Wisdom from Trajan’s Market

Of Quasi-Contract and Taxonomy

The New Digest is thrilled to feature this guest contribution from Dr Rajiv Shah. Dr Shah has a PhD in law from the University of Cambridge. His doctoral thesis was on the law of Unjust Enrichment.

Suppose you owe me £100 but accidentally pay me twice. Under the law of all western based legal systems – which covers every country in the world other than Saudi Arabia – I then owe you £100.

Historically, in England, the action by which you could recover the £100 from me was known as the action for money had and received. The pleading for this action required the plaintiff to aver that the defendant had promised to pay that sum of money. But in the scenario of a mistaken payment there rarely was such a promise. Instead, English law implied that on receipt of a mistaken payment the recipient had promised to pay it back. It did not matter if there was no actual promise, the law implied the existence of such a promise. It was a legal fiction.

Until about the 1960s, almost everyone in the English speaking world classified such a claim as being ‘Quasi-Contractual’. That is still the classification used in some continental legal systems (see for example Art 1300 of the new French Civil Code). Common law scholars have almost unanimously abandoned the classification of such an obligation as ‘quasi-contractual’. It has been replaced by ‘Restitution’ or ‘unjust enrichment’ (there is a difference between those two but it does not concern us here).

The gist of the argument for rejecting ‘quasi-contract’ is that this cause of action has nothing to do with a contract: there is no agreement between the parties, and whilst the law did historically rely on a fictional promise this was a mere fiction in order to make a remedy fit within an existing form of action.

The point, to be clear, is not that this is not contractual. No one ever claimed that this was a contractual form of liability – the use of the term ‘quasi-contract’ rather than ‘contract’ makes that clear. Rather the objection is that ‘quasi-contract’ gives the erroneous impression that this form of liability is close enough to a contract when actually it has nothing to do with it.

This point seems obvious. So obvious. How could previous generations of lawyers have fallen into such a trap?

When faced with such a situation it is best to approach it with humility. As Lord Reed, the now President of the UK Supreme Court, put it about this area of the law: ‘Although judicial reasoning based on modern theories of unjust enrichment is in some respects relatively novel, there are centuries’ worth of relevant authorities, whose value should not be underestimated. The wisdom of our predecessors is a valuable resource’ (Investment Trust Companies v HMRC [2017] UKSC 29 at [40]).

A study of Roman Law can help us elucidate this point. It is often said that Roman Law had a law of contracts whereas the (contemporary) common law has a law of contract. What this means is that Roman law had different rules applicable to different types of contracts, for example a sale, loan, or a mandate,^[1] whereas, the common law has a general set of rules that is applicable to all types of contractual arrangements. This is most clearly reflected in legal education. A law student in the English speaking world is not typically taught the law that is applicable to various types of common contracts but is instead taught the general law that is applicable to all contracts. A typical contract law course consists of the rules on formation of an agreement (offer, acceptance, certainty), requirements for the agreement to be legally binding (intention to create legal relations, consideration), the identification of the terms of a contract and their interpretation, the various ways in which a contract ends (discharge, breach, frustration), and remedies (expectation v reliance damages, specific performance, injunctions).  The study of particular types of contracts – such as sale or agency – is typically reserved for an advanced course (in Cambridge, where I studied, it was called ‘Commercial Law’).

By contrast a student of the Roman law – and indeed this was the way contracts in Roman law was taught at Cambridge (Cambridge and Oxford are almost unique in the common law world in making the study of Roman law compulsory and in teaching it as a living subject rather than as legal history) – would be taught about particular contracts. So for example the rules relating to sale, loan, and mandate would be taught to students and only then would attempts to generalise the rules be introduced.

Such a difference is not an inherent one between the Roman law and the common law. Indeed, the common law also used to have a law of contracts, the shift to the law of contract was a Victorian innovation.

For our purposes there were two main subcategories of contract in Roman law: contract in re and contracts consensu. Contracts in re arose from the delivery of a thing and contracts consensus arose from the agreement of the parties without the need for any further formalities. This does not mean that contracts in re did not require the consent of the parties – they did – but the transfer of a thing was another salient feature.

The contracts in re were commodatum (loan of a returnable thing), mutuum (loan of a fungible/consumable thing), pignus (pledge) and depositum (deposit).

A loan of money was a mutuum, this obliged the recipient of money to return the equivalent sum though (unlike with a commodatum) he was not expected to return the exact same coins which he had received.

What if, instead of a loan, a payment had been made in error? The Digest answers thus: ‘He, also, who, through the mistake of the person who made the payment, received something to which he was not entitled, is bound as in the case of a loan, and is liable to the same action as that to which a debtor is liable to his creditor. It should not, however, be understood that he who is responsible in a case of this kind is bound by a contract; for anyone who pays money by a mistake does so rather with the intention of discharging an obligation than of contracting one.’ (D.44.7.5.3, my emphasis).

Although the Corpus Iuris Civilis did not use that term, later scholars of Roman law called such a relationship promutuum given its obvious similarity to mutuum: in both cases a thing (here money) is transferred from one person to another, and in both cases the original owner does not intend to make an outright gift. It is true that a mutuum has the additional feature that the recipient knows and intends to return the money when he receives it and this is not a present in a promutuum. Nonetheless, the absence of that fact is not critical and liability can still be established based on the two former factors.

Mutuum and promutuum are incredibly similar and the only difference between them is one which is not critical to establishing liability. Therefore, although it is clear that promutuum is not contractual it makes a lot of sense to call it quasi-contractual.

The common lawyers who employed the label of quasi-contract were well versed in the classical legal tradition and that is why they it made sense to them to refer to it as quasi-contractual. Here, there is no need to speculate that this is the reason for their use of that term as they explicitly told us that was the reason for their use.

In Sinclair v Brougham Lord Dunedin explains it as such: ‘And coming to the case of money, while mutuum was proper loan, pro-mutuum covered the cases where money was had and received without contract, and a special form of action for the common case of the payment of a supposed but non-existing debt was known as condictio indebiti. Now, the English law, having no quasi contracts, got over the difficulty in such cases as the action for money had and received by the fiction of a contract.’ (Sinclair v Brougham [1914] AC 398 at 432). In another case, Fletcher Moulton J put it as ‘the obligation arises re and not consensu’ (Nash v Inman [1908] 2 KB 1 at 8).

When the new taxonomy – under which a contract was defined entirely in terms of a legally binding agreement – took hold, the language of quasi-contract to describe such obligations no longer made sense and so the label quasi-contract had to be abandoned.

Whilst some taxonomies are better than others, I do not think that there is a right or a wrong answer when it comes to taxonomies. The point of this post is certainly not to argue that against the prevalent theory and taxonomy of contract in the common law world. The point is merely that we can benefit from understanding that there are different ways of categorising the legal system and that, when seeking to decipher the wisdom of our predecessors we should not import modern meanings into the terms they are using even when it might seem obvious that we are using the same term.

There is a final point about taxonomy. We have a natural tendency to seek to identify common features about all the instances in a category. However, in doing so we run the risk of forgetting other important features and we also run the risk of thinking that those common features are always going to be salient. It is absolutely true that agreement was a common feature of both contracts in re and contracts consensu. But whilst agreement was critical to liability for contracts consensu, that was not the case for (at least some) contracts in re. A taxonomy focused solely on agreement has the advantage of simplicity but this comes at a cost.

---

^[1] Roman law also had a general form of contract called the stipulatio. However, that general form did not swallow the specific forms of contract.