.3.1.Introduction: Monism and Article 94 of the Constitution
The Dutch are widely known to have a very friendly constitutional climate for international law. As we said before, international law takes precedence even over the Constitution itself. This friendly climate essentially originates from the traditionally rather monist approach of the Dutch legal profession. As early as 1919, the Supreme Court expressed its opinion that international law as such is automatically applicable in the domestic legal order. There is thus no need for any kind of conversion to norms of national law.23 Not only are treaty provisions as such accepted as valid law as a matter of customary law. They are also recognized to be of a higher order. Accordingly, the courts generally assume that unless Parliament expressly deviates from its international obligations, it must clearly have intended any provision in its Act to be consistent with a given treaty. This assumption is the basis for the courts’ usual practice to interpret national law as far as possible in a way consistent with the rights laid down in conventions such as the ECHR. And it is this practice that has given rise to a few of the most celebrated but also deeply notorious (some might even say activist) Supreme Court judgments. On such occasions it may well read in the statute some highly detailed rules that have very little to do with either the text of the statute in question or its legislative history.24
To turn back to the supremacy rule: should Parliament legislate expressly against the text and the prevailing interpretation of a treaty, the treaty irrefutably takes precedence over the conflicting statute. This has arguably always been the case but as from 1953, there has been a clear provision in the Dutch Constitution empowering the courts to disapply the statute in question. This provision is currently laid down in Article 94 of the Constitution, which reads as follows:
‘Statutory regulations in force within the Kingdom shall not be applicable if such application is in conflict with provisions of treaties that are binding on all persons or of resolutions by international institutions’.
The key question, which is ultimately for the courts to decide upon, is what exactly constitutes a provision of a treaty ‘that binds all persons’. The importance of the answer to this question lies in the fact that the courts may not disapply the national statute if it ‘only’ conflicts with provisions of international law that do not fit this description. According to Article 93, a treaty ‘binds all persons’ when it is proclaimed and in so far as it contains provisions that may by their very nature be eligible to ‘bind all persons’. This only shifts the issue to what kind of provision would be ‘eligible to bind all persons’.
.3.2.‘Eligible to bind all Persons’ and Judicial Lawmaking
In the current case law of both the Supreme Court and the highest administrative courts, this requirement comes down to two questions.25 First of all, whether the contracting state parties have expressly agreed upon the nature of the treaty provision. This is seldom the case, however. The courts therefore usually convert the question into a matter of justiciability. Does the text of the provision provide the courts with judicially manageable standards to decide the case? In the words of the Supreme Court in its 1986 landmark judgment concerning a major railway strike: ‘does the provision require the legislature to legislate on a certain subject or is it by its very nature eligible to function as “objective law” without further ado?’.26 The real question thus becomes whether the courts are able to derive from the provision some clues as to how to decide cases without having to engage in extensive judicial lawmaking. This brings us near the heart of our subject in this paper. Because if the courts decide wrongly on this issue, they might end up having to decide the case by reading into the treaty detailed rules which the treaty itself is really unable to yield. And they may then be legislating rather than judging the case, which makes them vulnerable to charges of judicial activism. The key criterion (whether the treaty provision textually provides a sufficient degree of manageable standards) therefore theoretically serves as a preliminary question for the courts to solve in order to keep them away from political territory.
What complicates matters, however, is that the decision whether a particular treaty provision is likely to ‘bind all persons’ is generally a ‘yes or no’ decision. Once the courts consider a provision to be self-executing (which we will, for the sake of simplicity, use interchangeably for the phrase ‘binding on all persons’), they consider themselves bound by such a ruling in further cases. Both the circumstances and the context of a specific case are therefore irrelevant when it comes to the question of the self-executing nature of the treaty provision. Deciding whether or not the provision is self-executing is pretty much like deciding whether the patient is pregnant. She either is or is not, but that has little to do with the circumstances. Yet, this may confront the courts with a dilemma. Because although the text might produce a clear outcome in one case, it might equally fail to do so in the next. Phrased differently: the text might yield some clear standards, but those standards might prove insufficient in a particular national context. A clear example is furnished by the principle of non-discrimination as laid down, for instance, in Articles 26 ICCPR and 14 ECHR. These provisions provide the applicant with a relatively clear right so it is usually equally clear for the government what it must or may not do. The question whether a given statute constitutes unlawful discrimination might sometimes pose a challenge to the courts, but usually not one they cannot handle by using a balancing test. The text of these provisions may therefore be considered self-executing. Having met this challenge, however, the court might then face the equally difficult task of providing a remedy for the violation. In some cases there might be several different outcomes of the case, each of which could be equally lawful.
Suppose that the court holds that the exclusion of a certain group of people from a tax exemption is unjustified. Because it is clear what the government should not have done – exclude people from a benefit granted to others – the treaty provisions give the courts relatively clear guidance as to whether there is a violation. They are therefore ‘binding on all persons’. However, just disapplying the statute would either not provide the applicants with a remedy or it would take the courts in political territory because it would grant a benefit to a large group of people where the legislature might just as lawfully have denied it to anyone. The principle of non-discrimination only requires after all that both groups are treated the same, not that they should both have the tax benefit. In such cases the ‘binding on all persons’ requirement itself does not prevent the courts from having to engage in positive lawmaking.
This dilemma raised some discussion in legal literature on the question whether the decision to mark a provision as self-executing ought to be contextual (depending on the characteristics of a given case) or dichotomic by nature. The Supreme Court has never been very explicit on the subject. Several authors concluded from the above-mentioned judgment in the 1986 Railway Strike case that as it was either the agreement between the contracting parties or the text of the treaty provision which was decisive, it must logically follow that the nature of the case in question was not a relevant factor in the decision whether the treaty was self-executing or not. In their view, the Supreme Court took a dichotomic approach.27 Others maintained quite the opposite. In a case in 1984, the Supreme Court had for the very first time in its history explicitly acknowledged the fact that it had a lawmaking role to play.28 But it pointed out that this lawmaking role would have been outstretched had it accepted the claim of an applicant who felt discriminated against and invoked the non-discrimination clause of paragraph 26 of the ICCPR to acquire a right to Dutch citizenship. The Court made it clear that it would have to choose between different outcomes, each of which were equally consistent with the non-discrimination requirement of Article 26. Since that would involve a choice the Court took to be essentially political by nature, it granted that the going practice of the government constituted a different treatment between men en women but it refused to rule on the question whether that constituted a violation of Article 26. Most scholars then concluded that the Court had meant to say that Article 26 was not self-executing in that particular case as it had otherwise refused judgment which the courts are not allowed to do under Article 13 of the General Provisions Act 1829.29
Meanwhile, the general feeling has turned to the dichotomic view. It is important to note in this respect that the Supreme Court itself seems to have abandoned its practice of refusing to rule on the question whether there is a violation. It is still very reluctant to provide a remedy (other than an informal declaration of incompatibility) in cases where that would involve political decision-making, but it does deal with the argument of complainants that the statute in question is incompatible with fundamental human rights law.30 And so it reviews statutory legislation on the basis of treaty law – thereby implying that the treaty is self-executing – even in cases were the remedy remains a political issue. The Court moreover confirmed its new course in its Yearly Report of 1995-1996.
To sum up, fundamental rights review in the Netherlands primarily relies on international human rights documents such as the European Convention and the ICCPR. These treaties automatically have legal effect in the Dutch legal order. Courts may, on the basis of Article 94 of the Constitution, review Acts of Parliament for their compliance with Convention rights if the treaty is proclaimed and in so far as the individual provisions are self-executing. A provision either is considered self-executing at all times or it is not. The key criterion is whether the treaty provision textually provides a sufficient degree of manageable standards for the courts to decide the case upon. The ‘binding all persons’ requirement therefore theoretically serves as a preliminary question to be solved by the courts in order to keep them from having to decide between several political outcomes. However, because the specific constitutional characteristics of a given case do not play a role in deciding the issue whether or not a particular treaty provision is self-executing, the courts may frequently be confronted with a provision that in itself may provide some clear standards but which may nonetheless force the court to engage in positive lawmaking in certain specific situations. These days the courts are very aware of this dilemma and they have tried to cope with it in a careful manner. Before we turn to the case law of the Supreme Court and its reception by legal scholarship, let us first say something about the historical reception and current position of European human rights law in the Netherlands, as they are closely connected to the way the Dutch courts carry out their lawmaking role.
.3.3.The Increasing Role of the European Convention in National Case Law
Although the Netherlands has usually lived up to its relatively monist tradition, it does not follow that the European Convention was always given the full weight in practice it ought to have had on a purely formal basis. As we have said before, with the introduction in 1953 of the current Article 94 of the Constitution, it became common ground that treaty law clearly takes precedence over any kind of legislation. Only a year later, on 31 August 1954, the Kingdom of the Netherlands joined the ECHR and yet, for nearly thirty years the courts remained very reluctant indeed to apply the Convention, let alone disapply legislation violating it.31 Until the 1980s, the judiciary was so cautious that there was hardly one case where the Supreme Court found a violation of a Convention right.32 If a Convention right was involved, the Court would either try to refer to a comparable right in Dutch law or it would deny the self-executing nature of the Convention right. It was also common practice to interpret Convention (or indeed Covenant) rights in such a way that they had either a very narrow scope or a very broad limitation clause.33 Conflicts between national legislation and human rights treaty law thus seemed very rare in the 1960s and 1970s. This led E.A. Alkema to conclude in 1980 that the courts had played only a very limited role in the implementation of the ECHR.34 However, things started to change rapidly soon after Alkema reached this conclusion and already in 1988 the story sounded very different.35 After a remarkable decision of the Maastricht District Court in 1977, disapplying a provision of the 1935 Road Traffic Act due to it violating Article 8 of the ECHR, an era began in which the courts overcame their initial reluctance within a few years.36
The Supreme Court was no exception. In 1980 it ruled that Article 959 of the Civil Procedure Code was to be interpreted in the light of Articles 8 and 14 of the European Convention. The legislature had knowingly established a difference in procedural treatment between cases concerning the custody of legitimate and illegitimate children. In the latter case, it was impossible for relatives of an illegitimate orphan to appeal against a decision of the local magistrate withholding custody. The Civil Procedure Code granted a right to appeal only to legally recognised kin and the legislature had always explicitly taken the view that there was no kinship between illegitimate children and family members of the parents.37 The Court considered the views on the justification of this different treatment of legitimate and illegitimate children considerably changed. This was reflected in the case law of the European Court of Human Rights, notably in its 1979 Marckx judgment.38 This judgment thus served as an argument to replace legislative history as the appropriate method of interpretation. The Supreme Court might have made law in the sense that it created a right to appeal for relatives of illegitimate children. But it is clear that the Court’s understanding of the word ‘kinship’ was rooted firmly in the case law of the European Court interpreting the Convention which, as we know, takes a clear precedence over national law. The same story applied when in 1982 the Supreme Court spontaneously introduced the duty for parents to justify their decision not to let their underage children enter marriage.39 Where refusing their consent would be evidently unreasonable, the courts were allowed to substitute the parents’ withheld permission, ignoring Article 1:36 (2) of the Civil Code which prohibited the courts from allowing a marriage where one of the parents objected to it. Again, this judgment was backed up by several decisions of the European Commission on Human Rights.
Halfway through the 1980s, the Court’s case law was at its peak in terms of self-consciousness. In 1984 it actually went one step further when it explicitly ordered the District Courts to set aside Section 1:161 (1) of the Civil Code, thereby fundamentally interfering in Dutch family law. This provision requires the courts when allowing a divorce to appoint both a guardian and a supervising guardian, consequently implying that parental authority ends with the divorce. On the basis of Article 8 of the Convention, the Court maintained that it should be possible for the courts to leave (joint) parental authority intact when such a course would be in the best interest of the child in question. It such cases the District Court had to set aside Section 1:161 (1), thus effectively allowing for dual custody.40 What was remarkable about this case – which, incidentally, is called the dual custody case – was that this time the Supreme Court had no clear mandate from either the European Court or the Commission when it held that the application of Section 1:161 (1) of the Civil Code violated the Convention. A more marginal and abstract review by the Court – leading to a different outcome – would probably have sufficed.41 Furthermore, the case showed that the Court was prepared to make full use of its power under Article 94 of the Constitution to ignore an Act of Parliament in order to issue relief based on the violation of the Convention.42 The Dutch judiciary evidently was no longer reluctant but appeared to be downright eager to apply Convention law. Some years later, in 1986, the Court issued its famous – or infamous – so-called Spring decisions.43 They showed that the Court had not only overcome its reluctance to apply the Convention. It also developed a rather more self-conscious attitude towards legislation and its own ability to regulate certain areas of law such as family law. The decisions will be elaborated upon in the next section and we will consequently leave it at this for the moment.
The 1980 s are usually regarded as the high watermark in the Supreme Court’s case law concerning fundamental rights review. They showed some of, what few have called the more ‘activist’ judgments of the Court. But they marked the beginning of a slow retreat as well. In some cases, by contrast, it exercised considerable restraint. For instance in the dual custody case we mentioned previously, the Court categorically refused to engage in judicial lawmaking (or rather in a positive sense in any case), and was only prepared to set aside the impugned statutory provision.44 The same year, 1984, witnessed the citizenship case, where the Court refused to remedy an alleged violation of Article 26 of the International Covenant because there were several ways of dealing with the unequal treatment (if there was indeed a difference in treatment) and choosing would mean encroaching on the policy prerogative of the legislature.45 We already touched on this judgment because it has led most authors to believe that the Court had applied the self-executing argument of Article 94 of the Constitution as an instrument to avoid entering into political territory. From the 1990s onwards, the Court explicitly recognised that it was not empowered to set aside national provisions for their inconsistency with Convention law, purely on the basis of its own interpretation of the Convention. In other words, it considered itself unable to offer claimants a broader understanding of the European Convention than the prevailing interpretation offered by the European Court.46 Accordingly, judicial lawmaking without a clear mandate by the European Court of Human Rights remains a phenomenon of the previous century.47
Together with the – as some might say – highly activist ‘Spring’ decisions, this case law created a difficult legacy, both for the Court itself and for legal scholarship. It did confirm that the Supreme Court considered itself competent to assume a lawmaking role – certainly in a negative, but sometimes even in a positive sense. But it raised questions as to what extent the Court was allowed to play such a role and what ought to be its obligations towards the victims of human rights violations. These questions will be discussed in the next section. What may be concluded from the current one is that although the judiciary was reluctant at first to apply the human rights treaties, it gradually overcame its cold feet. The 1980s constituted a phase wherein the Dutch courts accepted the human rights treaties, particularly the European Convention, as a judicially enforceable Bill of Rights for the Netherlands.48 Of course, the 1983 Constitution already provided a civil rights charter, but due to the ban on judicial review and its broad limitation clauses, it had only a limited role to play except perhaps for the political branches. The European Convention provided the courts with an enforceable equivalent.
To some extent, this came as a real novelty to them. For decades the relationship between the courts and Parliament had largely been shaped by the existence of Article 120 of the Constitution, prohibiting the courts from reviewing any Act of Parliament. For all its particularities and exceptions, that provision constituted a bright-line rule for the courts to rely upon. Never before had they been confronted with the difficulties concerning the boundaries of their role with respect to the prerogatives of the legislature. Not to such an extent as they were confronted with in the 1980 s and the years to follow in any case. Their approach to this new question was initially not unequivocal or clear. Legal arguments concerning the positioning of the courts, the Supreme Court in particular, and Parliament scattered among several already existing doctrines. The Court and legal scholarship for instance tried to cope with some of the constitutional difficulties by using Article 94’s self-executing requirement in a somewhat dexterous manner. They also tried to fit in the Supreme Court’s new role in the discussions about its lawmaking role in general, which primarily took place in the fields of civil and criminal law but certainly not constitutional law.49 This attracted the attention of constitutional scholars to the debate on the lawmaking powers of the judiciary. And it is that debate to which we too will now turn our attention.