6—Public Body Unjust Enrichment Claims in France; Lessons for England and Wales



Public Body Unjust Enrichment Claims in
France; Lessons for England and Wales



IT MIGHT BE thought that the solution proposed in the previous chapters is too elaborate. After all, it will be remembered that in his final book Birks argued that the swaps cases necessitated a change to the civilian ‘no basis’ approach,1 and this might therefore be thought to be a simpler and hence preferable alternative. Certainly it is true that it is in public body cases that the English courts have come closest to adopting such an approach,2 and for this reason alone it would be worth examining a system which does follow this ‘absence of basis’ model, albeit that this has not been the result of conscious reflection of the Birksian kind. However, there is a second variable which also makes the French system a useful comparator to add to the investigation of public body unjust enrichment cases in England. In France, the division between public and private law could be found even in the Ancien Régime, and the Conseil d’Etat was founded as a check upon the administration by Napoleon in 1799. Thus, unlike the English system which came to the idea of such a distinction only recently, in France the distinction between public and private law has long been well established.

The Public/Private Divide in France

Under the Ancien Régime in France, there early on appeared jurisdictions specialising in certain categories of administrative litigation, such as the bureaux de finances for fiscal matters. Although the monarchy attempted to forbid the ordinary courts (Parlements) from dealing with administrative affairs, this instruction ‘remained something of a dead letter’.3 Following the Revolution, the Assemblée constituante responsible for fixing the rules of administrative litigation was thus faced with the decision whether, in the light of the poor reputation enjoyed by the administrative courts under the Ancien Régime, it should create new courts of this genre or not. It decided that the active administration itself would decide administrative cases and in order to preserve this jurisdictional monopoly it passed the Loi of the 16–24 August 1790, of which Title II, Article 13 declares that the functions of the ordinary courts are distinct from the administrative functions. The principle of the separation of the administrative and judicial authorities was completed by a decree of the 16 fructidor4 an III.5 However, it was soon realised that some judicial check was needed upon the administration and this was supplied by the creation of the Conseil d’Etat by Napoleon in 1799. Originally the Conseil d’Etat, which took over from the pre-revolutionary Conseil du Roi, was a more bureaucratic body, providing expert advice on the drafting of laws and regulations. However, it also had the duty of ‘resolv[ing] difficulties which might occur in the course of the administration’6 and in time this enabled it to build its independence as a true court, though its jurisdiction as such was not recognised until the beginning of the Third Republic (1870–1940).7 Thus, ultimately, there were two separate branches of courts; the ordre administratif (administrative courts) and the ordre judiciaire (the ordinary courts).8 During the nineteenth century the lower administrative courts developed and the juges judiciaires, or the ordinary judges, began to judge certain types of administrative litigation. In 1872 the Tribunal des Conflits was established to resolve questions of jurisdiction arising between the Conseil d’Etat and the Cour de cassation9 (the highest level ordinary court). The relationship between the jurisdictional divide and the consequently different rules applicable in each system has something of a chicken and egg quality to it, but nevertheless Auby et al argue that the duality of jurisdiction is maintained because of the difference between the rules of public and private law and the need for judges to be expert in one or the other.10 Indeed the training is different for public law judges from that for civil or criminal law judges; they have a placement in the administration and practical experience of active administration.11

The result of this jurisdictional division is that, whereas in England the starting point was to follow the Diceyan orthodoxy (to examine the rules concerning restitution and unjust enrichment and then to apply them to public bodies in the same way as to private parties), in France the starting point is the reverse. The first relevant division, taking the French legal system as a whole, is between public and private law just as was suggested in relation to the English system.12

Today the central characteristic of French public law is the presence of the administration as one of the parties to the dispute,13 but the converse does not hold; the administration can be party to a case which falls within the jurisdiction of the ordinary courts, as will be seen below.14 The term ‘administration’ includes not only the national government, but also local authorities, public agencies, public enterprises, public services (ranging from port authorities to universities, railways and other commercial activities) and quangos (autorités administratives indépendantes).15 For instance, as a close parallel of the English discussion, the regulatory functions of the French Conseil de la Concurrence (Competition Council) are public services controlled by public law.16

However, as well as this organic separation there is also a functional separation between public and private law; indeed Debbasch and Ricci describe the public private divide in France as ‘oscillating’ between these ‘two poles’.17 The distinctive feature of public law was held in Blanco to be the presence of ‘public service’, though this has since undergone radical transformation to the point where certain authors have referred to it as a ‘pseudo-criterion’ devoid of real value.18 In principle it contains various sub-criteria. The public service must satisfy a public need; the activity in question must be carried on by a public authority (though as in England the line between public and private bodies has become very blurred in recent years) and ‘the authority must have recourse to methods and prerogatives which would be excluded in relations between private parties, that is des prérogatives exorbitantes du droit commun’.19 This leads Vedel and Delvolvé to answer complaints that the public service criterion is nebulous by suggesting that, rather than looking for one single distinguishing feature, one should look instead for the presence of a more general concept of puissance publique (public authority or power).20 Their statement that French public law is characterised by the existence of the administration’s simultaneously greater and lesser powers brings us back to the same definition of the public/private divide as was chosen in chapter one.21

Thus, as with our ‘paper clips’ example in English law22 it is not the mere fact that a body is public, but rather the public interest in the activity undertaken by it, which attracts the operation of the public law rules in France. This means that in France, even more so than in England23 it is the activity itself and the question whether that activity is performed in the public interest under a separate legal regime that can be relevant.24 For example, in another two parallels with England, France has had problems with defining the role of sports’ regulatory bodies25 and citizens may raise the illegality of an administrative decision as a defence to a criminal action (collateral challenge).26

Once the relevant competence has been established, it is nevertheless possible that an accessory question may arise which, if it were the principal question, would raise the competence of the other branch of jurisdiction. There are two categories of such accessory question, questions préalables, which may be resolved by the judge hearing the principal question and questions préjudicielles, which may not and which must then be sent to the competent judge. Accessory questions of private law raised before the administrative judge always constitute questions préjudicielles and must therefore be sent to the ordinary judges. As for the competence of the ordinary courts in administrative matters, issues of interpretation of administrative regulations are questions préalables, whereas interpretation of individual acts are questions préjudicielles, except where competence has been expressly granted by legislation. Matters of assessment of legality of administrative acts, whether regulatory or individual, are all questions préjudicielles, except in relation to regulations that inflict grave harm on individual liberty or property rights. What would be known in England as preliminary references to the ECJ under Article 267 TFEU (ex 234 EC) are deemed to be questions préjudicielles from either branch of court.27 The existence of this system exemplifies clearly the point made in relation to the English system that it is not always possible to fit the concerns of a case exclusively into either a public or a private law event, and that expertise from two different fields may sometimes be necessary in order to decide a particular case.28

The French Law of Unjust Enrichment

As for its rules on unjust enrichment, in France the most relevant area is that of the quasi-contracts, three of which are covered by Title IV of the Civil Code. There are some who would argue that these quasi-contracts are unified by the principle of ‘ enrichissement sans cause’,29 but as compared to the French rules on contract and delict they do not have much of a common ‘régime’; only their rules on capacity, prescription periods and proof are shared.30 Indeed, the category of quasi-contracts may be open to additions which have little to do with the concept of unjust enrichment or enrichment without cause as it is understood in the English system. For example, in 2002 it may have acquired a new member31 in the form of pseudo-gain or ‘pseudo-wins’. This arose from a case involving the well-known practice of mail-order companies who include in their mailings notices indicating that the recipient, addressed by name, has won a sum of money. In this particular case the recipient sent off the requisite slips only to discover that this merely granted him entry to a pre-draw from which the winners would subsequently be drawn. The Cour d’Appel had rejected his claim in delict for the winnings on the basis that they did not represent his loss, but a mixed chamber of the Cour de cassation held that in rejecting his claim the Cour d’Appel had violated Article 1371 of the Code, the article concerning quasi-contracts, and the lottery organisers were thus held to be liable for the winnings they had originally indicated.32

Although the whole category of quasi-contracts cannot therefore be regarded as matching the English conception of unjust enrichment or even as forming a unified category of ‘enrichment without cause’, there are nevertheless three quasi-contracts which do seem more closely related to it, two of which are particularly relevant in examining claims similar to those considered in the previous chapters.33

Enrichissement Sans Cause

Of the quasi-contracts it is certainly enrichissement sans cause which bears the most resemblance to the English law of unjust enrichment. This quasi-contract has Roman roots, and has become known as the action de in rem verso. However, Flour and Aubert warn against assuming that this term has exactly the same import as in Roman law.34 The most striking aspect of this quasi-contract is that it is not laid out in any one article of the Civil Code35 (though there are several disparate articles that can be explained on this basis)36 but instead, and very unusually for French law, it became an autonomous source of obligations through jurisprudence. It was in the arrêt Boudier (Boudier judgment) of 1892 that this development took place and the action de in rem verso became a separate cause of action. Although the ambit of the claim as first conceived was extremely wide, subsequent case law has narrowed it, at first severely and then in a more balanced and moderate way, so that now there are three ‘material elements’ and four ‘jurisprudential elements’ to the claim. The material elements are that the claimant should have been impoverished37 (indeed he or she is generally known as the appauvri); that the defendant should have been enriched (hence enrichi)38 and that there should have been a causal link between these two elements,39 though if the enrichment has passed through the patrimoine of a third party, as in the Boudier case itself, this will not block the claim. Of the four negative ‘jurisprudential elements’ the first, as the name of the action implies, is that there must have been no legal cause of the enrichment.40 Most frequently a cause is constituted by a contract between the enrichi and the appauvri, but contracts between the enrichi and third parties can also count as causes of the enrichment, provided that they give the enrichi the right to hold on to the benefit.41 The cause can also be a legal rule,42 a legal decision or the intention to give (l’intention libérale).43 The second negative requirement is that of ‘subsidiarity’, which originally meant that the action in enrichissement sans cause would not be available to the claimant if he or she had any other cause of action, even if that cause of action was blocked for some reason.44 This requirement derived directly from the concern that the new quasi-contract would allow the existing rules of law to be overturned or bypassed, and in this sense could be regarded as another instance of the French rule of non-cumul des responsabilités45, but it is perhaps here that the law has relaxed most in recent cases. For example, where the principal claim would have been brought against a third party who is insolvent, it is now possible for the claimant to bring a claim in enrichissement sans cause against the enrichi instead, though the appauvri must prove the insolvency.46 The third technique for containing the action in enrichissement sans cause was to block it when the enrichment and impoverishment were due to the fault of the appauvri.47 Traditionally the fault of the appauvri deprived him of any claim,48 but some more recent judgments of the Civil Chambers of the Cour de cassation have shown the increasing flexibility common to all the jurisprudential elements, holding that the claim of the appauvri will succeed but he or she will also have to indemnify the enrichi for any harm caused to him or her.49 Finally, if the appauvri has acted even slightly in his or her own interest in addition to the interest of another, he or she will not be able to bring an action in enrichissement sans cause.50 When the action succeeds it is subject to a ‘double ceiling’ on recovery and the claimant can only recover the lesser of his impoverishment and the defendant’s enrichment.51

Répétition de l’indu

This quasi-contract, also based on Roman Law (this time the condictio indebiti)52 deals with situations where a legal subject has received a sum or benefit which is not due to him,53 most often as a result of an error.54 The recipient-defendant is known as the accipiens and the payer-claimant as the solvens. According to Article 1235 of the Civil Code ‘every paiement supposes a debt: that which has been paid without being due is subject to recovery’.55 However, in this context, ‘paiement’ has a very narrow meaning,56 so that receipt of services will not give rise to a claim for répétition de l’indu. Obviously, restitution could be made of their value,57 but the obstacle to doing so may be that it is ‘repetition’, rather than ‘restitution’, that is required. Arts 1376 and 1377 of the Code also distinguish between objectively and subjectively undue sums,58 in the sense that if a sum is objectively undue there was no obligation to pay at all, whereas if it is only subjectively undue, a sum of that nature was owed, but it was either owed by someone other than the solvens, or it was owed to someone other than the accipiens. Such instances of répétition de l’indu subjectif therefore tend to arise in situations involving three parties, the solvens, the accipiens and the true debtor/creditor, and recovery in such situations is not always as straightforward as in situations of répétition de l’indu objectif.59 Although there are some aspects of répétition de l’indu which do not reflect unjust enrichment reasoning (such as the absence of recovery for services noted above, or the fact that it is not necessary to prove overall enrichment or impoverishment in order to bring a claim),60 Mazeaud and Chabas refer to répétition de l’indu as being ‘tinged with’61 enrichissement sans cause, and the two quasi-contracts intertwine in three-party cases.62 Certainly there seems to be enough of a relationship between the two for cases of répétition de l’indu involving public bodies to provide a further parallel with the cases examined in the previous chapters. Importantly, however, the absence of a need to show enrichment or impoverishment means that the French defence of passing on (répercussion) cannot apply and thus when this cause of action is used in order to reclaim invalid taxes or customs duties it is not necessary for the claimant to show that it bore those charges personally.63

Gestion d’affaires

Gestion d’affaires (based on the Roman negotiorum gestio) on the other hand, does not provide such an apt parallel, either for unjust enrichment in general or for the cases under investigation here. Article 1372 of the Code Civil states that:


When one voluntarily manages the affairs of another, whether or not the owner knows of the management, the manager contracts the tacit agreement to continue the management that he has begun, and to complete it until the owner is in a position to make provision for it himself; he [the manager] must equally concern himself with all the appurtenances of this same matter. He subjects himself to all the obligations which result from an express mandate given to him by the owner.64

Gestion d’affaires thus differs from the English conception of unjust enrichment in that it leads to reciprocal obligations; the manager (gérant) must for his or her part complete the undertaking and everything dependent on it,65 taking the care of a bon père de famille. In addition, even though the action brought by the manager appears to be closer to unjust enrichment reasoning than those brought by the master (‘maître’ or, inelegantly, ‘géré’), there are also occasions where it applies when the master has not really been enriched at all (for example repairs are carried out by the manager on a house which then burns down)66, and in any case the amount of recovery is governed by the expenses incurred by the manager.67 Conversely, unjust enrichment and enrichissement sans cause are largely concerned only with the existence of the enrichment, whereas the rules on gestion d’affaires are also concerned with the circumstances in which any enrichment came about, in the sense that it must have been intended for the benefit of the master.68 Indeed, rather than being based on unjust enrichment or enrichment without cause, it appears that the best justification for the quasi-contract of gestion d’affaires is just that someone who has proved his altruism should have his expenses ‘reimbursed . . . it is . . . a matter of encouraging useful initiatives and of discouraging those which are impetuous’.69

As for cases involving public bodies, while of course it is possible that a case of gestion d’affaires could involve a public body,70 French lawyers regard it as a particularly inapt quasi-contract for the public law context.71

Enrichment without Cause and Public Bodies

How, then, have the two key variables of the long-standing public/private divide and the ‘without cause’ approach of French law affected its treatment of cases parallel to those considered in the first five chapters? And what lessons are there for the English system which does not share these variables?

Quasi-Contracts are Automatically Adjusted in their Application to
Public Bodies

The first, and most obvious lesson is that whereas in England the whole aim of part one was to argue for the different treatment of unjust enrichment claims involving public law events, in France the sphere of public law already contains separate administrative versions of each of the private law quasi-contracts.72 These separate versions of quasi-contract are of course inspired by their counterparts in the civil code, to the extent that Chapus refers to them as ‘enfants adoptifs’ in administrative law,73 but they are nevertheless distinct enough for Moderne to state that he is able to study them independently, without referring to private law except for the purposes of comparison.74

Of the three basic quasi-contracts, gestion d’affaires is, as noted above, applied least often in administrative law. Once it is recalled that this cause of action most often applies in situations where the manager has to act because the master is absent, the reason becomes obvious; ‘the administration is never absent’.75 In addition to this, while French law may generally be less concerned with individuality and subjective devaluation than English law, there is every reason not to tolerate the interference of unqualified third parties in public affairs for which the administration is responsible.76 Given the very specialist nature of public services and division of competence between public bodies there are, furthermore, scarcely any examples of gestion d’affaires where the claimant is a public body that has managed the affairs of another public body.77 In many cases the Conseil d’Etat therefore prefers to appeal to the more general principle of enrichissement sans cause, and only in rare cases where this is not possible is the theory of gestion d’affaires invoked. It is particularly interesting that in criticising this tendency, Delacour accuses the administrative judges of ‘confusing the particular case with the genre’.78 This is correct in French terms, since, as noted above, in private law enrichissement sans cause is still separate from the other quasi-contracts as a result of its subsidiarity, but it is interesting that administrative law adopted the quasi-contracts after their initial development in private law and in doing so it has moved closer to the English position, ‘amalgamating’, in Delacour’s opinion, the three quasi-contractual sources ‘to the profit of enrichissement sans cause’.79

Répétition de l’indu, on the other hand, has been used far more often since the decision of the Assemblée Plénière80 on 1 December 1961, in Société Jean Roques81 and in the other branch of courts the Cour de cassation has declared that ‘répétition de l’indu’ is an institution common to private law and to internal public law’.82 Amselek notes that: ‘[I]n this domain one observes, over the long term, a remarkable, as if inexorable tendency of public law to align itself progressively with private law’.83

As in English law,84 the areas of application of répétition de l’indu have been expanded as a result of EC law and the case of Café Jacques Vabre85 (concerning restitution of wrongly-paid import duties on soluble coffee extracts) is well-known in France as being the first case in which first the Cour d’Appel of Paris and then the Cour de cassation checked a national law for compatibility with Article 90 (ex Art 95) of the Treaty of Rome, holding the law inapplicable in the case as a result of the incompatibility.

Unless specific procedural rules apply, whenever the case in répétition de l’indu is brought by a private party against a public body the administrative procedure is used.86 The general exceptions to this rule are of course questions préjudicielles,87 and there are other specific exceptions.88 Within the administrative procedure, répétition de l’indu is a recours de pleine juridiction,89 meaning that the annulment of a decision is sought and the court has full power to replace the challenged decisions with a decision of its own and to require the administration to pay money to the claimant.

In addition to this use of the administrative procedure for répétition de l’indu in the administrative context, the burden of proof is reversed. This contrast with private law is less marked in practice since the claimant (solvens) no longer has to prove that he, she or it paid in error in cases of répétition de l’indu objectif,90 but nevertheless its existence in principle demonstrates the general system of modifying the public law conceptions of these private law events. It should finally be noted that the dispute over whether répétition de l’indu is simply a specific instance of enrichissement sans cause exists in administrative law as well as in private law.91 Despite its later pre-eminence among the administrative quasi-contracts, enrichissement sans cause was, in Moderne’s view, originally not popular because the concern that it would overturn existing rules was even greater in administrative law than it had been in private law.92 However, it was eventually introduced in Ministre de la Reconstruction et du Logement c Société Sud-Aviation.93 Just as in private law, saved expenses can constitute enrichment,94 but the conditions here are slightly different; there must be a real utility of the enrichment (which appears to be largely left to the discretion of the administrative judge)95 and the public body must have ‘assented’ to it. This is relatively easily satisfied since it can be found in the mere ‘tolerance’ of the work by a public body,96 and because, according to Furet, public interest and general utility are more important in public law than the existence of a profit of the kind that would be required in the private sphere.97 Instead, the idea behind these requirements is that the administration cannot be enriched without knowing it; it must be able to know and to control all forms of intervention in matters within its competence and guard against the interference of individuals where necessary.98 As such it is only the equivalent of free acceptance as a rejoinder to subjective devaluation of non-monetary benefits,99 and merely appears to be unusual in the French system because as a whole that system is less receptive to this form of individualism.100 It is true that these requirements only apply in situations where the public body is the defendant, whereas it has been argued here that claims involving public law events should be treated consistently regardless of whether the public body is the claimant or defendant. However, the rule could perhaps be explained simply on the basis that public entities and private parties are, as Furet notes,101 likely to value things differently. Enrichissement sans cause, like répétition de l’indu is a recours de pleine juridiction within the competence of the administrative courts, unless it relates to supply, when it will be presumed to be within the competence of the ordinary courts unless something about the situation ‘endows it with an administrative character’.102 The principle of subsidiarity of the action in enrichissement sans cause is also more nuanced in administrative law than in civil law so that it only applies in certain circumstances, but these include preventing claimants from bringing actions too easily against solvent public body debtors when they may have other options elsewhere.103

If Courts Have to Choose Between Two Events, Both of which are
Actually Relevant to the Basis of the Claim, Some Courts Will
Choose One, and Others the Other

In England, of course, this choice was not initially made. In both Woolwich104 and the swaps cases there were two relevant actions; one for judicial review and one in unjust enrichment. It was only when the two actions were condensed into one hearing105 that the courts chose to view the whole issue from the private law perspective alone. If this initial existence of two actions was not enough to show the relevance of both events, an exercise in comparative law has already shown,106 even without considering the French experience, that if faced with a choice between the two, it is not surprising that some systems should choose the public approach while others choose the private; this is simply the inevitable result of the fact that the choice is itself misleading. From this point of view the fact that much of the French experience is closer to that of Canada, and the opposite of England, is only further evidence in the same direction.

In France, as has recently been the case in England,107 the reason for choosing between the two events has been the time limits applicable to each of them. The first set of cases to consider here concerns taxes on alcohol manufacture108 and a tax known as the ‘supervignette’ imposed on vehicles greater than 16 CV.109 Both sets of taxes were declared by the ECJ to be contrary to Community law. In a series of cases brought before the Chambre Commerciale of the Cour de cassation, the court (and in particular in the supervignette cases Advocate General Michel Raynaud) held that when the ECJ has held a fiscal charge to be contrary to Community law, all actions for reimbursement consequently brought by taxpayers concerned must be considered as a pure civil action for répétition de l’indu, which does not raise any preliminary question of fiscal law. No question of the extent of the taxpayer’s fiscal obligations or the basis of the tax is raised. As a result the special rules of fiscal contentieux (litigation) such as the rules on time limits from the book of fiscal procedures are not applicable.110

So far the cases are exactly parallel to Woolwich where the two events were similarly split. Unlike in England, in the French system it was then decided that when the nullity of the tax had not yet been established, the case was essentially fiscal, even though the taxpayer was claiming restitution of a sum which he claimed to have paid unduly. The restitutionary claim was seen as being accessory to the principal issue of validity, the argument being that the questions to be decided by the judge are exactly the same as if the taxpayer had not yet paid the amount, and so the fiscal rules should apply to the case.111

Thus initially the French experience fell half way between that of Canada and England; cases in which the invalidity has already been established follow the latter, while cases in which the invalidity is at issue follow the former. However, the subsequent intervention of the administration tends even further towards the Canadian model. On 29 December 1989 Article 36 of the Financial Law of that date amended Article L190 of the book of fiscal procedures to provide that:

The rules of this chapter are to apply to the investigation and judging of all actions dealing with the discharge or the reduction of a tax, or to the exercise of rights of reduction founded on the non-conformity of a rule of law which has been applied, with a superior rule of law.

When this non-conformity has been revealed by a court decision, the action for restitution of sums paid or in payment of rights of deduction not exercised or the action for reparation of prejudice suffered can only bear on the period subsequent to 1 January of the fourth year before that in which the decision revealing the non-conformity took place.112

At first sight these two paragraphs might be thought to maintain the distinction, but this was not how the administration chose to interpret them. In its instruction of 10 May 1990,113 it held that Article 36 also subjects the cases mentioned in the second paragraph to the fiscal procedure. Amselek strongly objects to this ‘refiscalisation’ of the cases, and suggests that the Cour de cassation and Conseil d’Etat might not always fully apply them in situations where they ‘restrain abusively, to the profit of the State, the paying parties’ right to restitution’.114 He also argued that these rules were contrary to the jurisprudence of the European Court of Justice on the obligation of Member States to provide equivalent and effective remedies for breaches of EC law,115 though the French courts have subsequently held that they are not.116

It is certainly easy to understand why the four year time limit of Article 36 might be thought abusive when compared to the default private law time limit of 30 years, and in addition, as was argued in relation to Canada this absolute prioritisation of the public law aspects of the case over the private law issues is just as undesirable as the reverse prioritisation chosen by the English courts.117 It was also argued that the law should not adopt one rule for cases where the public body is the defendant and another where it is the claimant.118 However, if, as suggested in chapter five, all public body cases involving unjust enrichment were to be subject to a special time limit of this kind which represented a compromise between the short time limits currently applicable in English public law and the six year time limit applicable in private law, the four years available in the supervignette cases would not actually be so problematic.

In addition to the tax context, a similar choice to prioritise nullity over the rules of répétition de l’indu occurred in the context of social security payments. In France, the office of social security is known by the acronym URSSAF.119 Although this is an entity responsible for a mission de service public, its pay-offices, or caisses are more like our provident societies and have traditionally been regarded as private bodies.120 Claims involving the caisses are therefore brought in the civil branch of the courts, but can nevertheless be subject to a special procedure. Thus Article 42 of the décret No 58-1291 of 22 December 1958 provides for a one month time limit to apply to certain claims brought by employers concerning social security payments. Consequently the social chamber of the Cour de cassation held that where a health insurance pay-office had fixed a tariff of payment and this had not been challenged by an employer within these time limits, the matter could not be reopened by way of a later action in répétition de l’indu to reclaim the social security payments (cotisations) in question.121 Clearly, then, here too the French courts have prioritised the issue of nullity and its applicable time limits over the payment of an undue sum in a manner which is the reverse of that ultimately chosen by the English courts.

Given their attitude to nullity in these instances, it is interesting, therefore, that faced with the issue of contractual nullity, the French administrative courts have aligned themselves with the approach taken in England, distancing themselves from the approach of the private law courts in situations of purely private contractual nullity.

On 24 September 2002 the first civil chamber of the Cour de cassation held that restitution between the parties to a failed contract is not dealt with by the law of quasi-contracts, but by the ‘rules of nullity’ themselves.122 This meant that instead of the 30 year time limit for répétition de l’indu applying to restitution under annulled contracts the court applied the five year time limit applicable to the action for annulment. In administrative law, on the other hand, it is the quasi-contracts which perform this role. Thus in Société Citécable Est123 the Conseil d’Etat famously held that if a contract with the Administration is annulled,124 the co-contractor has the right to reclaim from the Administration the reimbursement of its ‘dépenses utiles’, (the useful enrichment as described above) on the basis of enrichissement sans cause. Not only can a claimant use enrichissement sans cause, it can also sue for fault-based damages following the annulment of a contract, as long as the total reclaimed under these two heads does not exceed what it would have received if the contract had been enforced. Even more strikingly (for French lawyers at least)125 the parties have the right to bring such claims regardless of the stage that the proceedings have reached (‘including for the first time on appeal’)126, and despite the fact that the rules of enrichissement sans cause do not belong to the set of rules known as being d’ordre public, ie of such importance that they can be raised by either of the parties or the judge at any time in this way.127 Only expenses specially incurred for the purpose of completing the contract can be reclaimed; even if a company has been founded solely for the purpose of carrying out a public contract, the company’s ordinary everyday expenses will not qualify. However, the burden of proving the lack of utility of the work or expenses rests on the public body.128 According to Guglielmi, in principle both the amounts claimed on the basis of enrichissement sans cause and through fault-based liability ought to be reduced to take account of the fault of the claimant, and he consequently criticises a decision of the Tribunal Administratif de Bordeaux for having reduced only the amount of fault-based damages in a case where the claimant company had agreed to conclude the contract in the knowledge that it was illegal.129 Moderne also notes that some authors dispute the necessity for the public body’s assent in cases of annulment, pointing out that if the contract has been annulled ab initio it is difficult to say that assent has not also been annulled.130

Enrichissement sans cause, of course, applies not only to annulled contracts,131 but also to situations where contracts have not been validly concluded, and where the contractor has exceeded its initial contractual framework in conditions which cannot give rise to contractual liability.132 It could also be the case that the work required by the contract has turned out to be more onerous than was anticipated,133 or is seen as being supplementary to the contract.134 Such supplementary work is usually the result either of the initiative of the contractor or of an order given by the administration.135 However, there cannot be recovery if the schedule of the contract requires the contract to be unchanged, or if certain formalities for changing it are not followed136 and it is difficult to see these ‘supplementary work’ cases as being wholly based on enrichissement sans cause because the work done must be totally remunerated.137

Indeed it is unclear exactly why the annulment of public contracts should be dealt with using enrichissement sans cause when other contractual issues are not, nor why the rules concerning public contracts should have diverged from both the private law and cases concerning taxes or social security payments. It seems unlikely to be because the quasi-contract rejected in all those other cases was répétition de l’indu,138 whereas here administrative law has used enrichissement sans cause,139 although the latter is certainly a more general quasi-contract able to deal with the conferral of benefits through work and services. There are, however, three potential answers which may go some way to explaining the choice to use enrichissement sans cause in relation to public contracts.

One is that in fact public and private law may not, after all, have diverged to such a great extent. Commenting on the decision to reject the operation of répétition de l’indu, Aubert argues that the Cour de cassation was simply concerned to ensure that restitution cannot be claimed without prior annulment of the contract and thus this annulment must occur first, within its own (shorter) time limits. On his reasoning there would therefore be no bar to the operation of répétition de l’indu in a situation where the contract had previously been annulled,140 and the difference between public and private law would then not be so great, because the cases in the Citécable Est line all refer to co-contractors whose contracts are already null.141

An alternative possibility is that if the law has diverged that is simply a coincidence. Chapus asks whether of the three quasi-contracts enrichissement sans cause has produced the most abundant jurisprudence in administrative law because it is not specifically referred to by the Civil Code,142 and it may just be that the jurisprudential and less rigidly-defined nature of enrichissement sans cause has allowed the administrative courts to develop their own version of it more independently of civil law concerns than they felt able to do in the case of répétition de l’indu, for example. In addition, the administrative courts may have felt that they could avoid the application of Article 1304 of the Civil Code143 when they did not feel that they could equally avoid the application of the more specific Article 42 of the Social Security décret or Article L190 of the Livre des Procédures Fiscales in the other nullity cases discussed above. The explanation for the different treatment of contracts when compared with the tax and social security cases could be that this is just a further illustration of the same unsurprising split as occurred between the Canadian and English courts; both the nullity and the enrichment144 are relevant, so if courts must choose between them, some will choose one and others will choose the other.

A slightly more radical alternative would be to see the different choices made by the public and private courts as reflecting different understandings of the nullity involved. It is evident that both branches of court regard annulled contracts as being void ab initio, and thus as effectively never having existed,145 but it is nevertheless the case that the five year prescription period applicable in private law courts comes from Article 1304 of the Civil Code, and only concerns the nullity of ‘conventions’.146 To some extent, therefore, paradoxically, the non-existent convention is still allowed to determine the time limits applicable to the declaration of its own non-existence. However, for the Conseil d’Etat in Citécable Est, the nullity of the contract meant that a decision to resile from it could not be annulled, because that decision could not have existed in the first place, and further that recovery was governed by the wholly independent concept of enrichissement sans cause. At least in terms of principle, then, the nullity applied by the Conseil d’Etat in Citécable Est appears to be more absolute, and therefore more in keeping with the ab initio voidness discussed in part one147 than the rules of nullity used in the Cour de cassation’s decision of 2002.148

As for the practical effects of these two approaches, however, it seems that the difference between a simple unravelling of the contract and an action in enrichissement sans cause is the quantity of recovery on the part of the claimant.149 The rules of nullity used by the civil law are objective and literal, so that the parties have to return everything they still hold, and where they cannot make specific restitution they have to return the value of the thing at the date it was transferred with, in some cases, an allowance for its use in the interim.150 With enrichissement sans cause, however, recovery is only of the lesser of the appauvri’s loss and the enrichi’s gain, measured in terms of ‘useful expenses’ in the case of public bodies. Oddly, the loss is measured at the time it occurred, whereas the enrichment is measured at the time of the claim (unless the enrichi was in bad faith) so that if the appauvri has done the work but it is destroyed, damaged or lost before the claim is brought, the enrichi will not be regarded as having been enriched at the time of the claim and thus the appauvri will be unable to recover.

In a sense, then, conversely, the rules of nullity used by the civil law actually represent a more literal unravelling and reversal of the annulled contract, and more effort is focused on returning the parties to the position they would have been in had the contract not existed there than under the rules of enrichissement sans cause used by the Conseil d’Etat. On the other hand, once it is acknowledged that this unravelling is largely going to be impossible when the enrichment constitutes work done by the appauvri, it is possible that the rules of enrichissement sans cause are actually more consistent with the purpose of annulling the contract in the first place. For example, taking the facts of Citécable Est, if the contract is simply void for procedural reasons, as in that case itself151 the work will presumably still have counted as being useful to the public authority, whereas if the contract related to something the public authority was not permitted to do in the first place152 then it may not be so easy to argue that the result was ‘useful’ to that public authority and therefore that it constituted an enrichment. Recognition of the relevance of enrichissement sans cause could therefore prevent the public body from paying for something it should not have contracted to receive, in circumstances where the whole point of annulling the contract was to prevent such an exchange from taking place, whereas the rules of nullity would require the public body to pay the private contractor the value of anything that it could not return specifically, despite the fact that paying for such work was precisely what the annulled contract required it to do. Indeed if the value of the work and goods transferred had gone up in price, the public body could actually end up paying more for those goods and services than it had contracted to do under the annulled contract. Obviously, however, such a conclusion would depend on the courts’ precise definition of ‘usefulness’ in enrichissement sans cause claims, and in any case it runs the risk that the public body could be left with the work without having to pay for it, to the detriment of the private contractor, though this in turn could be seen as an incentive for private contractors to ascertain carefully the power of the public body to enter the contract in the first place.