Editor’s Note: Today we publish the 2016 Report on Dutch constitutional law, which appears in the larger 44-country Global Review of Constitutional Law, now available here in a smaller file size for downloading and emailing.
—Nick Efthymiou, Erasmus University Rotterdam; Roel de Lange, Erasmus University Rotterdam
In this contribution, we will first give a general picture of the Constitution of the Kingdom of the Netherlands, with a focus on the absence of constitutional review of primary legislation by the courts. Secondly we will discuss two issues that were important in constitutional developments in 2016. These are a referendum on an EU Association Agreement with Ukraine and a court case involving climate change. They illustrate major issues in Dutch constitutional law. The Netherlands does not have a constitutional court, so in this respect our contribution will differ from most other country reports.
II. Constitutional History
Article 115, par. 2 of the Dutch Constitution as revised in 1848 introduced the formula that ‘statutes are inviolable’. The Hoge Raad (Supreme Court) of the Netherlands judged in 1961 that this wording (then Article 131, par. 2) meant that it had no power to oversee the constitutionality of a certain Act of Parliament, including the procedural aspects. The appellant had claimed that this Act was not in fact an Act of Parliament because it had never been approved in a proper manner and according to the appropriate constitutional procedural rules. The Hoge Raad ruled that it had to assume that the legislature itself had considered the constitutionality of the Act including compliance with the provisions regarding legislative procedure, and that courts have no power to second-guess the interpretation that the legislature itself had given to the Constitution. Article 131, second paragraph, intended to protect primary legislation against constitutional review by the courts. It has always been uncontroversial that this did not apply to all other types of legislation, such as delegated legislation by statutory instruments, municipal legislation, by-laws enacted by professional and economic public authorities, and ministerial legislation. All of these types of rules could be constitutionally reviewed by all courts.
During the 1960s and 1970s, proposals were made to revise the Dutch Constitution, including one by an Official Commission on the Revision of the Constitution to include judicial constitutional review of primary legislation. There was some academic support for this, and the Hoge Raad also appeared to be sympathetic to the idea. Most of the proposals considered constitutional review by ordinary courts, in parallel to their powers with regard to the effect of international law in the Dutch legal order. A separate, specialized constitutional court was hardly ever considered.
However, little support was found for this in political circles.
Meanwhile, developments in the field of human rights led to a further shift in opinion among lawyers. Early observers of fundamental rights protection and its potential functions had already noted that the need for judicial protection of fundamental rights would increase if the government became more and more dominant in its relationship to Parliament, even to the extent that parliamentary participation in primary legislation was no more than symbolic because majority coalitions in Parliament would most often just give their blessings to legislative proposals from the government rather than seeing themselves as co-legislators.
From 1976 onwards, since the European Court of Human Rights (ECtHR) found the first violation of the ECHR by the Netherlands in the case of Engel, Dutch lawyers became accustomed to breaches in the inviolability of Acts of Parliament.
This was reflected in the debates in Parliament on the revision of the Constitution which led to the revised Constitution of 1983. The government’s proposal to replace article 131, second paragraph by the more modern wording of article 120 was accepted: ‘The courts shall not review the constitutionality of Acts of Parliament (statutes) and Treaties’. Did this exclude judicial review of primary legislation against fundamental unwritten principles of law?
In the cause célèbre, which is the case of the Harmonisatiewet, the Hoge Raad concluded that the limits that followed from the ‘traditional place of the judiciary within the political system’ prevented constitutional review of primary legislation against fundamental principles of law.
III. The Constitution Today
This settled the issue of the scope of the new article 120 of the Constitution. Courts were not allowed to review the constitutionality of statutes, neither with regard to their content nor to the legislative procedure, and this included a review in the light of unwritten fundamental constitutional principles.
An important assumption is that the legislature itself has a keen eye for issues of constitutionality. In practice this turns out not always to be the case, especially—but certainly not exclusively—in the field of fundamental rights. There is probably room for improvement in a procedural and a substantive way. Advisory bodies pay attention to constitutionality of legislative proposals, but Parliament is not obliged to follow this advice. There is a role here also for the indirectly elected First Chamber of Parliament.
- The Rise of the ECHR
The Engel case, mentioned above, concerned military disciplinary law, and the violation was only minor. Nevertheless, this alerted Dutch lawyers of the potential of the ECHR as a standard for review of primary legislation. On the basis of article 94 of the Constitution, norms of every hierarchical status can be reviewed against the ECHR, and the awareness of this Convention in the Dutch courts increased. The 1985 Benthem judgment of the ECtHR gave a new impulse to the constitutional effects of the ECHR in the Netherlands. The arrangement of appeal in administrative cases was revised, the Council of State was restructured, and the scope of judicial review was expanded under the influence of article 6 ECHR. This was extended even further by the 1995 judgment in the case of Procola, a case concerning the Luxembourg Council of State which bears a strong resemblance to its Dutch counterpart. In the Netherlands it has had a strong impact at an institutional level. In the course of 25 years, the Dutch Council of State was more or less completely restructured.
In other areas, the ECHR proved equally important. Freedom of speech, protection of family life, detention regimes, and the law of criminal procedure, all saw judgments by the ECtHR with strong impact on the Netherlands.
The ECHR grew increasingly important and the Constitution lost part of its impact and its significance in legal practice. The impact of the ECHR functions as a ‘bypass’ for the lack of judicial constitutional review of primary legislation. That said, it happens only exceptionally that a piece of primary legislation is found to violate ECHR law. More often, the complaints before the Strasbourg court concern administrative action, sometimes court decisions.
The ECtHR considers itself the ‘final authoritative’ interpreter of the Convention, and the Netherlands takes that seriously. In practice, this means that the cases concerning all countries are relevant.
- Constitutional Fundamental Rights and Treaty Rights
Due to the rise of the ECHR, the fundamental rights in the Constitution have become less important in practice. A striking example is the protection of property rights: art. 14 of the Constitution now only protects against certain forms of arbitrary expropriation without compensation, whereas art. 1 First Protocol (‘A1P1’) gives a much broader protection.
In 2010, the Staatscommissie Grondwet (Official Commission on the Constitution) published recommendations to strengthen the constitution by incorporating treaty law, e.g. the rights of access to justice, freedom of expression, and protection of confidential communication. In 2016, the government presented a legislative proposal to introduce a provision on access to the courts into the Constitution. This bill is still pending.
Privacy protection also is mainly a matter of treaty law (now including the Charter on Fundamental Rights of the EU).
- Interpretation in Conformity with Higher Law
In practice, rather than ‘disapplying’ national law in case of conflict with directly effective treaty law, courts prefer interpretation of national law in conformity with treaty law. In itself, this is a familiar part of the canon of interpretation, as we can see from the German examples in interpretation in conformity with the Constitution and the British example of the Human Rights Act (courts should interpret and apply national law in conformity with the ECHR ‘so far as it is possible to do so’), as well as from the European Court of Justice’s case-law in the Marleasing line. In the Dutch context, some spectacular examples can be found of these figures. One of the reasons is that interpretation in conformity avoids the irreversibility problem: if a Dutch court would disapply a statutory provision, the legislature will have difficulty to repair the provision. Interpretation in conformity seems to leave more room for dialogue, both between courts and the legislature, and between national and European courts. The courts strive for consistency and a ‘treaty-conform’ result.
IV. Major Cases
- Climate Change: The Urgenda judgment
The Dutch corporation Urgenda (a contraction of Urgent Agenda) advocates sustainability and innovation. In 2012, Urgenda requested the Dutch government to reduce Dutch greenhouse gas emissions in 2020 by at least 40% in comparison with 1990. When the government did not meet the request, Urgenda started a lawsuit against the Dutch state in 2013. On 24 June 2015, the Dutch district court (rechtbank) at The Hague issued its ruling (ECLI:NL:RBDHA:2015:7145). It ruled that the state acted unlawfully towards Urgenda because of its climate policy. It also ordered the state to reduce Dutch greenhouse gas emissions in 2020 by at least 25% in comparison with 1990.
In September 2015 the state appealed against the ruling. On 9 April 2016 it sent its statement of grounds of appeal to the appeal court (gerechtshof) at The Hague. The procedure before this court will start in 2017.
The Urgenda judgment led to many reactions, even outside the Netherlands. In the Netherlands, many constitutional lawyers wrote about the judgment. From a constitutional point of view, the judgment is highly relevant for the idea of separation of powers (trias politica in Dutch legal parlance).
In its judgment, the district court remarked on separation of powers (paragraphs 4.94-4.102). It acknowledged that a central issue in the judgment was whether ordering the state to reduce emissions thwarted the separation of powers (4.94). It stressed that there is no complete separation of powers in the Dutch legal system. Courts are sometimes obliged to judge the acts of democratically legitimized political institutions. However, when such acts entail policy considerations of divergent interests concerning the organization of society, judicial restraint or abstinence is required (4.95). Urgenda’s claim—asking the court to order the state to cut Dutch emissions—called for restraint (4.96). However, the claim was not outside the domain of the courts. Urgenda asked legal protection and courts should offer this, although the issue of emissions involved political decision-making (4.98). The court also stated that the facts on which the state and Urgenda agreed necessitated wider action by the state (4.99). According to the court, Urgenda’s claim did not amount to an order to the state to take specific legislative or policy measures. Should the court order the state to cut Dutch emissions, the state would still be free to decide in which way it would act (4.101). The district court concluded that separation of powers was no obstacle for ordering the state to cut emissions, as the state would still have discretion (4.102).
In paragraph 15 of its statement of grounds of appeal, the state argued that separation of powers was in fact an obstacle for the court order. The Urgenda claim concerned Dutch climate policy. This policy involved many policy and political considerations, so courts should exercise restraint (15.1, 15.12, and 15.13). Another reason for restraint was that the order to cut emissions had consequences for natural and legal persons who were not a party in the proceedings (15.4). Because of this, the district court ought not to have given the order it did (15.5). The state further argued that a cut in Dutch emissions could only be achieved through acts of Parliament. This made the court order an order for legislation, which courts are not allowed to give (15.14).
Constitutional lawyers were not always positive about the judgment. According to Roel Schutgens, professor of jurisprudence at Radboud University Nijmegen, the district court went too far. The court did not pay sufficient attention to the discretion of the Dutch government in matters of climate policy. The actual policy of the government amounted to a reduction of Dutch emissions in 2020 by 20% in comparison with 1990. The court ought only to have given the order it did if it were obvious that the government could not reasonably have come to its actual policy. According to Schutgens, this was not obvious at all. By giving the order, the court violated the separation of powers.
According to Geerten Boogaard, lecturer in constitutional and administrative law at Leiden University, the judgment created constitutional difficulties. These difficulties concerned judicial construction and separation of powers. The judicial construction in the Urgenda judgment was perhaps too autonomous. And in relation to the separation of powers, the court perhaps talked out of turn and should have left matters of climate policy to political institutions. It would, according to Boogaard, have been better if it had confined itself to a declaratory judgment that the state acted unlawfully if it did not reduce Dutch emissions in 2020 by at least 25%, in comparison with 1990.
The consequences of the Urgenda judgment for separation of powers and Dutch climate policy are unclear, as the state has appealed. What does seem clear is that the judgment might not be in line with Supreme Court (Hoge Raad) jurisprudence. It also seems clear that the court might have made itself vulnerable by the amount of fact-finding in its judgment. This made it easier for the state, in its statement of grounds of appeal, to criticize the court’s interpretation of the facts.
One final remark should be made about the procedure before the courts. The district court gave its order to the state in 2015. The procedure before the appeal court will start in 2017. Doubtless at least one of the litigants will go the Supreme Court after the appeal. By the time the Supreme Court will reach a verdict, we might be close to the year 2020. By then, ordering the state to reduce emissions in 2020 by at least 25% in comparison with 1990 might be pointless.
- The Referendum on Ukraine
On June 27, 2014, the European Union and its 28 Member States made an Association Agreement with Ukraine. This extensive treaty (486 articles, 44 annexes, and 3 protocols) was agreed simultaneously with Association agreements with Moldova and Georgia. These are part of the activities of the so-called Eastern Partnership of the EU, aimed at promoting stability and prosperity in those three countries and Belarus, Armenia, and Azerbaijan. The EU has Association Agreements with 23 states and groups of states, including for example Lebanon, Israel, Turkey, and Central America.
An Association Agreement facilitates free movement of goods, services, capital, and persons, analogous to the freedoms that form the economic core of the EU Treaties. In some cases it has in the past been a first stage of a process that resulted in accession to the EU, but this is by no means automatic.
On July 1, 2015, an Act on Consultative Referenda came into force in the Netherlands. In it, arrangement was made for consultative referenda. It enables referenda with regard to statutes, i.e. legislative decisions of Parliament and government acting together. If a statute has been made, has received royal assent, and has been published, it can be subjected to a referendum before its entry into force. In the Netherlands, the approval of treaties of major significance normally takes the form of a statute. Although there is evidence that it is politically difficult and risky to make foreign policy the object of referenda, the Dutch legislature decided not to include treaty-approving statutes in the list of statutes that are excluded from being subjected to a referendum. Article 11 of the Consultative Referendum Act 2015 states: ‘If it has been definitively established that a referendum has led to an advisory opinion to reject, a bill will be presented as speedily as possible. The bill will propose that the statute will be repealed or it will regulate its entry into force.’ It is up to the legislature to review its own earlier legislative decision, and there are two possible outcomes: either it accepts the advice of the referendum electorate and repeals the statute, or it rejects the advice and sets a date for the entry into force of that statute. The referendum result is therefore clearly not binding on the legislature. Furthermore, the Act requires a turnout of at least 30% of the electorate for any referendum result to be valid.
Almost immediately after the entry into force of the Consultative Referendum Act, a political grouping centred around a rightwing website (GeenPeil) started a campaign to subject the Act of approval of the Ukraine Association Agreement to a referendum. Legislative approval of this association agreement had taken place by Act of Parliament (statute) of 8 July, 2015 (Stb. 2015, 315). The GeenPeil campaign was motivated by a strong anti-EU sentiment, and based on the contention that the Assocation Agreement was the first step in the process of accession of Ukraine to the EU. The initial request—for which 10.000 signatures are required—was quickly successful, and ultimately 427.939 signatures supported the referendum application. The Kiesraad (Electoral Commission) found that among them there were sufficient numbers of valid signatures to make the application succesful (the statutory threshold is 300.000). The referendum was held on 6 April 2016. More than 4 million voters took part, of whom 61% (2,509,395) voted against the approval of the Association Agreement, 38,2% (1,571,874) voted in favour, and 0,79% (32.344) voted blank.
The turnout was 32,8%, which made the result valid.
- Aftermath of the Referendum
While the constitutional status of the referendum result is that of a non-binding opinion, a few days before the referendum most political parties let it be known that they would consider themselves bound to the outcome. After the result and the turnout of only 32,8%, Parliament and government were confronted with a major political problem. They decided not to stick to the non-binding character of the outcome, but to go into a process of re-negotiation with the other EU Member States, all of which have ratified the treaty. Unsurprisingly, progress in this area was not forthcoming. Finally, after more than 8 months, the Dutch government came forward with a declaration by the European Council of 15 and 16 December 2016, in which some of the concerns that were voiced during the referendum campaign were addressed. The Dutch government presented this as a ‘decision’ by the European Council.
The Conclusion of the European Council of 15 December 2016 is worded as follows (under point 23-24):
23. After having carefully noted the outcome of the Dutch referendum on 6 April 2016 on the Act of Parliament approving the Association Agreement and the concerns expressed prior to the referendum as conveyed by the Dutch Prime Minister, the European Council takes note of a Decision of the Heads of State or Government of the 28 Member States of the European Union, meeting within the European Council (Annex), which addresses these concerns in full conformity with the Association Agreement and the EU treaties. 24. The European Council takes note that the Decision set out in the Annex is legally binding on the 28 Member States of the European Union, and may be amended or repealed only by common accord of their Heads of State or Government. It will take effect once the Kingdom of the Netherlands has ratified the agreement and the Union has concluded it. Should this not be the case, the Decision will cease to exist.
Despite the not unambiguous wording of the Conclusion and the Annex, this document can only be read as an interpretive statement of the EU side of the contracting parties. The interpretive statement was not co-authored by the Ukraine government. Therefore, in accordance with art. 31, par. 2 (b) of the Treaty of Vienna of 1969 it can probably only be relevant for the interpretation of the treaty if the other party—in this case Ukraine—will have accepted it.
The conclusion of the Dutch government was that a bill to regulate the entry into force of the Act of approval was to be submitted as soon as possible. After the entry force of this Act, ratification of the Treaty could take place. After this the Association Agreement itself provides (Article 486, par. 2) that it will enter into force on the first day of the second month following the date on which the last act of ratification was deposed.
The analysis presented here was confirmed by the Dutch Council of State’s Advisory Division on January 18, 2017. The Council of State suggested to the government to provide further clarification with regard to the legal status of the Decision of the 28 Heads of State or Government. The government added a few lines to its Explanatory Memorandum. Parliament will debate the matter after the parliamentary elections to be held on March 15, 2017.
All in all, the picture is that the Dutch government and coalition parties, by binding themselves in advance to the outcome of a referendum before they could know what the turnout and the result would be, have manoeuvred themselves into a conundrum on a European level, incurring in a diplomatic result that is unconvincing and most likely gratuitous. This must be part of a learning process with regard to referenda, and with regard to the relationship between constitutional norms (and provisions regarding the status of referenda) and political realities.
 ECHR 8 June 1976 (Plenary), Appl. No. 5100/71 a.o., Engel and others v The Netherlands.
 Hoge Raad 14 april 1989.
 The directly elected chamber is called the Second Chamber.
 ECHR 28 September 1995, Appl. No. 14570/89 Procola v Luxembourg.
 De Wet 2008.
 ECtHR 9 June 2009, Appl. No. 33401/02, Opuz v Turkey, par. 163: the ECtHR will see ‘whether the national authorities have sufficiently taken into account the principles flowing from its judgments on similar issues, even when they concern other States’.
 Gerards 2011:58 sq.
 ECJ 8 april 2014, joined cases C-293/12 & C-594/12: the Directive on data retention was declared invalid by the ECJ. Following this judgment, national legislatures have to react speedily, or cases will be brought before the national courts. E.g. Provisional Court The Hague, Judgment of 11 March 2015, ECLI:RBDHA:2015:2498.
 Donner 1982.
 De Lange 2005.
 See also De Wit 2012.
 See e.g. https://www.theguardian.com/environment/2015/jun/24/dutch-government-ordered-cut-carbon-emissions-landmark-ruling and Lord Carnwath’s speech about Climate change and the courts, held on 26 November 2015. See also Roel Schutgens, Urgenda en de trias (Urgenda and the separation of powers), NJB 2015/1675, (p. 2270-2277).
 The Dutch state referred to Supreme Court (Hoge Raad) jurisprudence to argue the latter point. See HR 21 March 2003, ECLI:NL:HR:2003:AE8462 (Waterpakt) and HR 1 October 2004, ECLI:NL:HR:2004:AO8913 (Faunabescherming/Provincie Fryslan).
 See Roel Schutgens, Urgenda en de trias (Urgenda and the separation of powers), NJB 2015/1675, (p. 2270-2277).
 See Geerten Boogaard, Urgenda en de rol van de rechter. Over de ondraaglijke leegheid van de trias politica (Urgenda and the role of the courts. The unbearable emptiness of the separation of powers), Ars Aequi January 2016, p. 26-33.
 See the following Supreme Court jurisprudence, as mentioned by the Dutch state: HR 21 March 2003, ECLI:NL:HR:2003:AE8462 (Waterpakt); HR 1 October 2004, ECLI:NL:HR:2004:AO8913 (Faunabescherming/Provincie Fryslan); HR 6 February 2004, ECLI:NL:HR:2004:AN8071 (Vrede c.s./Staat); HR 9 April 2010, ECLI:NL:HR:2010:BK4549 (Clara Wichmann/Staat).
 Because both the EU and all of its Member States are parties to the treaty, it is a ‘mixed agreement.’
 Act of 10 March 2015, Stb (Official Journal) 2015, 123.
 The exclusion list mentions budget statutes, revisions of the Constitution, and statutes regarding the monarchy as its most important items.
 The Dutch electoral register is linked to the population register, so no separate registration for any election or referendum is required.
 Kamerstukken (Parliamentary documents) 21501-20, nr. 1176, p. 5-7.