £100 a year was contractually binding on him, she was forced to argue that the law on promissory estoppel prevented him going back on his promise. By this time Alfred Denning had been promoted to the Court of Appeal, and Denning LJ ruled that the wife’s claim must fail:

‘Much as I am inclined to favour the principle stated in the High Trees case, it is important that it should not be stretched too far, lest it should be endangered. That principle does not create new causes of action where none existed before. It only prevents a party from insisting upon his strict legal rights, when it would be unjust to allow him to enforce them, having regard to the dealings which have taken place between the parties.’

Since Combe v Combe was decided, the line that promissory estoppel only works to stop someone going back on a promise not to enforce his strict legal rights against the promisee, and not to stop someone going back on any other type of promise, has been pretty solidly adhered to by the English courts. The only slight weakening came in Williams v Roffey Bros & Nicholls (1991), when Russell LJ said that he would have ‘welcomed’ the opportunity to consider whether a contractor was estopped from going back on a promise to pay a sub-contractor more for getting work done on time. But otherwise it seems to be accepted (as it was by the Court of Appeal in Baird Textiles v Marks & Spencer plc (2002)) that only the UK Supreme Court can change the law so that promissory estoppel can operate as a sword, and not just a shield.

The position is, of course, different in the United States (under s 90 of the Restatement 2d of Contracts)and Australia (thanks to the decision of the High Court of Australia in Waltons Stores (Interstate) Ltd v Maher (1988)). In those countries, if A makes a promise to B with the foreseeable effect that B relies on that promise in some way, B may be entitled to seek some remedy against A if A attempts to resile from his promise. As it is the fact that B has relied on A’s promise that impels the American and Australian courts to grant B a remedy, one would expect that remedy to be tailored towards ensuring that B is left no worse off as a result of relying on A’s promise. The easiest way of doing that is to order A to perform his promise, or put B in the financial position she would have been in had that promise been performed – on the basis that B will not regret relying on A’s promise if the promise is actually kept, or is as good as kept – and that tends to be the remedy (a remedy protecting B’s ‘expectation interest’) that the American and Australian courts have adopted in promissory estoppel cases even though a lesser remedy targeted at just giving back to B what she has given up as a result of relying on A’s promise (a remedy protecting B’s ‘reliance interest’) would sometimes be just as effective in protecting B’s position, and less burdensome to A.

Proprietary estoppel

We now enter the third of the rooms in the house of Estoppel that this essay wants to look at. This room contains the law on proprietary estoppel, which applies when B has relied on a belief that she has, or will have, an interest in land that belongs to A.

If we go back to the section of this essay dealing with the law on estoppel by representation, we saw there that if B relies on A’s representation that B has an interest in land that currently belongs to A, B can take advantage of the fact that A is estopped from going back on that representation to bring a claim against A for trespass to land. It was not so much of a leap for the law to say that in such a case B could also bring a claim asking the court to recognise that she actually did have the interest in A’s land that A told her that she had. Obviously, this could not be done through the law on estoppel by representation because, as we have seen, that area of the law is simply concerned with rules of evidence on what someone can and cannot say in court. So the law on proprietary estoppel was born, under which area of law claimants could go to court to ask for the court to recognise that they had an interest in land when they had been led to believe by the landowner that they had such an interest and had relied on that belief.

The law on proprietary estoppel, unlike the law on promissory estoppel, could therefore work as a sword, not just as a shield – people could bring claims to be granted interests in land under the law on proprietary estoppel, and not just defend themselves from having claims brought against them. But the roots of proprietary estoppel in the law on estoppel by representation meant that for a long time, it was thought that the law on proprietary estoppel would only apply to protect claimants who had relied on a belief that they had an interest in land that belonged to the defendants. In Ramsden v Dyson (1866), Lord Kingsdown suggested that the law on proprietary estoppel might apply to protect a claimant who had relied on a belief that he would be given an interest in land belonging to the defendant:

‘If a man, under a verbal agreement with a landlord for a certain interest in land, or, what amounts to the same thing, under an expectation, created or encouraged by the landlord that he shall have a certain interest, takes possession of such land, with the consent of the landlord, and upon the faith of such promise or expectation, with the knowledge of the landlord, and without objection by him, lays out money upon the land, a Court of equity will compel the landlord to give effect to such promise or expectation’ (emphasis added).

However, this was rejected by the majority. Lord Cranworth LC held that:

‘If a stranger begins to build on my land supposing it to be his own, and I, perceiving his mistake, abstain from setting him right, and leave him to persevere in his error, a Court of equity will not allow me afterwards to assert my title to the land on which he had expended money on the supposition that the land was his own… [But] if a stranger builds on my land knowing it to be mine, there is no principle of equity which would prevent my claiming the land with the benefit of all the expenditure made on it. There would be nothing in my conduct, active or passive, making it inequitable in me to assert my legal rights.’

It was not until the decision of the Court of Appeal in Crabb v Arun DC (1976) that it was held that the law on proprietary estoppel could be used to compel A to keep a promise that A would give B an interest in A’s land when B had relied on that promise. In that case, the claimant owned land next to the defendant council’s land. A fence separated the two pieces of land and a road ran along the council’s side of the fence. When the fence was being constructed, the council gave the claimant a point of access through the fence onto the road, so that he could access and exit his land through that point of access. But the claimant – who was thinking of selling the top half of his land that contained the point of access onto the road – asked the defendant council to grant him another point of access to the bottom half of his land. The council agreed that they would do this, and the fence when it was finally constructed contained two gaps, corresponding to the point of access to the top half of the claimant’s land that the council granted the claimant, and the point of access to the bottom half of the claimant’s land that the council had said they would give him. The claimant then sold the top half of his land. The council subsequently fell out with the claimant and blocked up the gap in the fence through which he was planning to access and exit the remaining bottom half of his land that he retained. The claimant sued, arguing that under the law on proprietary estoppel, the council could not go back on its promise to give him a point of access to the bottom half of his land. The Court of Appeal agreed. It held that the claimant should be given a right of way from his land onto the road running along the defendant council’s land. It also ordered that the claimant should be given the right of way free of charge, as the council had behaved so disgracefully in preventing the claimant gaining access to the bottom half of his land for the six or so years that it had taken to take the case to court.

After Crabb, there was nothing to stop the law on proprietary estoppel being used to grant people interests in land where they had relied on a promise that they would be given such an interest in the land. But Crabb emphasised as well that where such a promise was made, what remedy the promisee would be entitled to under the law on proprietary estoppel was discretionary. As Scarman LJ observed in Crabb, the remedy granted to the claimant in such a case would depend on what was the ‘minimum’ required to ‘satisfy the equity’ raised in favour of the claimant as a result of his relying on the defendant’s promise. Sometimes that would involve giving the claimant the interest he was promised. But sometimes that would involve giving the claimant less, or more, than he might have expected to get from the promises that were made to him.

Crabb was a ‘more’ case, where the unpleasantness of the defendant’s conduct led the court to give the claimant more than he might originally have expected to get. Pascoe v Turner (1979) was a ‘what the claimant expected’ case. The claimant was granted title to the house which she used to live in with her ex-lover, the defendant, and which the defendant had told her he would give her when he walked out on her. However, the Court of Appeal would have given the claimant less than that given her minimal reliance on the defendant’s promises (putting some curtains up) had the defendant not been subsequently so unpleasantly vindictive in trying to throw the claimant out of the house. Gillett v Holt (2001) was a ‘less’ case. The claimant was promised that he would get the defendant’s farming business when the defendant died, and on that basis had worked for the defendant for 40 years. The defendant then went off the claimant, dismissed him as an employee, and changed his will so that the claimant would no longer get anything from him when he died. The claimant sued, and the defendant was ordered to hand over to the claimant the farmhouse and 105 acres of land on which the claimant had been living with his wife before being dismissed, as well as £100,000 – but the claimant did not get the entire farming business.

In allowing: (1) a claimant to sue on a promise that he would receive an interest in land; and (2) leaving it up to the courts’ discretion to determine what remedy to give the claimant, the law on proprietary estoppel has lost touch with the idea that estoppel is about stopping someone doing something. Proprietary estoppel is about green lights, not red lights. It gives the green light to disappointed promisees to sue, so long as the promise in question was a promise to give them an interest in land, and it gives the green light to the courts to decide for themselves what is the ‘minimum equity’ required to ‘do justice’ to those promisees.

Perhaps disturbed by the thought that the law on proprietary estoppel had lost its way, Lord Scott attempted to bring this area of law under some kind of control in Cobbe v Yeoman’s Row (2008). The claimant and defendant had reached an agreement ‘in principle’ that the defendant would sell the land to the claimant if he obtained planning permission to develop the land. But a lot of the details of the deal still needed to be settled, and the defendant – unhappy with those details – pulled out of the deal after the claimant had done a lot of work trying to obtain planning permission. In an attempt to obtain some sort of remedy for his client, counsel for the claimant made five different types of claim against the defendant. One of the claims was that the claimant was entitled to a remedy under the law on proprietary estoppel.

The House of Lords held that the law on proprietary estoppel did not apply in this case. Lord Scott gave the leading judgment. He argued (at [14]) that ‘An “estoppel”  bars the object of it from asserting some fact or facts…that stands in the way of some right claimed by the person entitled to the benefit of the estoppel. The estoppel becomes a “proprietary estoppel”…if the right claimed is a proprietary right…’. As the claimant was not arguing that he had a proprietary right in the defendant’s land, and was not seeking to stop the defendant’s raising an argument that would deny that he had such a right, then the law on proprietary estoppel was irrelevant here. Moreover, Lord Scott held (at [18]) that for the law on proprietary estoppel to apply, a claimant would have to received assurances that he had a ‘certain interest’ in the defendant’s land, and there was no certainty in this case that the claimant would have any kind of interest in the defendant’s land as whether or not he would obtain such an interest would depend on whether his negotiations with the defendant were successful. Furthermore, Lord Scott expressed the view (at [29]) that the law on proprietary estoppel could not be used to sue on an agreement which a statute had rendered invalid and unenforceable.

The limitations that Lord Scott attempted to place on the operation of the law on proprietary estoppel in Cobbe v Yeoman’s Row led two academics, Ben McFarlane and Andrew Robertson, to announce ‘The death of proprietary estoppel’ in [2008] Lloyd’s Maritime and Commercial Law Quarterly 449. However, normal service in the law on proprietary estoppel was resumed just a year later when the House of Lords decided Thorner v Major (2009). In that case, David Thorner worked for 30 years unpaid for his father’s cousin, Peter Thorner, on Peter’s farm. At various times, Peter said things that made David believe that he would inherit the farm on Peter’s death. However, Peter died without having made a valid will; the one will he did make, which did leave the farm to David, he had destroyed when he fell out with one of the legatees. David claimed that he was entitled to the farm under the law on proprietary estoppel given that he had done substantial work for Peter because he had been led to believe that he would inherit the farm on Peter’s death.

Consistently with his judgment in Cobbe v Yeoman’s Row, Lord Scott was unhappy about allowing Peter to take advantage of the law on proprietary estoppel to obtain an interest in the farm: ‘I would prefer…to confine proprietary estoppel to cases where the representation, whether express or implied, on which the claimant has acted is unconditional…’ (at [20]). Cases involving ‘representations…of future benefits, and subject to qualification on account of unforeseen future events’ should be dealt with by other areas of the law, such as the law on when a claimant can claim a constructive trust over property belonging to another.

However, this time Lord Scott was in the minority. The leading judgment was given by Lord Walker, who saw no problem in the fact that Peter’s promises (such as they were) that David would inherit the farm on his death must have been conditional on nothing else happening that would require him to dispose of the farm (for example, to fund his being cared for in his old age). He quoted (at [57]) Hoffmann LJ in Walton v Walton (1994): ‘equitable estoppel…does not look forward into the future and guess what might happen. It looks backwards from the moment when the promise falls due to be performed and asks whether, in the circumstances which have actually happened, it would be unconscionable for the promise not to be kept.’ As Peter had now died, his promises that David would inherit had fallen due to be performed, and in the circumstances it would have been unconscionable if David did not obtain the farm.

The House of Lords’ decision in Thorner seems to make clear that Lord Scott’s attempted retrenchment of the law on proprietary estoppel has failed, and this area of the law has been restored to the position it occupied before Cobbe v Yeoman’s Row was decided (though Lord Walker’s strictures in Cobbe v Yeoman’s Row (at [46]) that the law on proprietary estoppel is not ‘a sort of joker or wild card to be used whenever the court disapproves of the conduct of a litigant who seems to have the law on his side’ will survive).

Knocking down the walls

We have now looked at each of the three different rooms in the house of Estoppel that this essay has been focussing on. I hope that why we have these rooms, and why they have developed in the way they have, has been made relatively clear. But looking at each of these rooms in isolation creates a tendency to think of each room in isolation and to simply concentrate on what aspects of each room could do with improvement or tidying up, given what that room is trying to do. But comparing the rooms with each other raises a number of very difficult issues which are going to have to be addressed at some stage, either by the courts or by Parliament.

First, should the law on estoppel have ever gotten into the business of stopping people going back on their promises? Did cases like Hughes v Metropolitan Railway and the High Trees case in fact impede the rational development of the law of contract by relieving the law of contract from the necessity of determining whether cases like Foakes v Beer (1884) were rightly decided in denying contractual effect to promises not to sue for the balance if a debt was paid in part? And given that Foakes v Beer decided not to recognise that such promises were contractually binding, was it legitimate for a first instance judge like Denning J to rule that such promises were binding anyway under a different area of the law?

Secondly, could the law on promissory estoppel learn something from the law on proprietary estoppel and not be so one-dimensional in terms of its response to the fact that A’s promise to B that he will not enforce his strict legal rights against B has been relied upon by B? Instead of preventing (estopping) A from enforcing his strict legal rights, could the law allow A to enforce his strict legal rights but as a condition of doing so, require A to indemnify B against the consequences of having relied on A’s promise? The decision of the Privy Council in Ajayi v R T Briscoe (1964) provides the courts with the justification they would need for taking such a step: there is nothing wrong with A’s enforcing his strict legal rights, so long as B’s reliance on A’s promise can be unwound. But would adopting such a ‘tailored’ response to the fact of B’s having relied on A’s promise render the law unacceptably uncertain?

Thirdly, can it be justified that the claimants who have relied on a promise to give them an interest in land can sue on those promises, while claimants who have relied on a promise to give them money cannot? It is hard to see why promises in relation to land are privileged – in terms of the sanctions against breach that they attract – over other kinds of promises. Either the law on proprietary estoppel needs to be scaled back, along the lines suggested in Cobbe v Yeoman’s Row, or the law on promissory estoppel needs to be expanded, as has already happened in the United States and Australia. But the current position seems unsustainable.

The doctrine of consideration

What is it?

Unless a promise is made in a deed, it will not be contractually binding (though it may still give rise to legal consequences under the law on promissory estoppel, or the law of tort, or public law) unless it is supported by consideration. This is the doctrine of consideration. Most contract textbooks will trot out the following definition of when a promise will be supported by consideration, taken from the case of Currie v Misa (1875):

‘A valuable consideration, in the sense of the law, may consist either in some right, interest, profit, or benefit accruing to the one party, or some forbearance, detriment, loss, or responsibility, given, suffered, or undertaken by the other’ (per Lush J).

This is so wide as to be useless as a definition. For example, it won’t help you to determine whether or not the following promise is supported by consideration:

A is thinking of buying a car. B tells A ‘If you buy a car, I’ll give it a free servicing every year.’ A buys a car. B’s promise played some part in A’s decision to buy a car.

Whether or not A’s buying a car provided consideration for B’s promise, so as to make it contractually binding, depends crucially on whether B’s promise was made with the object of getting A to buy a car. If B was a car dealer, and A was on B’s car lot, and B made the promise with the object of persuading A to buy a car from him, then there is no doubt that if A does buy a car from B, then B will be contractually bound to give the car a free servicing every year. But if B was A’s friend, and B was simply offering to give A’s car a free servicing if he happened to buy a car, and B was not trying to get A to buy a car when he made his promise, then A’s buying a car will not provide consideration for B’s promise, and B will not be contractually bound to give A’s car a free servicing every year.

Here is my definition of when a promise will be supported by consideration:

There will normally exist consideration for A’s promise to B if: (1) A’s promise was made as part of an agreement reached between A and B under which they each promised to do things for the other; (2) A invited or requested B to do x and A made his promise to B in order to induce B to do x, and B was so induced.

Two points need to be made about this definition.

First, it only defines when we will normally find that there was consideration for A’s promise. There may be occasions when there B will not have provided consideration for A’s promise even though the above definition applies to their case; for example, where what B promised to do for A or what B was induced to do by A’s promise was something that B was legally bound to do for A anyway.

Secondly, it is not possible to simplify the above definition without eventually getting confused. For example, it is sometimes said that the doctrine of consideration requires that something be given in return for A’s promise, or that A receive some quid pro quo for his promise. And consideration is consequently often referred to in the books as the ‘price of the promise’. I think this is misleading. In situations covered by the second limb of my definition (under which there will be consideration for A’s promise to B if A was inviting or requesting B to act in a particular way, and A made his promise to B in order to induce B to act in that way, and B was so induced), B’s act isn’t really given in return for A’s promise (which it would be if B said to A, ‘If you promise to do y for me, then I will do x for you’). It’s normally the case that A’s promise is given in return for B’s act: in these kind of situations, A is usually saying, ‘If you do x, then I promise I will do y for you.’ However, as a rough and ready working definition of when consideration will be provided for A’s promise, it’s acceptable to think of consideration as something that is given in return for A’s promise.

Why do we have it?

It is, again, often observed in the contract textbooks that jurisdictions on the European continent don’t have a requirement that a promise be supported by consideration before it will be legally binding. So why does the common law have such a requirement? To answer this question, we need to know a bit about legal history.

So – in England, in medieval times, there were local courts and royal courts. Justice in the local courts was random and arbitrary. A plaintiff (nowadays, claimant) who brought a claim against a defendant in a local court would lose his case if the defendant could find enough people to swear (through a process known as ‘compurgation’ or ‘wager of law’) that he was not liable. So powerful defendants tended to be immune from ever being held liable in a local court. The royal courts were better: they made an attempt to find out what had actually happened in a case brought before them, and tried to come up with a reasoned conclusion as to what the legal outcome of the case should be. However, a plaintiff who wanted to sue a defendant in the royal courts on the ground that the defendant had done him some wrong (or ‘trespass’) would face the problem that the royal courts were only interested in dealing with cases where the defendant’s trespass was ‘against the King’s peace’. In other words, they were only interested in remedying violent wrongs. Normally, this wasn’t much of a problem. The plaintiff who wanted to complain that the defendant had done him wrong would simply bring a claim of trespass in the royal courts and tack onto his complaint that the defendant had acted violently, in breach of the King’s peace, in doing him wrong. The case would then be heard and no one would be that bothered about whether the ‘contra pacem’ part of the claim was made out.

However, there were some claims where it was simply impossible for the plaintiff to allege with a straight face that the defendant had acted violently in doing him wrong. This would be particularly the case where the plaintiff wanted to complain that the defendant had performed some service (such as shoeing the defendant’s horse) badly. It was simply not possible for the plaintiff to say ‘The defendant did me wrong by undertaking to shoe my horse, and then in breach of the King’s peace he shoed the horse so badly that it was made lame.’ The peaceful nature of the defendant’s undertaking and failure to do a good job made it impossible to allege that the defendant’s wrong was committed violently, in breach of the King’s peace. So plaintiffs who wanted to sue for this kind of wrong were shut out of the royal courts and had to sue in the local courts instead, where – as I have said – it was a matter of chance whether or not justice was done.

This changed from about 1350 onwards. From then on, the royal courts showed themselves willing to hear ‘actions for trespass on the case’ (or ‘actions on the case’, for short) under which the plaintiff would not need to allege a breach of the King’s peace to get the royal court to hear his claim. The actions on the case that the royal courts then started hearing including a number of claims where an essential element of the plaintiff’s claim was that the defendant had undertaken, or promised, to do something for him. These claims were known as assumpsit claims (‘assumpsit’ being Latin for ‘he promised’ or ‘he undertook’). The sort of assumpsit claims that the royal courts were willing to hear included:

(1) From 1369: claims that the defendant had undertaken to perform some service for the plaintiff and had injured the plaintiff’s person or property in performing that service.

(2) From 1449: claims that the defendant had undertaken to look after the plaintiff’s property, and that he had failed properly to look after it.

(3) From 1450: claims that the defendant had deceitfully deprived the plaintiff of money by promising to sell the plaintiff land, accepting the purchase price of the land, and then selling the land to someone else or (from 1504) refusing to hand over the land.

(4) From 1520: claims that the defendant had deceitfully induced the plaintiff to hand over goods to a third party by promising to pay for those goods, and then failing to pay for them.

(5) From 1550: claims to enforce the defendant’s promise to pay a debt that he owed the plaintiff in consideration of the plaintiff’s supplying the defendant with certain goods or services.

(6) From 1577: claims to enforce a bet under which the defendant promised to pay the plaintiff money, and the plaintiff promised in return to pay the defendant money, depending on what the outcome of a game was; the plaintiff having won the bet, he sought to enforce the defendant’s promise.

In all of these cases, it was never enough for the plaintiff simply to say ‘the defendant promised’ and sue the defendant for failing to keep his promise. Assumpsit claims were never simply based on an assumpsit: they were always assumpsit plus something else. An attempt was made in 1400 to bring an assumpsit claim for a carpenter’s mere failure to keep his promise to build a house, but that claim – and others like it brought in the following few years – was dismissed on the basis that mere failures to keep a promise could only be sued for in covenant, for which the promise had to be made in a deed.

We owe the doctrine of consideration to the fact that plaintiffs were not allowed to bring assumpsit claims simply on the basis the defendant promised to do something for the plaintiff: something else had to be established for the plaintiff to be allowed to bring his claim. The doctrine of consideration both expressed that fact, and attempted to define what that ‘something else’ had to be before an assumpsit claim could be brought. The vagueness of Currie v Misa’s definition of what would amount to consideration may be attributable to the fact that assumpsit claims could be brought in so many different situations. The fact that an assumpsit claim could be brought in situations (3) and (5) accounts for why definitions of what amounts to consideration say that there will be consideration for the defendant’s promise if the defendant has obtained some benefit from the plaintiff. The fact that an assumpsit claim could be brought in situation (4) accounts for why such definitions also say that there will be consideration for the defendant’s promise if the plaintiff has incurred some detriment as a result of the defendant’s promise. And the fact that an assumpsit claim could be brought in situation (6) – basically, a bilateral executory contract where two people have made reciprocal promises to each other – accounts for why definitions of what amounts to consideration say that the plaintiff’s making a promise to the defendant can amount to consideration for the defendant’s promise to the plaintiff.

This brief history lesson should teach us three things.

(1) Rationale of consideration

The variety of situations where an assumpsit claim could be brought makes it very unlikely that the royal courts allowed those claims to be brought for the same reason in every case. Given this, it is unlikely that we can come up with a unified explanation of why promises that are supported by consideration are legally binding. It is unlikely that the reason why a promise which forms part of an agreement under which the parties have undertaken to do things for each other will be legally binding is the same as the reason why a promise that was made with the object of persuading the promise to act in a particular way and was successful in persuading the promise to act in that way will be legally binding. The first kind of promise is likely to be legally binding because there is a considerable social interest in making reciprocal agreements binding: modern life would be impossible if such agreements were not legally binding. The second kind of promise is likely to be held to be legally binding in order to avoid the material harm that would be suffered by the promisee if he were induced to rely on a promise which was then broken. The first kind of promise is held to be legally binding in the public interest; the second kind of promise is held to be legally binding in order to prevent the promisee being harmed.

(2) Non-enforcement of gratuitous promises

The reasons why assumpsit claims could not be brought for mere breach of a promise were purely formal in nature. There already existed a cause of action that allowed plaintiffs to sue defendants for merely breaching a promise, and that was covenant. The royal courts refused to allow assumpsit claims to be brought for the mere breach of a promise because they did not want: (1) to undermine the rules limiting when someone could bring a claim for breach of covenant by allowing claims for assumpsit to be brought by the plaintiffs who could not bring themselves within the rules for bringing a claim for breach of covenant; and (2) to provide plaintiffs who actually did have a good claim for breach of covenant with an alternative claim in assumpsit. Neither of these reasons, being purely formal in nature, were actually good reasons for refusing to allow a claim for the mere breach of a promise to be brought in assumpsit. So it still has to be established that the modern law of contract does have a good reason for refusing to enforce gratuitous promises that have not been made in a deed.

(3) Limits on when the courts will find that there is consideration for a promise

I have argued that the doctrine of consideration originated in: (1) the fact that a claim in assumpsit could not be brought simply on the basis that the defendant had failed to keep a promise, and that something more needed to be established before the plaintiff could bring such a claim against the defendant, and (2) the need to explain to litigants what that ‘something more’ involved. Given this, it’s hard to understand why at some point before the end of the 18th century, the doctrine of consideration became immutable in that its list of situations where there would be consideration for a defendant’s promise became closed. (I say ‘before the end of the 18th century’ because by the time Pillans v Van Mierop (1765) was decided, not even a judge as bold as Lord Mansfield tried to argue that a written promise to pay a third party’s debts was supported by consideration; instead he argued that such a promise did not need to be supported by consideration to be binding – a suggestion that was disapproved in Rann v Hughes (1778).)

There is a saying that ‘the categories of negligence are never closed’ (Donoghue v Stevenson (1932), per Lord Macmillan). What this means is that the law does not take the view that there is a limited list of situations in which one person can sue another for negligence and if your case doesn’t come within that list, then you have no case. It is a pity, for the rational development of the law, that we do not have an equivalent saying that ‘the categories of consideration are never closed’, so that if a claimant who wanted to sue a defendant for breaching a promise would not necessarily be prevented from doing so just because his case does not fall within the list of situations where the courts have recognised in the past that a defendant’s promise will be supported by consideration. But the categories of consideration are now definitively closed, and a claimant who wants to enforce a promise not made in a deed and whose case does not fall within one of the two limbs of my definition of consideration, set out above, is forced to cast about for some other area of law, such as the law on estoppel or public law, as a way of getting an effective remedy; the law of contract will not help him out. I suppose this does make the law more certain – the law of negligence is notoriously unstable precisely because ‘the categories of negligence are never closed’ – but it does not help the law develop in a clear and rational way. Ideally, we would have one area of law dealing with the issue of when a promise will be legally binding, but at the moment we have three or four grappling with this issue because the area of law that is most apt to deal with this issue – the law of contract – has stopped developing.

Of course, if it were true that ‘the categories of consideration are never closed’ then the start of this essay would look very different. We would have to adopt something like Patrick Atiyah’s view of the doctrine of consideration (see Atiyah, ‘Consideration: a restatement’ in his Essays on Contract (1986)), and say that ‘A defendant’s promise will be supported by consideration if the courts think that there is a good reason to enforce it even though it was not made in a deed. As the law stands at the moment, the courts will hold that a defendant’s promise is supported by consideration if the promise is made as part of a reciprocal agreement under which the parties to the agreement have promised to do things for each other, or if the promise was made with the object of persuading the plaintiff to act in a particular way and was successful in that object. However, it has been argued that there are other kinds of promise which should be enforceable, and it has been suggested that the courts should also recognise that a promise is supported by consideration if…’ All this very much represents the road not taken by English law. There are only two situations in which the courts will recognise that a promise is supported by consideration and it is hard to see nowadays – given the current state of the law – how that list of situations could, or will ever, be added to by the courts.

Should we have it?

I think academics and judges who argue that we should abolish the doctrine of consideration are not thinking straight. In a case where A makes a promise to B, we might want to find that that promise is legally binding because of: (1) something that B has done, or (2) something that A has done.

If we are going to find certain promises binding because of something the promisee has done, then we will need the law to specify when a promise will be legally binding because of something the promisee has done – and that is basically the function that the doctrine of consideration performs today. If we abolished the doctrine of consideration, we would just need to reinvent it (though perhaps in a modified form) to deal with the issue of when a promise will be legally binding because of something the promisee has done. Of course, it is always possible that we would never want to enforce a promise because of something that the promisee has done, and that our reasons for enforcing a promise would always be based on what the promisor has done in making that promise. But this is hardly likely.

No – the real question is whether the current law on deeds and consideration needs to supplemented. And that question turns on this one: Do we think that the courts should ever recognise that a promise not made in a deed is legally binding when the promisee has done absolutely nothing that might give us reason to want to enforce that promise? If the answer is ‘yes’ then the law needs to be reformed so that it says that ‘A promise that has not been made in a deed and is not supported by consideration will still be legally binding if…’ where what comes after the ‘if’ sets out the conditions under which we would want to enforce a promise not made in a deed even though the promisee has done absolutely nothing for his part to make us want to enforce the promise. If the answer is ‘no’ then the current law on deeds and consideration does not need supplementation – though the doctrine of consideration may need modification if it is too restrictive at the moment in specifying the circumstances in which we will enforce a promise based on what the promisee has done.

So – should we ever recognise that a promise not made in a deed is legally binding even though the promisee has done absolutely nothing that might give us a reason to want to enforce the promise? In Pillans v Van Mierop (1765) (which, remember, was disapproved by the House of Lords in Rann v Hughes (1778)), Lord Mansfield and Wilmot J thought the answer was ‘yes’. Two reasons were given by Mansfield and Wilmot as to why we might want to enforce such a promise. These remain the two principal reasons that anyone has ever been able to think of as to why we might want to enforce a promise when the promisee has done nothing that might give us a reason to want to enforce it.

(1) Commercial demand

In Pillans v Van Mierop, the plaintiff bank (Pillans & Rose) effectively lent a merchant, White, £800 so that he could buy some goods from someone called Clifford. They did so on White’s assurance that the defendant bank (Van Mierop & Hopkins) would guarantee the debt. The defendants subsequently indicated that they would guarantee White’s debt. However, White then went bankrupt and the defendants declined to cover the debt White owed the plaintiffs. The plaintiffs sued the defendants. Lord Mansfield observed that the case was ‘a matter of great consequence to trade and commerce’ and held that the defendants’ promise should be found to be legally binding as ‘It would be very destructive to trade, and to trust in commercial dealing if they could [breach their promise].’

There are situations where the needs of the marketplace do seem to demand that gratuitous promises be held to be binding even though they are not made in a deed. The most obvious example is the promise that a bank makes when it issues a documentary credit. The way a documentary credit is as follows. Seller is shipping 10,000 widgets to Buyer. Seller wants some assurance that he will paid for those widgets when they are shipped. The fact that Buyer is contractually bound to pay for the goods won’t be of much good to Seller: Buyer is in another country and may be of doubtful credit. So Buyer gets a Bank to issue a documentary credit to Seller, under which Bank promises to pay Seller the price of the widgets when Seller presents Bank with documentary proof (in the form of what’s called a bill of lading) that he has shipped the widgets to Buyer. Now – Seller will not have provided any consideration for Bank’s promise to pay under the documentary credit. However, it is still commercially vital that Bank’s promise to pay be legally binding as the entire system of international trade would fall down if banks started to refuse to honour documentary credits that they had issued. For this reason, documentary credits are legally binding even though they are not issued in the form of a deed, and are unsupported by consideration. So in this area at least, the demands of commerce have caused the courts to supplement the law on deeds and consideration with a special rule making documentary credits legally enforceable.

(2) Intention to be bound

The demands of commerce were not the only reason Lord Mansfield thought that the promise made in Pillans v Van Mierop should be legally binding. He seemed to take the view that we should enforce any promise that is intended to be legally binding. If a promisor makes a promise intending to be legally bound by it, then that gives us sufficient reason to want to enforce the promise. Lord Mansfield argued that ‘the ancient notion about the want of consideration [making a promise non-binding] was for the sake of evidence only: for when it is reduced into writing, as in covenants, specialities, bonds, &c. there was no objection to the want of consideration.’ The same point was developed by Wilmot J who argued that consideration was normally required to make a promise binding ‘in order to put people upon attention and reflection, and to prevent obscurity and uncertainty…it was intended as a guard against rash inconsiderate declarations: but if such an undertaking was entered into upon deliberation and reflection, it had activity; and such promises were binding.’

The view that a promise should be held to be legally binding if the promisor intended to be bound by it is most strongly associated nowadays with Charles Fried’s Contract as Promise (1981). Two arguments in favour of this view can be gleaned from Fried’s book.

(i) Moral theory. Fried argues that breaking a promise is morally wrong because someone who makes a promise ‘has intentionally invoked a convention whose function it is to give grounds – moral grounds – for another to expect the promised performance. To renege is to abuse a confidence he was free to invite or not, and which he intentionally did invite. To abuse that confidence now is like (but only like) lying: the abuse of a shared social institution that is intended to invoke bonds of trust. A liar and a promise-breaker each use another person.’ Fried sees contract law as existing to give effect to the moral obligation that a promise-maker comes under to keep his promise.

(ii) Transfer theory. The first, introductory, chapter of Contract as Promise suggests a different argument in favour of the position that promises that are intended to be legally binding should be legally binding. Fried argues that contract law is a natural extension of the ‘liberal premise that individuals have rights’ in that it allows us to dispose ‘of these rights on terms that seem best to us.’ So – some academics argue – if I intend to be bound by my promise to paint your house, then I am trying to transfer to you my right to decide whether or not I will paint your house. This was a right which was initially given to me under the law and which – so long as I had it – made it unlawful for you to force me to paint your house. But now I have decided to transfer that right to you by making a legally binding promise to you to paint your house. Given that I have decided to transfer to you my right to choose whether or not I paint your house, why should the law get in the way of my doing that? But that is what the law does when it says that my promise is not – despite my intentions – binding on me because you have provided no consideration for it. Holding that my promise is not binding on me is the equivalent to the courts’ declaring that a gift that I made to you was invalid because you never gave me anything in return.

Both arguments suffer from problems. The moral argument may go too far in that it might be taken as indicating that any seriously made promise should be legally binding, rather than one that was intended to be legally binding. It also unclear why in the area of promises, the law should get involved with requiring people to do the right thing when it does not, for example, sanction much more serious forms of moral wrongdoing such as adultery and failing to raise one’s children properly.

The problem with transfer theories of contract law is that it is not actually clear that I have a right to decide whether or not to paint your house. What I do have is a right that you not punch me, or imprison me, or threaten me with being punched or imprisoned – and all of those rights make it virtually impossible for you to force me to paint your house. But that does not mean that I have a right – equivalent to my rights that you not punch me, or imprison me, or threaten me being with being punched or imprisoned – to decide whether or not to paint your house. So if I promise to paint your house, intending to be legally bound by that promise, I am not attempting to transfer to you something that was originally mine. I am trying instead to create something entirely new, and it is not clear why the law should assist me in doing that just because that is what I want to do.

Summary

Students might find the following summary of this lengthy essay helpful:

(1) The doctrine of consideration tells us that promises not made in a deed will only be contractually binding if they are supported by consideration.

(2) A promise will only be supported by consideration, if (i) the promise was made as part of an agreement under which both parties to the agreement promised to do things for each other; or (ii) the promise was made with the object of persuading the promisee to act in a particular way and was successful in achieving that aim.

(3) It is unlikely that the law enforces promises (i) and (ii) for the same reason. It is likely that a promise that falls into category (i) is enforced for reasons of commercial convenience. It is likely that a promise that falls into category (ii) is enforced in order to protect the promisee from suffering harm as a result of relying on that promise.

(4) Proposals to abolish the doctrine of consideration should be dismissed as nonsensical so long as it remains the case that we will want, in certain situations, to enforce a promise because of what the promisee has done; the function of the doctrine of consideration is to identify what those situations are.

(5) The doctrine of consideration may require modification in so far as it fails currently to identify all of the situations where we would want to enforce a promise because of what the promisee has done. However, such modification is impossible due to the fact that before the end of the 18th century, the categories of situation where the courts would find that there was consideration for a promise became closed, and limited to the two situations identified in (2), above.

(6) The doctrine of consideration and the law on deeds may require supplementation in so far as it is the case that we should enforce a promise not made in a deed even though the promisee has done absolutely nothing that would give us a reason to want to enforce that promise. It may be that certain such promises should be enforced when the needs of commerce demand; it is more doubtful that such a promise should be enforced simply because the person making the promise intended to be legally bound by it.