The personality of the Directive – one of the EC’s three main legislative instruments – has undergone major alteration in its short history. We shall see two competing trends, noted by Dashwood, – the need to secure the potency of Community law, and the need to follow faithfully the provisions of the EC Treaty – leading to an unholy hierarchy of remedies which, on detailed analysis, reveal the extent of the flaws in the current law.
The earliest relevant cases are those of Van Duyn  and Ratti  in which the ECJ laid out the criteria for what is known as direct effect for unimplemented directives. That is to say, essentially, that directives may have direct effect when a number of conditions. Using EC directive 64/221 as the stimulus, the court recognised, in Van Duyn, that rights conferred by implemented directives are enforceable. However, it is the ‘judicial Freudian slip’ in these two cases that is of most interest.
In Van Duyn, the reasoning behind the decision was that, if rights were not enforceable, EC law would lose its effectiveness as national courts could simply ignore the directives. In Ratti, however, the court changed its tack. The reasoning in this case was the estoppel principle: that states should not be able to profit from their own failure to implement a directive. This is the background to what we may consider the orthodox position of directive effectiveness. Direct effect may be found if it is vertical, that is to say an individual invoking it against the state, but not horizontal. Individuals may not enforce their rights against other individuals as this would compromise the doctrine of legal certainty. Dashwood notes that the reasoning in these cases was not contentious, as it related only to the exercise of the state’s public powers: immigration control and labelling of toxic products.
Marshall v Southampton HA  is considered a crossroads by many commentators. In this case the distinction between vertical and horizontal effect was blurred slightly. Mrs. Marshall sought a remedy against her employer, the health authority, for unfair retirement rules, relying on directive 76/207. Loyally applying the estoppel principle, the notion of the state was broadened significantly (criteria for determining what counted as an ‘emanation of the state’ were formulated in Foster v British Gas). The health authority was labelled part thereof. This, we may say, was the first step on a slippery slope. Though the intention was obviously good – the desire for a fair remedy when a strict interpretation of the no horizontal effect rule would have prevented it – the decision undermines the estoppel principle. It is very tenuous indeed to say that a health authority had any capacity whatsoever to implement the directive. Thus, if we remove the gloss painted on the matter by the ECJ, we can see the first signs of horizontal direct effect slipping in, under the guise of vertical effect.
This should be the first point of reflection on whether the law is sound. Clearly, at this point it is not; indeed, many of the facets of the law seem contrary to many fundamental tenets of European legal tradition. Firstly, the law is capricious as there is arbitrary discrimination between public and private sector employees. Secondly, these decisions seem to run contrary to any properly set culpability requirements. In R (ex. P BT) v HM Treasury, it was recognised that the failure to implement directives need not be a deliberate failure on the part of the member state. Indeed, if a directive is unclear (as was the case in P v S and Cornwall CC, in which it was unclear whether an anti-discrimination directive mentioning “men and women” applied to a transsexual), surely that is the fault of the European institutions. To punish a member state for this failure runs contrary to all that we understand about legal orthodoxy. Finally, a major question has not been resolved: the court, in Marshall, did not choose clearly between general direct effect and the particularity of directives.
More confusion arises between cases such as Faccini Dori, in which the court emphatically refuses to allow horizontal effect, and the notable trends of increasing abstraction in EC jurisprudence and of indirect effect.
In Van Colson and Marleasing, the courts developed a rule known as the consistent interpretation rule, the subtler, more sophisticated sister of indirect effect, which we shall examine later. In Van Colson, the ECJ held that national courts had “to interpret and apply the legislation adopted for the implementation of the directive in conformity with the requirements of Community law, in so far as it is given discretion to do so under national law”. Marleasing extended this to all national legislation. Already we can see the growing force of the unimplemented directive, and the growing force of abstract principle. The ECJ had recognized the difficulties in doing away with the no horizontal effect rule, and has instead empowered the directive as a constructive tool. Just as the British courts will assume that Parliament did not intend to legislate contrary to the Rule of Law, they must now assume that Parliament – even if the domestic law was enacted prior to the directive (following Marleasing) – did not intend to legislate contrary to the directive. It is not difficult to note the increasing legal importance of the unimplemented directive.
Pfeiffer is the next big milestone. In its judgment, the court was quick to emphatically reiterate the principle of consistent interpretation; they then took it to the next level. Following Pfeiffer, not only do the national courts have to read relevant pieces of national law in the light of appropriate directives, but they must interpret the entirety of national law in that way. Some have welcomed the Pfeiffer judgment as a return to the orthodoxy that horizontal effect is not to be permitted, but this seems naïve. The most interesting part of Pfeiffer is the attitude the ECJ expresses towards directives, and their legal force as a whole. In Océano Grupo Editorial, Advocate General Saggio said that “[u]ltimately, the national court’s function as a Community court of ordinary law entails entrusting it with the delicate task of guaranteeing the primacy of Community law over national law. The need to prevent the harmonising action of the Community directives from being compromised by Member States’ unilateral behaviour…” We may submit then, that the Pfeiffer rationale goes beyond mere harmonisation.
Indeed, this takes us back to the nature of the directive as envisaged by the Treaties. It was recognised that there was a need for discretion in member states to allow national law to adapt to the legal output of the EC institutions, and as such, the Directive was created as separate from the directly binding Regulation. As such, alteration to the law was envisaged as being undertaken by the executive and/or legislature of each of the member states in keeping with the democratic traditions of the member states. By compelling the national courts to adapt the entirety of domestic law so as to render it congruous with the directive certainly undermines democracy; this is even more acute if we accept the ‘democratic deficit’ arguments that surround the EC. This is what John Spencer might call ‘legal overkill’; the ECJ has been so desperate to achieve its end of rendering community law more potent that it has conceivably undermined one its primary aims.
Let us not, however, be too apocalyptic. The ECJ seems to have tempered this approach to direct effect in the 2006 case of Pupino, stating specifically that domestic courts could not be expected to create liability “contra legem”. This is similar rationale as that in Duke, in which courts where dissuaded from finding direct effect were there was none.
It is important to trace the legal position through these cases; contradiction and ambiguity are evident. Despite the Faccini Dori ban on horizontal effect, Van Colson, Marleasing and Pfeiffer find an alternate route which is then tempered by Pupino. It is very difficult to find one set principle arising from these cases.
The Pupino rationale, however, leads us into an important point: the hierarchy of effect. In providing that courts should not rule contra legem – the opposite of the obligation in direct vertical effect cases – the ECJ has made it clear that direct effect is the “better” remedy, clearly conscious of the questionability of its approach regarding incidental horizontal effect and the consistent interpretation rules. In addition, post-Francovich (and the subsequent authority that embellishes the ruling therein – Factortame No.5), the ECJ could have resolved that state liability was a solution to these problems. A number of commentators have noted that the court sees this as a last resort.
An individual can sue a state for damages caused by the state’s failure to implement a directive if the directive creates rights for the individual, those rights are clear and there is a causal connection between the failure to implement, and the damage sustained (see Francovich). Added to this, the breach must be a sufficiently serious one (following Factortame No.5). States will not be liable if the rights are unclear (see Denkavit).
Nonetheless, we may safely submit that state liability is not a satisfactory solution, given that, though the claimants are in some part recompensed, the directive itself is not necessarily given effect.
What this hierarchy reveals is a consciousness of the weaknesses of the law on the part of the court. Indeed, the jurisprudence reveals their hesitation over the role of indirect effect, and the rule of consistent interpretation. The conflict between the need to preserve the nature of the Directive, as well as make EC law potent is evident. We should respectfully pose the question, if the ECJ isn’t sure about the guiding principle in all of this, how can they expect anyone else to be?
The confusion seems to lie with what followers of Wade would call the Rule of Recognition. What is the supreme political fact from whence the judges find the authority for their judgments? In some of the above cases, the Treaties have been the guiding beacon, those in which legal orthodoxy has been maintained. In others, however, it seems that abstract principles have been the guiding authority.
This latter point was taken to the extreme in Mangold v Helm . In that case, the German Labour Court referred a question related to fixed term contracts. Germany was given until 2006 to implement Directive 1999/70, which carried a “standstill obligation” and which provided that when such contracts were terminated, substantive reasons had to be given provided the employee was over 58. Though there was held to be no breach on this last point, the fact that Germany had introduced a law extending this to those over 52 was held to be both a breach of the standstill clause – thus highlighting another part of the effectiveness of non-implemented directives – as well as a breach of a “general principle of EC law” against age discrimination.
This is problematic. Firstly, the rules reducing the age limit were due to expire only a month after the implementation had expired. This can hardly be said to be a serious breach in light of Factortame No.5. Secondly, the source of substantive rights is not concrete. The Charter on Fundamental Rights has the ambiguous appellation of “solemn declaration”, rather than any legal standing, and it contains no reference to age discrimination in the workplace. Thirdly, it surely cannot be right that these principles of such importance are decided solely by judges.
It seems that the ECR has taken EC law to an extremely abstract level. It obliges the courts to have mind to absolutely intangible concepts which appear to have no authority outside of the minds of the judges in Mangold. In Wade’s words, “if that is not revolutionary, constitutional lawyers are Dutchmen”.
If the rule of recognition in EC cases is intangible principles then we are at a very uncertain juncture. This was recognised by Advocate General Mazak in the recent Palacios de la Villa decision. The court has, he said “set foot on a very slippery slope, not only with regard to the question whether such a general principle of law on the non-discrimination on grounds of age exists but also with regard to the way it applied that principle”.
To conclude, it seems that it is quite uncertain precisely what legal effect unimplemented directives may have as the ECJ has made great inroads into the domain of horizontal effect, constructing this highly artificial hierarchy of effect. The trend, to the dismay of those who seek legal certainty and orthodoxy, seems to be that they are becoming increasingly powerful as the Pfeiffer and Mangold decisions demonstrate.
Clearly, unsatisfactory is not the word to describe the law’s present position. The hierarchy of effect is evidence that the court’s rule of recognition is unclear and, until it is clarified, constitutional lawyers seem destined to study a continuing stream of contradictory case law. Though judgments following Palacios de la Villa seem more hopefully, we shall hope that the ECJ doesn’t don a Midas costume, and that everything it touches doesn’t turn to Mangold.