1 A Worst-Case Scenario

There are three reasons for considering government liability a fertile topic for comparison. Every system of law must face and solve problems concerning the liability of public authorities and public officers.⁵⁰⁹ The solutions can be similar or different, but legal scholarship has, until now, focused on their diversity, making government liability a ‘worst-case scenario’. However, over the last seven decades, the autonomy of the State, of every State when addressing issues concerning liability has been limited by the obligations stemming from membership of both the EU and the Council of Europe. Both these aspects will be addressed in turn. Dicey stretched the contrastive approach when he criticized

‘the opposition specially apparent in the protection given in foreign countries to servants of the State or … of the Crown, who, while acting in pursuance of official orders, or in the bona fide attempt to discharge official duties, are guilty of acts which in themselves are wrongful or unlawful’.

‘the fourth and most despotic characteristic of droit administratif lies in its tendency to protect from the supervision or control of the ordinary law courts any servant of the State who is guilty of an act, however illegal’.⁵¹⁰

The problem was that he focused on less recent French norms, essentially as they were before 1848, whereas they had been changed in 1875. Well before the first edition of his treatise, much had already been done through the jurisprudence of the Conseil d’État concerning liability. Moreover, for all the prestige of French administrative law, in other parts of Continental Europe, lawyers and judges were used to thinking of tort law as preeminently private law. Dicey’s account of English law also failed to acknowledge the fact that the English law of the day accorded extensive immunities from being sued in tort to the Crown, even if government officials did not benefit from them. In a similar vein, as early as the 1950s, Davis cast doubt on ‘whether English courts at any time would have held the Prime Minister liable personally on account of exercise of discretionary powers’.⁵¹¹ However, for some, government liability is still marked by differences.⁵¹² It can, therefore, be viewed as a sort of worst-case scenario in the sense illustrated by Shapiro, ie, as the ‘body of known legal phenomena most likely to falsify’ a conjecture or position, in our case, the hypothesis that a common core exists and does not consist only in vague ideals.⁵¹³

This is all the more interesting in the light of European integration because the choice of solutions to the problems concerning public authorities is no longer entirely left to each individual State. The measures adopted by both the Council of Europe and the EU influence this choice. After issuing various acts concerning the exercise of administrative powers, the CoE Committee of Ministers adopted a recommendation on public liability.⁵¹⁴ Although the recommendation was not binding on the member States, it was interesting for two reasons. The first of these was the intent to increase the liability of public authorities as evident in the first principle laid down by the Recommendation: reparation should be ensured ‘for damage caused by an act due to a failure of a public authority to conduct itself in a way that can be reasonably expected from it in law’, with the presumption that a failure occurred ‘in case of transgression of an established legal rule’. The EU has also adopted some measures that have had an impact on national systems of liability. Some of these measures are sector-specific as they concern a given field, such as public procurement. Since 1989, EU directives require member States to ensure that if those who participate in government procurement suffer unjust damage arising from the conduct of public authorities, they can seek financial compensation. A more general limitation to the autonomy of the member States derives from the Court of Justice’s assertion of the general rule, ‘inherent in the system laid down by the treaties’, whereby a State that fails to respect the rights stemming from ec legislation is financially liable for the damage caused to those rights and cannot claim the sovereign immunity.⁵¹⁵ By virtue of the vertical effect of this rule, EU tort law upholds individuals’ interests against the State by imposing the respect of legality. Secondly, the legal system of the ec/eu is interesting due to its own regime of liability. The Treaty of Rome made the eec liable for all the torts it or its agents committed, and rested liability on the ‘general principles common to the laws of the member States’.⁵¹⁶ This could be interpreted as implying that the six founding States shared a general principle of tort liability. Its existence was implicit, too, in the renvoi to the shared general principles.

2 Further Consequences of Procedural Unfairness in Adjudication

We have already examined two cases concerning claims of procedural impropriety and unfairness; that is, the dismissal of a civil servant and the revocation of a license inaudita altera parte. They deserve further consideration because due to their ‘extreme nature’ in that government action is affected by ‘egregious procedural failings’.⁵¹⁷ The point of interest here is whether the arguments brought by the applicants can be endorsed by the courts also from the viewpoint of damages.

Our first hypothetical case concerned a disciplinary procedure ending with the dismissal of a policeman, Maurice. In this case, it was found that some standards generally associated with public law are relevant in the legal systems included in our comparison. An employee subjected to such a procedure is entitled to a clear statement of the charges against him and to a meaningful opportunity to make his case in a hearing, assisted by either a lawyer or an expert. The authority is also required to take into account all relevant considerations. If the final decision adversely affects the employee, and all the more so in the event of dismissal, the public authority must give a statement of reasons. There are cases where the misbehavior of the civil servant is so serious, and the facts are so manifestly clear, that a judge may take the view that whatever the description of the facts and related events the employee may give, it would make no difference. This was the standard, from a comparative perspective, set by the ecj in Alvis. As for reasons, in various legal systems, these might be furnished in another document, so long as a reference is made to it. Moreover, in various legal systems, including Germany and the UK, interest would be payable, but this would not be the case in France. It is equally uncertain whether Maurice would be able to obtain damages for loss of reputation. This would be so in Spain and Switzerland but not in Austria and Poland, while in the UK it would depend on whether the procedural failings were due to the inadequacy of sector-specific rules rather than an act of discretion.⁵¹⁸

That having been said, an area of agreement between legal systems can be found where none would have been expected on the basis of the ideas and beliefs about public law widespread until some decades ago. For example, though there is no general duty to provide an oral hearing in Italy, it would be required in a case of this kind. Similarly, in the UK, the common law does not impose a general duty to furnish reasons, but it would do so in this case. If the dismissal infringes these standards, which are associated with the public law values of propriety, fairness and transparency, also in the light of Article 6 echr,⁵¹⁹ it will most probably be annulled. The area of disagreement becomes even less significant if one considers that not only does the annulment produce its effects retrospectively, but it also implies that the public authority must pay compensation. However, this cannot be established simply in terms of the net wages Maurice would have earned without unfair dismissal, as would be the case in Germany.⁵²⁰ Especially in France, this is not permitted due to a principle stemming from public finances: public money cannot be given for a service that has not, in reality, be rendered.

The other hypothetical case concerns the revocation of a license or concession inaudita altera parte. The distinguishing elements of this case can be briefly mentioned. There is no concept of the employer’s prerogative, but a discretionary power related with the distribution of government largesse, in this case through the attribution of the reserved use of a part of the public domain. More importantly, the facts of the case confront us with an exercise of power in absence of any notice and hearing, as well as reasons justifying the decision taken by the public authority.

In this respect, at least two elements of diversity must be noticed. First, in some legal systems, including Austria and Germany,⁵²¹ a claim against government liability has a subsidiary nature and is, accordingly, only permissible if other legal remedies have failed to protect the claimant’s interests. Accordingly, only a limited space remains for a claim for damages. Second, in some places – for example, in Poland – both losses and lost profit would be included in the compensation the court may award, whereas in Austria, lost profit would be recognized only where the conduct of the public authority is characterized either by willful intent or gross negligence, and in France it would not be granted on the basis of the principle of the precariousness of any license concerning the public domain.⁵²² There is thus a disagreement among the compared legal systems, and not just regarding questions of detail but also the applicability of the provisions established by civil codes, if they exist.

However, from a comparison of the actual results from the viewpoint of judicial enforcement of the requirements imposed on public authorities, an area of agreement among legal systems can be found. First and foremost, all the legal systems adhere to the principle that administration is not exempt from liability. The point Wade made in the early 1960s – ‘if an authority exceeds it powers […] if its act, being without justification, constitutes a tort’ then it is liable for an action for damages – was thus of general importance for administrative law.⁵²³ Whether their functions, notably legislation, can still be considered immune outside the area governed by EU law is another question. Secondly, it emerges that procedural requirements such as the right to be heard and the duty to give reasons are everywhere deemed necessary in order to protect people from arbitrary interference by those who wield power over them. Infringement of these requirements entails the invalidity of the revocation due to the impossibility to understand how administrative action can be justified. The area of agreement is, therefore, more significant than expected.

3 Contracts: The Unlawful Exclusion of a Tenderer

We turn now to other ways in which administration is discharged, namely through contracts and police powers. We observed in Chapter 3 that the close of the nineteenth century saw a growing number of requests for judicial review of police power by individuals claiming that they had undergone abuse of power in the ways in which limitations were imposed either on personal liberty or freedom to do business, and we suggested that the development of procedural requirements permitted the courts to control the exercise of power through judicial review. The topic thus has a bearing on our argument that procedures can provide a better terrain for testing the existence of a common core.

Things are partially different as regards contracts, which normally enhance the opportunities for individuals and firms to achieve their own goals.⁵²⁴ As contracts are increasingly used by public authorities, which define therein the terms for spending a very large sum of public money, there are issues of fairness and propriety concerning their making.⁵²⁵ At first sight, the law governing administrative action appears to be characterized by both commonality and diversity. Whereas in some legal systems public authorities are subject to the general law of contract, albeit with some adaptations as a response to the specific needs of administrative activities, in others there is a more or less separate body of government contract law, concerning what the French call administrative contracts (‘contracts administratifs’). On the other hand, the solutions are largely influenced by international and supranational norms. There are international treaties, such as the Agreement on Government Procurement adopted in the context of the wto.⁵²⁶ There are EU directives, which are binding for its member States, as well as on the others that have stipulated agreements with the EU, such as Norway and Switzerland. These directives establish the general principles to be respected, including free competition and transparency for acquiring goods and services, and the specific types of procurement procedures, such as open tenders and private auction. They define detailed prescriptions concerning the phases or stages of such procedures.⁵²⁷ As a result of all these requirements, when public authorities make contracts, their freedom of choice is much more limited than that of private parties in order to ensure the respect of the principles of fairness and transparency so as to avoid any bias among parties and the mismanagement of public money.

In our hypothetical case, we seek to shed light on how contracting authorities exercise their powers in the context of public procurement law: the exclusion of participants from a tender procedure if they are considered unreliable and therefore unsuitable to be awarded a contract. Contracting authorities can exclude from a procurement procedure any operator who has been convicted for professional misconduct, including the failure to pay social security contributions. We thus imagine that during the procurement procedure initiated by the municipality of Mandeville, it receives information from the Department of Social Security (dss) showing that a bidder, Alphagroup, has systematically failed to pay social security contributions. Mandevilles’ officials use the information received from the dss, which is not informed about it, and drop its offer. It is only after the conclusion of the tendering procedure, when Alphagroup has access to the documentation held by the municipality, that it discovers that its offer had been dropped because of its alleged systematic failure to pay social security contributions. It then seizes the national court, arguing that: (i) the municipality had no right to use information against it without giving it a real opportunity to challenge it; and (ii) factually, the dss had made a mistake, insofar as the economic operator that failed to respect the obligations stemming from social security legislation was not Alphagroup but another one named ‘Alpha Group Ltd’. Alphagroup then brings an action for damages against the Mandeville municipality before the court. The municipality objects that if anyone is liable, it should be the dss. The questions that thus arise are, first, whether Alphagroup’s action based on procedural fairness would be likely to be endorsed by the court, and, if so, whether the court would conclude that the participant lost the chance to win a contract and, lastly, if the municipality is found liable, whether it can turn to the dss, in order to place the burden on them, or at least share it.

It is necessary to make a preliminary remark concerning the remedies available against the exclusion of a tenderer. All the national reports point out that the legal systems examined provide for interim relief, which is a requisite for EU States that was laid down as early as 1989 by directive 89/665.⁵²⁸ Some reports add that the request for immediate judicial protection constitutes a requisite, in the sense that if the applicant has not sought to obtain it, a judge will normally be unlikely to allow an action on grounds of liability. The underlying assumption is that if the tenderer had sought to obtain interim relief, the contracting authority would have an opportunity to rectify its decision. This is important in itself. It may also give rise to further consideration from the point of view of the question concerning the respective positions of the local and national authorities because the former had no reason to suspect that the information provided by the latter was wholly inaccurate. Alternatively, it might be argued that had the contracting authority correctly followed procedural requirements and thus heard the applicant, it would not have excluded it only on the basis of the information received from the Department. However, in our hypothetical case, we suppose that the participant has not been duly informed in a timely manner of the reasons supporting its exclusion precisely because we seek to understand whether the infringement of procedural requirements can give rise to government liability.

Not surprisingly, the results reached are largely those that experts in government procurement with a comparative background would have anticipated. The distribution of judicial powers is even more differentiated than in other cases. In the UK, for example, the firm’s action would obviously have to be brought before the ordinary courts. Other legal systems, including France and Italy, apply the general rule that action taken by administrative authorities must be challenged before administrative courts, though there are sometimes particular rules governing this type of judicial proceeding, in order to ensure its speed. There are still other legal systems, for example Austria and Poland, where, unlike the action for annulment, the action for damages falls within the jurisdiction of the ordinary courts. Moreover, as regards the standards of administrative conduct, in Germany the duty to accurately consider all the relevant facts is established by legislation, while elsewhere – notably in the UK – it is a judicial construction. Another crucial factor of diversity concerns the conditions for affirming governmental liability. There are differences depending on whether the liability is governed by the general rules, established either by civil codes or the courts (as in Germany and the UK, respectively) or is subject to a special legal regime, notably as in France. In Germany, for example, liability is sometimes imposed on the basis of illegality, and this solution can also be found in EU law, where there is no real discretion (otherwise, the breach would have to be sufficiently serious).⁵²⁹ In other legal systems, including France and Romania, authorities are liable if an administrative action is characterized by either fault or intentional wrongdoing, and the general understanding is that fault denotes illegality.⁵³⁰ Further differences arise in the field of public procurements. Within the EU, this field is the subject of both EU directives and the national legislation implementing them, while in Switzerland some cantons do not have special provisions, and others rule out compensation for lost profit.⁵³¹ In Germany, compensation for loss of profit might be awarded if the participant can show that if the error of assessment had not been made, it would have been ranked in first position, while in Italy it would be sufficient for the participant to show that it had a good chance of winning the award.⁵³²

However, some findings emerged that came as a surprise at least to some, while others were unexpected by the majority of participants in our seminar. Starting from the standards invoked by the applicant, the general position in domestic laws is that both the right to be heard and the duty to accurately consider all the relevant elements of fact would be important and enforceable for the purposes of judicial review. It is clear that a claimant who seeks to rely on a breach of procedural requirements will have to prove their existence. In terms of procedural fairness and propriety, however, the courts would most probably consider that there was a failure to inform the participant of the cause of exclusion, a breach not justified by any overriding public interest (as might be the case, for instance, in a criminal investigation). They would hold that, had the local authority duly informed the participant, it would have been able to show that there had been an error of fact. They would reach the conclusion that there was evidence of misconduct and that the participant consequently lost its chance of being awarded the contract. There is, thus, a causal link between the infringement of procedural requirements and the damage suffered by the participant. The participant would be awarded compensation for the costs incurred for preparing the offer and participating in the procedure. Lastly, everywhere, including in the legal systems in place under socialist governments until 1989 (such as Hungary, Poland and Romania), the responsibility of both local and national authorities would be recognized in the sense that liability should be shared. In conclusion, there is an area of agreement between legal systems that can be explained by the presence of two forces. Modern governments face similar challenges⁵³³ and often devise similar solutions. Similarity is further accentuated by the common standards defined by EU law.

4 The Violent Police Officer

Thus far, we have looked at various forms of administrative action that give rise to the adoption of legal acts, including administrative decisions, contracts, and acts of general applicability. But administrative functions and powers are not discharged only through legal acts. There is a wide range of activities consisting in the delivery of goods and services, such as free books in schools, and medical treatment in hospitals, respectively. This area is a manifestation of the ‘positive’ State. There is also a vast area of police powers that – manifesting the ‘negative’ State – implies interference with an individual’s rights. The latter area, which has seldom been examined from the comparative point of view,⁵³⁴ will be considered here.

Three preliminary points should be made before presenting our hypothetical case. Firstly, the phrase ‘police powers’ can be understood in either broad or narrow terms. In the broad sense, the police powers doctrine holds that States have the power to regulate the conduct of all citizens in order to protect a public interest. These powers include, for example, adopting rules concerning goods and other things that may constitute a danger to the public (such as chemical substances and inflammable material), general orders limiting the free movement of persons and goods, and individual measures, such as the confiscation/requisition of goods, or the suspension of a business concern. In the narrow sense, police powers to stop and search individuals are among the most contentious aspects of administrative action. They have always been contentious because they impinge on the individual’s liberty in a variety of ways, ranging from a request for information (‘did you see anyone leave that place?’) to the order to leave a place, which, if disregarded, may be followed by arrest for resisting an officer of the law, and a pecuniary penalty, while the most extreme power is to carry out an arrest in the context of a criminal proceeding.⁵³⁵ Secondly, it is highly problematic to compare the exercise of police powers within two or more States. It is one thing to discuss such powers in the context of a liberal democracy or a well-administered State but quite another to discuss them in the reality of an authoritarian government. Perhaps it may turn out that the degree of comparability varies from sub-topic to sub-topic, for example from police powers to patrol highways to the dispersal of an unauthorized public meeting. As a practical example, it seemed advisable to choose the first type of case. Thirdly, in this case, the exercise of power by public officers implies the use of coercion. This raises a whole host of issues, including when and how coercion may be legitimately used and the limits that must be respected in order to avoid arbitrariness. If these ways and limits are violated, can a police officer be said to be acting to fulfil a public duty, or is he or she acting in the same way as a private individual would? From this viewpoint, police officers are not regarded as law-enforcers and protectors of the peace but as law-breakers. The other question that thus arises is what consequences ensue in terms of liability for damages.

In our hypothetical case, we suppose that two police officers stop a driver, Agatha, and ask her quite ruthlessly to get out of the vehicle and show them her papers. Agatha vehemently protests and resists the officers’ request, stating that she is being treated unfairly. One of the two officers, without warning Agatha as required by Police Department rules, moves towards her, grabs her left arm, and twists it into an armlock. The torsion causes Agatha’s elbow to crack, and permanent injury ensues. She refuses any assistance from the police officers and is taken to hospital by some witnesses. Subsequently, Agatha sues the two police officers and the State for damages. The questions which arise are, first, under what conditions, and to what extent, would her court case be successful and, second, whether it would be relevant that the two policemen infringed the guidelines set out by the Police Department.

As might be expected, the comparison of police powers is more complex than, say, the adoption of authorization to sell electronic communications services, an area increasingly regulated by EU law. There are both general and particular legislative provisions on police powers. The latter concern, for example, the prevention of terrorism, with a further element of differentiation depending on whether there are also rules established by regional authorities, as is the case for example in Germany but not in Italy. There are also different rules governing government liability. Thus, in Austria, for example, a direct lawsuit against a police officer is not feasible if he or she was acting in the course of duty.⁵³⁶ On the contrary, in both France and Italy, the action would need to be brought against the State. The underlying rationale is not to accord special protection to public servants but to protect victims from an officer’s actions, with the further distinctive trait that the State will then take action against the responsible officer. Moreover, in the UK, as well as in other common law systems, our hypothetical case would be conceptualized as one of strict liability; that is, a case in which a person is legally responsible for the consequences that flow from the action carried out, even in the absence of any intent (mens rea). There is still another element of divergence concerning the rules governing the exercise of police powers. While in a legal system – for example in France –⁵³⁷ the guidelines defined by the Police Department serve to structure and limit the exercise of power, together with regulations, and their violation would entail the officers’ misconduct, in another only the latter are both legally relevant and binding, so their infringement gives rise to illegality. Other differences concern judicial review. There would be a different tendency to consider whether the police officers can be said to be acting to enforce the law. This would have an impact on the court’s jurisdiction. While in most legal systems the action would have to be brought before ordinary courts, in France the claim would be heard by administrative courts, unless there is misconduct on the part of the police officers, as opposed to ‘faute de service’ (that is, a service fault).

With these caveats, it is noticeable that not only is this hypothetical case regarded as potentially relevant to all the legal systems included in our comparison, but it is also soluble.⁵³⁸ While it is no doubt right in general terms to say that policemen have a duty to prevent crime and that the execution of this general duty involves various powers interfering with personal freedom, at least as far as the use of vehicles is concerned, such powers are not unlimited. If there is an area of public law in which, for all the importance of the ends, they do not justify all means, it is precisely that of police powers. First and foremost, not all means are permitted, but only those allowed by the law.⁵³⁹ Second, the use of coercion must respect some basic requirements, including a clear warning, except when an officer is under attack, which is not the case here. Without this necessary warning, the use of force is illegitimate, also because a warning would have probably avoided the injury. Third, the use of excessive force is prohibited according to a logic of proportionality. Fourth, in all the legal systems examined, an action for damages can be brought against the exercise of police powers, although the court’s willingness to uphold it will be different. Finally, if the infringements of the legal requirements just mentioned are proved, in addition to a causal link with the damage, that is the injury suffered by the victim, liability arises, and the damage can only be compensated for in money. This always covers medical expenses and any loss of profit, for example for the period of time in which the injured person is unable to work. It may also cover non-economic losses. The area of agreement is, therefore, broader than could be expected.

It does not rule out, however, the existence of two further differences which deserve a mention. In some legal systems including the UK, but not all, the courts would most probably also accord ‘exemplary’ damages.⁵⁴⁰ Damages of this kind express the court’s disfavor towards officers who abuse their powers and are intended to prevent the same officers or others from committing the same offences in the future. In other legal systems, notably in Hungary, the courts would probably be willing to accord more weight to Agatha’s conduct, especially her initial refusal to establish her identity and show the papers regarding the vehicle. We can probably see here the legacy of an authoritarian theory of the exercise of State power over the individual.

This is not without problems from the perspective of the echr. The point is well illustrated by the recent case of Vig v Hungary, in which the European Court showed its willingness to interfere with a refusal by domestic courts to acknowledge that the plaintiff had suffered harm caused by the exercise of police powers.⁵⁴¹ These powers were widely exercised against a group of people in the context of a festival organized by a community centre in Budapest. Police forces carried out ‘enhanced checks’, asking participants to reveal their identity. The plaintiff did so because he felt intimidated by the police but later brought a claim before the domestic courts. When his claim was rejected, he argued that his rights under Article 8 echr had been violated. Interestingly, his complaint was not that the police had exceeded their powers under national legislation, but that such legislation was incompatible with the Convention. The Court endorsed his argument in two ways. It observed that interference with the individual’s rights must be necessary in view of a certain objective, which was stated when the administrative action was authorized. This was problematic from the viewpoint of the proportionality of the challenged measure. Moreover, as the domestic court itself had affirmed, it had no power to review either the authorization of the enhanced checks or the operational plan defined by the police force.⁵⁴² The Court thus held that, in the absence of adequate safeguards offering the individual adequate protection against arbitrary interference, Article 8 had been breached, and awarded non-pecuniary damages to the applicant, in addition to the costs and expenses incurred during the proceeding.⁵⁴³ This ruling shows the persistence or return of a positivist-authoritarian theory of the exercise of State powers, as well as the Court’s willingness to challenge it. This is confirmed by other recent rulings, even in the more contentious area of anti-terrorism legislation. Thus, for example, in cases such as Gillian and Quinton and Vinks, the Court endorsed the individuals’ argument that the powers of authorization and confirmation of stop and search were neither sufficiently circumscribed by existing rules nor subject to adequate legal guarantees against abuse and were not, therefore, ‘in accordance with the law’, in contrast with Article 8.⁵⁴⁴ This is not without consequences, because the member States have a fiduciary duty to revise their legislation and practices when these have been found to be in breach of the Convention.

5 Conclusion

Our analysis of liability in various forms of administrative action has taken into account, on the one hand, ‘extreme’ cases, where governmental action is affected by ‘egregious procedural failings’ and is, consequently, more likely to be subject to liability,⁵⁴⁵ and, on the other, more ordinary cases, such as the exclusion of a bidder. Our analysis has sought to ascertain whether there are not only distinctive but also common traits. The former not only persist in general terms, but in some cases also appear more significant than they were some years ago. This is illustrated by the Hungarian case just examined, in which the conflict between the common values and principles, in the context of the echr, and national legislation and practices, is manifest.

At the same time, there is an area of agreement among the legal systems examined here, and it is not limited to the demise of old ideas and beliefs about public law, ‘as administrative authorities have virtually no immunity’,⁵⁴⁶ but extends to the duties of consideration and procedural fairness that public authorities owe citizens and other individuals, as well as social groups and legal entities. These duties reflect to varying degrees the common values recognized and promoted by the Council of Europe and the EU and support the view being advanced here that it is both factually unrealistic and normatively untenable to assume that domestic administrative laws are characterized only by innumerable differences or are even incommensurable. However, the infringement of these duties does not necessarily imply liability. In this respect, we find a further confirmation of the flexibility of procedural justice.⁵⁴⁷