In the Netherlands, the established rule is that there is NO system of precedent. Lower judges and the four highest courts¹ are in principle not bound by previous judicial decisions. This rule is laid down in Article 12 of the General Provisions Act (Wet algemene bepalingen) of 1829.

In the Netherlands, judgments of the courts are formally not a source of law (in that sense, ‘case law’ or ‘judge-made law’ does not exist). Nevertheless, especially the judgments of the Dutch Supreme Court (Hoge Raad) and the other highest courts are very authoritative and persuasive. As regards court cases, the declaratory theory is formally adhered to: the judge does not create new law, but states the law as it is (even though in everyday practice the fallacy of this approach is recognized, e.g. in case law where the Dutch Supreme Court specifically states that it will fill in a so-called ‘open-ended norm’; see HR 27 November 1981, NJ 1982, 503 (‘Pensioenvereveningsarrest’, also known as ‘Boon v. Van Loon’)).²

The formal adhesion to the declaratory theory is not only undermined by everyday practice, it is also mitigated by the formal recognition of what could literally be translated from Dutch legal doctrine as the ‘law forming task’ (rechtsvormende taak) of the judiciary. More and more the highest courts are enabled to declare their interpretation of a legal matter outside the context of a concrete case before them through various institutional arrangements. The oldest of these arrangements is the possibility for the Supreme Court to give ‘rulings in the interest of the law’ (cassatie in het belang der wet, which is of French origin). ‘Rulings in the interest of the law’ may occur in cases that were decided upon by a lower court, in which all parties involved refrain from lodging cassation appeal. Whereas such a case would normally not find its way to the Supreme Court, the Public Attorney may decide to bring the case before the Court in order for it to rule on the legal question at hand. Its ruling, however, carries no consequences for the parties in the initial proceedings. The only possible rationale behind this instrument is to allow the Supreme Court legal means to assert its position as an authoritative and persuasive interpreter of the law. In 2012 three Acts of Parliament were passed to strengthen this position further. The Act on Strengthening Cassation (Wet versterking cassatierechtspraak) and the Act on Preliminary Questions to the Supreme Court both focus on the Supreme Court. The first Act facilitates the Supreme Court in a fast track ruling procedure to decide on the inadmissibility of appeals in cases where such appeals are manifestly ill-founded and holds lawyers to a higher than usual standard of competence and education to be able to bring an appeal at the Supreme Court. The second Act contains an amendment to the Code of Civil Procedure (Wetboek van Burgerlijke Rechtsvordering) and allows lower judges to raise preliminary questions to the Supreme Court in civil proceedings where there is a multitude of claims of a similar nature. In the explanatory memoranda of both Acts the rationale behind these instruments was justified in the interest of the ‘development of the law’ (rechtsontwikkeling)³ or the ‘forming of law’ (rechtsvorming),⁴ as well as promoting ‘unity of case law’ (rechtseenheid).⁵ The fast track inadmissibility procedure facilitates the Supreme Court by letting the Court concentrate on more fundamental questions of law. The procedure on preliminary rulings aims to accomplish the same by permitting the Supreme Court to establish its line of reasoning in cases or fields of law where large aggregations of legal questions arise. With regard to these preliminary rulings, the amended Code of Civil Procedure obliges lower courts to take the preliminary ruling of the Supreme Court into account in their judgments.

Nevertheless, the Legislature still formally adheres to the general notion of Article 12 of the General Provisions Act. By way of the new legislation and providing the aforementioned rationale, the Legislature makes it more than implicitly clear that although lower courts are allowed to think for themselves, the general expectation is that they will follow the Supreme Court’s case law in the vast majority of cases.

In administrative law, the 2012 Administrative Proceedings Amendment Act (Wet aanpassing bestuursprocesrecht) and its Explanatory Memorandum evoke a similar expectation. This Act amends the General Administrative Law Act (Algemene wet bestuursrecht), and establishes three ‘Grand Chambers’ (grote kamers) of five judges for each of the three highest administrative courts (the Central Appeals Tribunal, the Trade and Industry Appeals Tribunal, and the Administrative Jurisdiction Division of the Council of State). Enabled by the fact that members of one of these courts almost always serve as replacement-members in the other two courts, the three separate administrative courts are to be represented in these Grand Chambers. According to Article 8:10a of the newly amended General Administrative Law Act, proceedings before these Grand Chambers are only justified in the interest of ‘unity of case law’ and the ‘development of the law’.^{6 , 7}

The conclusion of this can only be that Dutch constitutional doctrine is not free from ambivalence. On the one hand, Dutch judges are formally not allowed to ‘make’ law. On the other hand, the national Legislature expects some degree of ‘unity of case law’, a considerable contribution to the ‘development’ of the law, and in some cases the ‘forming’ of law. An exact boundary between especially the latter and judge-made law is not given in this respect.

The practice of application of Article 94 of the Constitution does offer some insight into this matter. Again, according to Article 94 of the Constitution, courts may not apply national law if ‘such application is in conflict with provisions of treaties or resolutions by international institutions that are binding on all persons’. In his farewell address, Supreme Court President S.K. Martens showed that in the vast majority of cases strict application of Article 94 offers no remedy to the parties involved.⁸ In most cases where national legislation was deemed in violation of international law (predominantly the European Convention on Human Rights), non-application of national law would lead to a ‘legal vacuum’. For instance, in a 2003 case of the Administrative Jurisdiction Division of the Council of State, the removal of voter rights of those under legal guardianship (usually mentally disabled persons) under Article 54 of the Constitution was deemed in violation of Article 25 of the International Covenant on Civil and Political Rights because of discrimination between the mentally disabled that were under guardianship and those who were not. The Administrative Jurisdiction Division of the Council of State could have chosen for non-application of Article 54. However, in doing so a situation would arise in which there were no rules for removal of voter rights of the mentally disabled at all. The Division argued that the choice for this or other possibilities of legally addressing this discrimination was not theirs to make. The variety of options was so extensive that according to the Division it fell outside the ‘law forming task’ of the judiciary to choose. In doing so, plaintiffs were told that they were right about the discriminatory nature of Article 54 of the Constitution, but that this was a matter for the national Legislature, to whom this issue was implicitly deferred. In 2008 a constitutional amendment was passed that allows persons under legal guardianship the right to vote.

There is an abundance of examples in which courts argue similarly. According to Martens, courts will only fill the void left by the non-application of national law ex-Article 94 of the Constitution when there is either only one possible way to remedy the incompatibility of national law with international law, or when the national Legislature has been made aware of such incompatibility by previous judgments, but did not act upon it. Case law on this is sparse, especially where it involves the latter option. This latter option does, however, provide some examples of prospective overruling, as we shall see below.

But let us return to the declaratory theory. As a result of the fact that the declaratory theory is formally adhered to, court rulings are, as a general rule, relevant retrospectively. The fact that the court formally only states the law as it is also appears where for a long time it was not the habit of the Dutch Supreme Court to indicate specifically that its interpretation of the law had changed; even the text of landmark cases often does not reflect that they contain a major change. The first case in private law where a different approach was chosen and where the Supreme Court indicated that its interpretation of the law had changed is HR 7 March 1980, NJ 1980, 353 (known as ‘Stierkalf-arrest’).⁹

In the Netherlands, there are no general judicial transition rules. General transition rules as regards judicial decisions may be hard to formulate since much depends on the specific circumstances of the case. J.M. Smits advocates a topical approach to the matter, distinguishing between different types of cases.¹⁰ It has been argued that at least two criteria should be taken into consideration in this specific area: (1) legal certainty and (2) a reasonable degree of predictability of changes in the court’s established interpretation of the law.¹¹

The Dutch Supreme Court has formulated transition rules on an ad hoc basis in a number of cases from 1981 onwards.¹² This happened for the first time in a private law case in 1981: HR 27 November 1981, NJ 1982, 503 (‘Pensioenvereveningsarrest’ or ‘Boon v. Van Loon’ mentioned above). The Court decided that entitlements to a pension of one of the spouses had to be taken into consideration in divorce proceedings when dividing the assets of the spouses. This judgment changed an approach based on a different rule dating from 1959 and, therefore, the Court specifically provided that the new interpretation of the law would only be relevant for the case at hand and future cases. The Court stated, amongst other things, that legal certainty would be endangered if the new interpretation could also be invoked by spouses who had previously been involved in divorce proceedings.

It has been a matter of debate whether this case can be classified as prospective overruling. In the end everything seems to hinge on the definition of prospective overruling: according to various scholars prospective overruling occurs only where the Court decides that its new interpretation is relevant for future cases but not for the case at hand. In Boon v. Van Loon the ruling was also relevant for the case at hand, so the new interpretation was valid ex nunc and, consequently, it is not a case of prospective overruling in the classic sense.¹³ Some Dutch authors nevertheless consider Boon v. Van Loon to be an example of what they call ‘qualified prospective overruling’.¹⁴

When it comes to questions of admissibility, the Criminal Law Division of the Supreme Court has recently been more specific in its approach to judicial transition rules. When the aforementioned Act on Strengthening Cassation entered into force on 1 July 2012, several cases were pending before the Supreme Court that were deemed to qualify for the new fast track inadmissibility procedure. Since no transition period was stipulated by the Legislature, it was up to the Supreme Court to decide whether or not pending cases could be subject to this new procedure.¹⁵ The Supreme Court (or at least, the Criminal Law Division of the Supreme Court) ruled that it would only apply the new procedure in cases in which cassation was filed for after 1 October 2012, because before that date the parties preparing a procedure before the Supreme Court could not have known if and when the Act on Strengthening Cassation would enter into force. Although this case law is prospective in nature, it cannot be seen as prospective overruling, given the fact that the Supreme Court did not overrule its own case law, but rather set a fixed date for application of new legislation.

A final example of a case that cannot be classified as prospective overruling is HR 24 April 1992, NJ 1993, 643. It concerns a case in which the Supreme Court was, nevertheless, concerned about the effects of its ruling in time. In the case, the Court determined when pollution of the environment (soil) can be viewed as a tort for which the State can claim damages from those responsible for the pollution (given the fact that the State will incur costs for cleaning up the polluted soil). The Supreme Court decided that such behaviour can only be considered a tort from 1 January 1975, since at that time it must have been clear for businesses that the State would take action in such cases to clean up the environment and incur costs, something that was according to the Court not clear before this date (although this date might in the opinion of the Court not be applicable in some cases of very severe pollution in which the required clarity may have existed before the date mentioned in the judgment).

The retrospective effect of judicial decisions has been criticized in the Netherlands in legal writing, but only to a limited extent.¹⁶ Generally speaking, the effects in time of judicial decisions have only received attention rather sparingly. This may also be due to the fact that the courts use various techniques to limit the effects of judicial decisions where such is necessary without using transition rules. In his PhD thesis, O.A. Haazen lists a variety of these methods, such as (1) unlimited retrospective effect of the judicial decision combined with an exception for the parties to the action in cases where the application of the rule would not be reasonable and equitable, (2) unlimited retrospective effect with an exception for specific types of cases, (3) retrospective effect for a specific time period only (e.g. 40 years), (4) unlimited retrospective effect unless an excusable error iuris can be proven, etc.¹⁷

Prospective overruling was first discussed in Dutch legal literature in 1950, in an inaugural lecture by J. Drion.¹⁸