The Urgenda Case is Separation of Powers at Work

separation of powers or more abstract considerations of what separation of powers requires.


Introduction
Judicial review of executive and legislative acts (or their absence) on the basis of necessarily open-textured norms, often fundamental rights, raises concerns with regard to separation of powers. Many argue that such review, which necessarily entails interpretation and (political) judgment on the part of the judge, may, at least at times, constitute an encroachment of the judiciary upon the powers of the executive and legislative. The Urgenda case incited a vivid debate on whether the judiciary overstepped the boundaries of its function.
In this short paper, I will take up several of Leonard Besselink's contributions defending separation of powers in light of the growing powers of the judiciary, resulting principally from two developments: First, international legal instruments increasingly play a role as constitutional sources in determining the validity of national rules; 1 and second, Europeanisation and internationalisation strengthen the grip of the (ordinary) judiciary over the legislature and, as a lesser concern, over the executive. 2 In Section 1, I concisely sketch a conceptualisation of separation of powers that I have developed elsewhere. 3 I explicitly link separation of powers with Rainer Forst's 209 theory on the right to justification as the very origin of all human rights. 4 This combination reveals the particular function of human rights litigation and courts in a constitutional democracy, namely to contribute to the legitimacy of the exercise of public power by requiring and allowing for further public justification when rights are restricted. Second, I briefly set out the main issues at stake in the Urgenda case and highlight why it concerns a highly exceptional situation (Section 2). Third, I discuss specific points concerning the separation of powers that arose from the Urgenda case: inherent limitations to the function of the judiciary, the prohibition to order legislation, the positive obligation to protect, minimum obligations, and the right to justification (Section 3).
I argue that the Dutch Supreme Court's Urgenda ruling 2019 is in line with the function of the judiciary. It cannot be criticised for undermining either the specific Dutch conception of separation of powers or more abstract considerations of what separation of powers requires.

Separation of Powers: a concise conceptualisation 5
Separation of powers is not a value as such. It serves other values, namely centrally the protection of collective and individual autonomy. 6 In this reading, it 'should not be understood as a pure instrument of restraining political power. It is also an instrument that constitutes this power.' 7 In a constitutional democracy, collective autonomy is enabled by collective will-formation and individual autonomy is protected by control that is able to place limits on what a majority may do to the individual. Expressed in a different way, ensuring the rule of law as opposed to the rule of men, as the expression of collective autonomy, is recognised as the core purpose of separation of powers. Democracy is more than majoritarianism. In fact, democratic legitimacy is not (necessarily) improved by (an ever) higher level of majoritarianism. Formal procedures, including those that reduce majoritarianism are crucial to the functioning of a (constitutional) democracy.
The rule of law 'has to be set, executed, applied and enforced in order for it to be a rule of law'. 8 Therefore, the rule of law is deeply intertwined with and emerges from specific political and judicial decision-making structures. This is where separation of powers comes in, not as a general blueprint but in the specific expression articulated and practiced in a particular jurisdiction.
Separation of powers goes to the heart of the tension between law and politics. 9 By allocating different functions to the three branches, executive, legislature, and judiciary, separation of powers aims at ensuring that this tension is perpetuated and that neither law nor politics dominates the other. By requiring (a degree of) independence between the three branches, separation of powers aims at protecting a space for political struggle over changing the law. It aims at avoiding that political struggle is suffocated by the constraints of (procedural and substantive) law. At the same time, separation of powers aims at protecting the ordering function of (pre-established) law and to avoid reducing it to just another argument in a political struggle. When politico-administrative decision-making is freed of the pre-established legal framework, determining procedural rules and individual rights, we can no longer speak of a 'constitutional' democracy or the rule of law. When the judiciary determines law at the very 8 B. van Roermund, 'Sovereignty: Unpopular and Popular', in N. Walker (ed.), Sovereignty in transition -essays in European law (Hart Publishing 2003), p. 37. 9 The very distinction between law and politics has repeatedly and convincingly been contested. However, the focus here is not on the (undisputedly) political nature of law but on the outer boundaries of the powers of the judiciary. For this purpose, while acknowledging that any such distinction is imperfect, it makes sense to highlight the tension between law-application, interpretation and development in an individual case as opposed to law-making, representation, and the ability to redesign a policy field in a generally applicable way. moment of adjudication, this would 'eviscerat[e] democracy at its point of application'. 10 In other words, a constitutional democracy and the rule of law require that a distinction is possible and made between the functions of the legislature and the judiciary and separation of powers is the tool to maintain this distinction.
One can go as far as stating that separation of powers, as well as the rule of law, rest on the basic ideal assumption that judges interpret and apply legal norms made in other -political and democratically legitimised -fora and that the judge's role is constrained by and benefits from a particular formality that connects her decisions to the legitimacy of the entire legal order. This assumption is an ideal in the sense that it does not and cannot exist. Law-applying is always also a form of developing the law that cannot be sharply distinguished from law-making. 11 Legislative power is bound by substantive and procedural (constitutional) law, which takes shape through interpretation by the judiciary. It is thus also compelled to engage in applying the law as part of the process of creating it.
The three branches draw from different sources of legitimacy. This justifies their different functions in the institutional practices that ensure, within a constitutional democracy, collective will-formation and control of public power. Judicial decisions are justified by reference to law. They benefit from the legitimacy of the whole legal order as long as they are convincingly reasoned to form a coherent part of existing law. Political reasoning does not require such embedding. Parliament enjoys direct democratic legitimation. Government makes its claim to legitimacy based on a combination of different The focus here is on the judiciary and its function, in particular in human rights litigation, of ensuring respect for the autonomy of all persons. The judge must oblige and enable the policy-maker to justify publicly her choices that restrict human rights. This is, in line with Forst's theory on the fundamental right to justification, 12 the particular function of the independent judiciary to ensure that the policy-maker respects the autonomy of all persons by offering them reasons for restricting their human rights.
In the remainder of this paper, I engage amongst others with the questions that Leonard Besselink addressed in his work, including in the context of the Urgenda case. What are the outer boundaries of the judiciary's powers? When does the judiciary encroach upon the powers of the other branches? How can it contribute to the legitimacy of public power in a constitutional democracy?

The Urgenda Case in a Nutshell
The Urgenda case concerned alleged violations of the Netherlands of the rights to life and to private and family life, under Articles 2 and 8 of the European Convention on Human Rights (ECHR), and a breach of its duty of care in tort law because it did not reduce greenhouse gas (GHG) emissions sufficiently. Judges in three instances (District Court, 2015; Court of Appeal, 2018; Dutch Supreme Court, 2019) determined that the Dutch State was required to reduce GHG emissions by at least 25% by 2020 against 1990 levels.
The District Court reasoned that the State's duty of care in tort, as an open-textured norm, must be interpreted in a manner that is, so far as this is possible, consistent with its obligations under international law. 13 The Dutch government appealed the decision, and Urgenda filed a cross-appeal with respect to the ECHR claim. In late 2018, the Court of Appeal upheld the decision of the District Court, while permitting Urgenda to make a claim under the ECHR. The Dutch government appealed to the Dutch Supreme Court, which affirmed in December 2019 the order of the District Court and the Court of Appeal's admission of Urgenda's claim under the ECHR. 14 The Supreme Court interpreted the Netherlands' obligations under Articles 2 and 8 ECHR by drawing on non-binding commitments of the Dutch State under the United Nations Framework Convention on Climate Change (UNFCCC), soft law sources such as Conference of the Parties' (COP) decisions, and 'scientific insights and generally accepted standards'. 15 It found the threshold test for a positive obligation satisfied, acknowledged that climate change poses a sufficiently 'real' and foreseeable risk of harm, and added that a degree of uncertainty regarding the timing, nature and extent of the specific harm that climate change will cause in the Netherlands cannot change this conclusion. 16 The Supreme Court confirmed that 'the genuine threat of dangerous climate change' 'constitutes "real and immediate risk"' 'that the lives and welfare of Dutch residents could be seriously jeopardised' 17 and identified this as 'an exceptional situation'. 18 It further confirmed the Dutch State had an individual responsibility 'to do "its part" in order to prevent dangerous climate change, even if it is a global problem'. 19 In its complex reasoning, the court relied on the 25-40% emission reduction target by 2020 formulated by the Intergovernmental  17 Ibid, para 5.6.2 (emphasis added). 18 Ibid, para 8.3.4 (emphasis added). 19 Ibid, para 5.7.1.
Panel on Climate Change (IPCC) 20 for developed countries for having a likely chance of holding temperature increase to below 2°C. This target had been endorsed in numerous decisions of the COP, 21 the European Union and the Netherlands. 22 It also considered the risk of tipping points resulting in irreversible change, 23 and the lack of available technology to remove CO2 from the atmosphere 24 and concluded that the 25% emissions reduction was indeed 'to be regarded as an absolute minimum' of the positive obligations under Articles 2 and 8 ECHR. 25 The District Court and the Court of Appeal had earlier argued the same outcome on the basis of Article 6:162 of the Dutch Civil Code and Articles 2 and 8 ECHR, respectively, equally relying on non-binding commitments, principles of international law, and scientific insights for the interpretation of these open textured legal norms.
Importantly, the Supreme Court considered the determination of 'whether the measures taken by the State are too little in view of what is clearly the lower limit of its share in the measures to be taken worldwide against dangerous climate change' a task of the judiciary. 26 It emphasised that in light of the grave threat of dangerous climate change to the enjoyment of human rights the principle of effective legal protection in Article 13 ECHR entails that 'the courts must examine whether it is possible to grant effective legal protection by examining whether there are sufficient objective grounds from which a concrete standard can be derived'. 27 Equally importantly for the present discussion, the Court pointed out that a State's duty of 'due diligence' required the Netherlands to substantiate that its emissions 20 The IPCC is the United Nations body for assessing the science related to climate change. See for its work: ipcc.ch, visited 2 February 2021. 21 Hoge Raad 20 December 2019, (Urgenda), ECLI:NL:HR:2019:2006, para 7.2.6. 22 Ibid, para 7.2.1-3 and 7.2.6. 23 Ibid, para 7.2.10. 24 Ibid, para 7.2.5. 25 Ibid, para 7.5.1. 26 Ibid, para 6.3. 27 Ibid, para 6.4. 215 reduction measures were 'responsible' 28 and that it 'ha [d] not been able to provide a proper substantiation of its claim that deviating from that target is nevertheless responsible'. 29 Nor, concluded the Supreme Court, had the State 'sufficiently substantiated' that the required reduction imposed an 'impossible or disproportionate burden', 30 including in light of the fact that some other EU Member States pursued 'much stricter climate policies', and the Dutch State had known about the required reduction since the first court decision in 2015. 31

Boundaries of Judicial Powers in Urgenda
While many commented on the Urgenda cases (2015, 2018, 2019), Leonard Besselink is one of the scholars who engaged in a serious constitutional analysis of the role of the judiciary in these decisions delivered by the District Court and the Court of Appeal. In a piece published in 2018, he challenged the District Court's reasoning that it could essentially give effect to non-binding international norms by reading them into open textured norms of the Dutch Civil Code. 32 He also took issue with the Court of Appeal's reasoning of reading non-binding norms into Articles 2 and 8 ECHR, questioning how this was different; why the Court explicitly mentioned the lower limit of 25%; and whether it did not fall within the discretion of the executive to find ways to reach the required outcome. 33 The Court of Appeal's ruling, in Leonard's view, still amounted to an order of the judiciary for the legislature, even if it primarily concerned an obligation for the executive. 34 28 Ibid, para 7.2.1. 29 Ibid, para 7.5.1 (emphasis added). 30 Ibid, para 7.5.3. 31 Ibid, para 7.5.3. 32 L.F.M. Besselink, 'De constitutioneel meer legitieme manier van toetsing: Urgenda voor het Hof Den Haag', 93(41) Nederlands Juristenblad (2018), p. 3078-3082. 33 Ibid. 34 Ibid. 216 In 2019, Leonard clarified that his earlier contribution was concerned with the prohibition of judges to legislate, 35 and not with the particular status of non-binding international instruments. 36 He pointed out that the Urgenda case concerned not so much interpreting national law in conformity with international law but the review of public policy in light of binding and non-binding international norms and offered contextual considerations on the legitimacy of the constitutional boundaries of judicial and legislative powers. In 2020, Leonard concluded that the Dutch Supreme Court's ruling brought about 'an undesirable change in the separation of powers between the judiciary on the one hand and the legislature and executive on the other.' 37 He cited the Supreme Court as explaining that in the Netherlands 'a tradition of non-review' exists for judges and explained in a detailed analysis of the Supreme Court's case law where the boundaries lie and how they developed over time. 38 In the following sections, I engage with issues that emerge from the Urgenda case and from Leonard's reflections on it, such as: does the judge in the Urgenda case overstep the limits of judicial power? May the judiciary establish a specific positive obligation and hold government to it? May the judiciary order government (and potentially the legislature) to act? What is the difference between a standard and a minimum obligation? Does the public have a right to justification of governmental policy departing from repeated political commitment and overwhelming scientific evidence?
Separation of powers is a general constitutional principle that is given effect in different legal systems in very different ways. No blueprint exists of how legal systems should divide competences and draw lines between the different branches in practice. The Dutch legal system, as every other, opts for a particular version of fundamental 217 rights review. The first two subsections reveal in many ways how the idiosyncratic features of a specific legal order have a role in establishing the boundaries of judicial action and assessing judicial action in light of separation of powers (Sections 3.1-2). The following three sections put forward considerations that would play a role in any legal order (Sections 3.3-5).

Prohibition to Order Legislation
The Dutch Constitution rules out ex post constitutional review, but makes an exception for reviewing laws in light of provisions of conventions and decisions of international organisations with the option of disapplying national law in the case of conflict. 39 The Dutch Supreme Court introduced further restrictions in Waterpakt concerning the implementation of an EU Directive. 40 Dutch courts may not, on the basis of their constitutional position, intervene in the process of formal legislation. They are prohibited from giving an injunction to legislate. 41 Acts of parliament are enacted jointly by the government and the States General based on political decision-making and a balancing of all interests involved.
This makes sense in light of the understanding of separation of powers introduced above. Courts regularly engage in the creative process of developing the law when they decide individual cases. However, the task of developing generally applicable laws, out of its own motion or on the basis of an impetus by the executive is reserved to the legislature. Disapplication of national laws because of a conflict with EU law is also different from an injunction to legislate. The former is limited to the parties of the case, while the latter would result in a generally applicable measure.
In Waterpakt, both sides agreed that it was necessary to amend or adopt a formal act of parliament in order to comply with the relevant 39 Articles 120 and 94 of the Dutch Constitution.

218
EU directive. In Urgenda, the point of whether the Dutch State could, as a matter of principle, comply with the Supreme Court's Urgenda ruling without having to adopt formal acts of parliament remained less clear.
What is clear is that Urgenda did not formally order the legislature to act but ordered the State as one of the parties to the judicial proceedings to comply with an obligation. 42 While most of the possible ways to reduce emissions would argulably have to rely on legislation, the question of how to comply with Urgenda is a choice of means and substance of the Dutch State, i.e. which of the countless factors affecting the emissions from Dutch territory should be addressed. 43 This was also the position of the Supreme Court, which interpreted the prohibition to order legislation, as established by Waterpakt, as preventing courts from 'interfer[ing] in the political decision-making process regarding the expediency of creating legislation with a specific concretely defined content by issuing an order to create legislation'. 44 On the means, it is relevant that the State had been on notice since the first Urgenda ruling in 2015. Meeting the absolute minimum of emission reduction would have been possible at an earlier stage with a much wider array of means. Yet, the Government failed to act earlier. On the substance, even if -particularly at the time of the decision of the Supreme Court in December 2019 -compliance was best achieved by adopting legislation (means), the decision did not demand any specific, concretely defined substantive content. This is indirectly, and perhaps ironically, confirmed by parliament's decision of 26 January 2021 to declare the execution of the Urgenda ruling 42 Besselink, 'De constitutioneel meer legitieme manier van toetsing', supra n. 32, p. 3079, 3081. 43 Absolutely any change in behaviour has an impact on the level of emissions. Hence, it is possible to reduce emissions by executive order. That as such cannot be disputed. Otherwise, the conclusion must be executive decree cannot affect behaviour. We can see this as a result of the restrictions to mitigate the COVID-19 pandemic -many of which were introduced by executive order. Even if, arguably, this may not be desirable from a separation of powers perspective.

219
'controversial' and so stopping the current caretaker government from executing it. 45 As complying with a court ruling, confirmed in three instances, can hardly be controversial in a constitutional democracy, the controversy must relate to the 'how' of giving effect to Urgenda.

Public interest organisations
Urgenda, just as Waterpakt, concerned public interest litigation. Urgenda is a campaign organisation, whose first and foremost purpose is to engage citizens with the pressing issue of climate change in many different ways and by different means, including litigation. 46 The political agenda setting powers of non-profit organisations, including through litigation, testifies to the socio-political consequences of public interest litigation. 47 In the Netherlands, public interest organisations enjoy standing to articulate the general interest before court. Courts then rule on situations that do not only concern the parties to the dispute but by definition raise issues of general interest and affecting the broader public. Some consider this a form of participation by citizens through courts as part of forming a diverse public sphere. 48 Other see particularly this form of general interest articulation to encroach upon the core task of the legislature. 49 220 Whether third parties, including public interest organisations, have the right to bring such actions is a question for the democratic constitution-maker or the legislature. The role such review may play in the overall constitutional structures that ensure separation of powers depends on many related choices, namely whether constitutional review is possible and what the precise mechanisms are that enforce fundamental rights guarantees. Arguably, the Dutch legal order gives comparatively generous standing to public interest organisations as a counterweight in structures of overall weak constitutional review.
The combination of a positive obligation, i.e. duty to protect, and standing for public interest organisations, may -as in Urgenda -allow third parties to bring judicial challenges with the objective of obliging state bodies to act, without necessarily having a claim that their own rights are infringed. I would agree that, as in any politically salient case, the judge should tread carefully. She should consider her function within the constitutional order and explain well how the individual decision rests on the legitimacy of the law, i.e. that a consistent reading in the light of the whole legal order virtually requires a judicial decision. Such decision may for example be required in order to counter an aberration that is made possible by a gap in or an ambiguity of the law. She should explain why this decision does not constitute a substitution of a legally defensible political judgment with a judicial judgment. In my view, the Supreme Court has in Urgenda provided the required justification for its ruling. My view is based on three central reasons that I explain in what follows.

Positive obligations and reflexive effects
In the Urgenda case, the Dutch Supreme Court required executive action (and arguably, at least at this late stage, also legislative action) based on positive obligations flowing from fundamental rights. Such positive obligations to protect fundamental rights are a particularly problematic issue for separation of powers. This is in particular the case when legislative obligations are concerned. Such obligations allow courts to intervene at the beginning rather than at the end of the law-making cycle, 50 where it is more difficult to argue that they protect individual autonomy from (illegal) interference by the collective. In other words, the court's intervention at the beginning of the law-making cycle does not benefit from the democratic legitimacy of parliament (or government). The argument that the court's intervention protects individual rights can also reasonably be made in the case of positive obligations. The futurelooking nature of enforcing a positive obligation in combination with the fact that a public interest organisation relies on rights without being itself directly affected makes the argument on the protection of individual autonomy more abstract in the Urgenda case than in many other cases. Yet, it does not make it unconvincing.
The Dutch Supreme Court took the position that it cannot be legally possible that there is a positive obligation and the government chooses to do nothing at all. 51 This is the case, in particular in light of the exceptional threat of dangerous climate change. 52 '[C]ourts can assess whether the measures taken by the State are too little in view of what is clearly the lower limit of its share' because this is in breach of its obligation. 53 This assertion constitutes a watershed moment in the ruling.
As Leonard Besselink convincingly pointed out, 54 all three Urgenda rulings, of the District Court, Court of Appeal, and Supreme Court, concerned similar and controversial effects of neither directly effective, nor legally binding international norms. When non-binding international norms are used for interpreting binding open-textured national (tort) law these effects are called 'reflexwerking' (reflexive effects) and when used for interpreting positive obligations under Article 2 and 8 ECHR they are called reflection ('weerkaatsing'). However, they raise the same concerns about the 50 Möllers, The Three Branches, supra n. 6, p. 138. 51 Hoge Raad 20 December 2019, (Urgenda), ECLI:NL:HR:2019:2006, para 6.3. 52 Ibid, para 6.3 referring to paras 5.7.2-5.8, which in turn refer to paras 5.7.9 to 5.6.2. 53 Ibid. 54 Besselink, 'De constitutioneel meer legitieme manier van toetsing', supra n. 32, p.

3081.
222 boundaries of judicial powers, because in both cases the judge gives effect to non-binding international law. The rather technical difference of whether this happens through national (tort law) or directly effective international law (ECHR) does not matter for the separation of powers criticism. Leonard argued that the judiciary should be more deferential to the policy maker when such reflexive effects are concerned 55 and that courts should be precluded from reviewing highly politicised pieces of legislation when no 'clearly defined directly effective rights of citizens' are at stake. 56 Much depends on framing here. In principle, it seems convincing that non-binding and non-self-executing international norms should not become binding positive obligations simply by being reflected in open-textured binding norms. 57 However, approached from the other end -as the Supreme Court did it -legally binding (albeit open-textured) obligations would be rendered pointless if it was possible for the State either not to act at all or fail to make an absolute minimum of effort. By definition, open-textured norms require further interpretation (specification?) when applied to the circumstances of the individual case. For such interpretation, the Court commonly relies on a range of means of interpretation that are not binding.
Another issue is the guidance that courts can draw from non-binding means of interpretation when construing the meaning of binding open-textured norms. Here, a lot depends on the specificity of the means of interpretation. Cases could be imagined, in which reflexive or reflective effects result from a wild growth of non-binding but de facto powerful norms agreed in complex exchanges of national executive actors in the international realm. Non-binding international norms are often criticised for being equally vague as the open-textured norms on which judges rely to justify their position on highly 55 Besselink, 'Naschrift', supra n. 36. See above for the discussion of reflexive effects. 56 Besselink, 'Annotatie Waterpakt-arrest', supra n. 40, p. 1452. 57 Read however L. Burgers & T. Staal, 'Climate action as positive human rights obligation: The Appeals Judgment in Urgenda v The Netherlands', 1 CSECL WP (2019), available at: papers.ssrn.com/sol3/papers.cfm?abstract_id=3314008, who rightly speak of the 'factual rather than legal significance' of the non-binding norms. 223 politicized issues. 58 However, this is not the case in Urgenda, in which the international instruments outlined minimum requirements in specific numerical form, namely 25%-40% reduction in 2020 compared to emission levels in 1990. The fact that the international instruments indicated a range does not contradict this. The absolute minimum of 25% could hardly be expressed more specifically. This also explains that the Urgenda ruling specifically obliges the Dutch State to reduce its emissions by at least 25%. The executive and the legislature are free to choose any higher level of reduction but not one percentage point lower -at least not without offering a justification! 59

Establishing an Absolute Minimum: the Task of the Judge?
Leonard Besselink did not mince his words, stating that '[t]he delegitimization by a judicial decision of […] democratically decided policy changes based on a balancing of major general interests by political organs of the state limits the democratic quality of political decision-making and affects the democratic quality of the rule of law. The judge is not called to make such balancing, if he could make it at all, on the basis of a case brought by a general interest organisation that limits itself to one general interest, no matter how great and important.' 60 This statement does not capture what the judge did in Urgenda. As explained above, I agree with Leonard that it is not the task of the judge to balance general interests and substitute her position for the one of the political decision-maker. Judges in Urgenda also did not balance general interests, nor did they 'delegitimize' the policy of the 58 This seems also to be Leonard's point on the District Court: Besselink, 'De constitutioneel meer legitieme manier van toetsing', supra n. 32, p. 3079. 59 See Section 3.5 below on this point. 60 Besselink, 'Rechter en politiek', supra n. 37, p. 144 (translation and emphasis by the author). See also, N. de Boer, 'Trias politica niet opofferen voor ambitieuze klimaatpolitiek', 72(1) S&D (2016), p. 40-48, at p. 41 (translated by the author), stating that a better climate policy cannot be reached by sacrificing separation of powers.

224
State. They declared inaction, i.e. failing to take an absolute minimum of action, illegal.
In this context, it is important to realize that 25% is not a 'standard' but the absolute minimum. In Urgenda, this absolute minimum was taken as an interpretive aid for applying a legally binding obligation and concluding that without justification this obligation is breached. That the court can only identify the absolute minimum threshold for legal policy-making is core to protecting separation of powers that allows political actors to determine what measures are possible, advisable, and desirable above this absolute minimum.
Prima facie unrelated to the issue of separation of powers, Leonard states that in terms of preventing harmful climate change the absolute minimum indicated by the Supreme Court is 'too little too late'. 61 He argues that Greta Thunberg is right and not the judges in Urgenda because they introduce a minimum standard as 'sufficient'. 62 I disagree on several levels. First, the whole ECHR, which concerns minimum protection rather than setting a standard, could be criticised in the exact same way. Second, in reality, Urgenda had the effect of reducing emissions from Dutch territory. The State did much less before Urgenda than after. In fact, as I state above, it waited until after the confirmation of the ruling in last instance before it took any measures. 63 Third, and this point brings it back to separation of powers, if the Supreme Court had decided on Greta Thunberg's complaint before the Committee on the Rights of the Child (to which Leonard refers), which inter alia aims to hold five states 'responsible for failing to prevent foreseeable human rights violations caused by climate change by reducing its emissions at the "highest 61 Besselink, 'Rechter en politiek', supra n. 37, p. 125. 62 Besselink, 'Rechter en politiek', supra n. 37, p. 129. 63 See: 'Kamerbrief over uitvoering Urgenda-vonnis', rijksoverheid.nl/documenten/kamerstukken/2020/04/24/kamerbrief-overuitvoering-urgenda-vonnis; for information in English, see: J. Watts, 'Dutch officials reveal measures to cut emissions after court ruling', The Guardian 24 April 2020, theguardian.com/world/2020/apr/24/dutch-officials-reveal-measures-to-cutemissions-after-court-ruling. 225 possible ambition"', 64 it would have overstepped its function and undermine the separation of powers [I do not say that it would have done this!]. Determining what the 'highest possible ambition' is in light of all (including prominently economic) interests is the role of the political decision-maker. In other words, you cannot have your cake (oblige the Dutch State to reduce emissions to the reasonably required level) and eat it (respect its judicial function under the separation of powers). The Supreme Court could not go further, even if it agreed with Thunberg that 25% is too little. 65 It was limited by the boundaries of its function within the constitutional democracy of the Netherlands to only identify the absolute minimum for legal action, rather than identify what is needed to mitigate climate change.
Leonard Besselink argued that 'judicial intervention is, in short, unauthorized when a multitude of alternatives exists to solve an issue of political policy'. 66 However, in the case of dangerous climate change (issue of political policy) there is only one possibility and no alternative: reduce emissions. The questions are only: how (by which means) and how much. The former was left to the policymaker. As regards the latter, very convincing arguments support the conclusion that it would be much more reasonable in light of the politically confirmed, overwhelming science to reduce 40% rather than 25%. Could it be justifiable to reduce less than the absolute minimum? Yes, a justification may be difficult in the 'exceptional situation' 67 of 'the genuine threat of dangerous climate change […] which could render part of the Netherland uninhabitable', 68 but nonetheless conceivable. However, as I discuss below, the State did not offer one. 69 226

Right to Justification
Human rights directly protect individual autonomy. However, human rights play a more complex role than keep-out-signs for policy makers. This role is linked to their very origin in a right to justification as explicated by Rainer Forst. 70 Forst developed a distinct conception of moral autonomy, that lies at the core of all human rights, and requires a reason-giving process when these rights are restricted. He concluded that the core of the human rights discourse is the understanding that being a person ('Personsein') depends on being treated as a being with a right to justification ('Rechtfertigungswesen'); this is as persons to whom one owes reasons. 71 Human rights are hence also a language of critique ('Sprache der Kritik'), which protects citizens from unjustified societal and political circumstances of oppression. Constitutional structures must allow for a reflexive engagement between those affected and those exercising public power. Those exercising public power in a way that restricts rights have to offer their citizens legitimate reasons. 72 This is where the function of the judiciary comes in. Judges should -pursuant to formal rules -oblige the policy-maker to justify her choices in public in light of their impact on human rights. This is where separation of powers comes in. It is needed to ensure the sincerity of the reason-giving process. It reconciles law and politics in practice in a way that one never fully dominates the other and must therefore still answer to the other. One driving question in the analysis of the suitability of any specific model of separated powers should therefore be whether it establishes a process of reason-giving. Another question, which may contribute more indirectly to the process of reason-giving, is whether and how public judicial proceedings activate citizens and motivate them to demand better justification for public policies. 70  In the specific case of Urgenda, one can hardly doubt the activating effect. In Urgenda, human rights were invoked, including to stir up public debate and press for such justification. While the case also concerned the protection of the human rights and hence individual autonomy of those most affected by climate change, it invoked human rights in favour of all Dutch residents, and arguably mankind as a whole. In other words, the court in Urgenda did not protect the rights of identifiable individuals from interference by the collective but the rights of the collective from present and inevitable future interference. Arguably, this blurs the distinction between individual autonomy and the collective in the function of the judiciary, as outlined above. 73 It does not take away from the underlying right to justification.
In the Urgenda case, the Dutch state was time and again invited to justify its choice not to develop a climate policy aimed at reducing emissions by the agreed absolute minimum of 25% needed to avoid dangerous climate change. Yet, the Dutch State refused to offer such justification.
In the words of the Supreme Court, 'under certain circumstances, the State must properly substantiate that the policy it pursues meets the requirements to be imposed, i.e. that it pursues a policy through which it remains above the lower limit of its fair share.' 74 Yet, '[t]he State has not provided any insight into which measures it intends to take in the coming years, let alone why these measures, in spite of the above, would be both practically feasible and sufficient to contribute to the prevention of dangerous climate change to a sufficient extent in line with the Netherlands' share. The State has confined itself to asserting that there "are certainly possibilities" in this context'. 75 On the contrary, '[t]he State acknowledges the fact' that 'any postponement of the reduction of emissions therefore means that emissions in the future will have to be reduce on an increasingly large scale in 73 A point highly relevant for separation of powers that we explore futher as part of our NORFACE project. 74 Hoge Raad 20 December 2019, (Urgenda), ECLI:NL:HR:2019:2006, para 6.5. 75 Ibid, para 7.4.6. 228 order to make up for the postponement in terms of both of time and size.' Postponement makes the necessary measures hence 'increasingly far-reaching and costly', as well as 'riskier'. 76 The point that the composition of the Dutch government has changed 77 since it repeatedly politically endorsed the very minimum standard that the judges used in Urgenda to interpret the government's obligation not to infringe fundamental rights, ignores that the Dutch government did not, despite ample opportunity in the Urgenda case, 78 dispute the science that underpins this minimum. 79 This is the nature of legal proceedings. Facts that are not disputed by either party are presumed as correct and accepted by both parties.
No reasonable justification was offered for the choice to postpone and herewith magnify the problem. The government could not justify why producing more emissions that stay in the atmosphere for many years to come and continue heating up the planet is on the balance of all involved interests a reasonable choice. The State remained focussed on formal legal arguments without offering any political considerations that could potentially justify its choice for staying below the scientifically identified necessary minimum. This is why the Dutch state lost. In a constitutional democracy the government is formally legally obliged to justify its actions and their consequences when asked in court to do so. Such justification takes place in light of binding law and within the procedural rules of the relevant legal system. The Dutch State could not or at least did not justify its actions. It certainly could not show why reducing emissions less until 2020 was a reasonable way of reaching the objective of reducing emissions more by 2030. However, the government also did not use the occasion to raise any other considerations that would 76 Ibid, paras 7.4.5 and 7.4.3. 77 Besselink, 'Rechter en politiek', supra n. 37, p. 144. 78 Certainly in first and second instance. 79 De Boer, 'Trias politica', supra n. 62, acknowledges a bit futher down: ' [de] wetenschappelijke inzichten stonden overigens niet ter discussie tussen de partijen (de Staat en Urgenda).' 229 allow the judge, the claimants, and the broader public to recognise the reasonableness of its choices.
On the contrary, the Supreme Court pinpointed the inherent contradiction in the state's argument: 'there may be serious doubts as to whether, with the 20% reduction envisaged by the State at EU level by 2020, the overall reduction over the next few decades, which the States itself believes to be necessary in any case, is still feasible.' 80 Finally, an inherent justification could have been that the demanded measures cross the limits of feasibility or place a disproportionate burden on the state. The European Court of Human Rights acknowledges this. 81 Yet, here the government confirmed that a much stronger reduction is still feasible.
In demanding justification, the Urgenda case is not exceptional but does what many human rights disputes do. The Dutch Supreme Court made this point very explicit: 'A state must take due diligence into account in its policy. The court can determine whether the policy implemented satisfies these requirements. In many instances found in ECtHR case law, a state's policy has been found to be inadequate, or a state has failed to provide sufficient substantiation that its policy is not inadequate.' 82 The State failed to comply with the right of justification of those affected by its deficient climate policy. For this reason, it was ordered to develop a policy that meets what the government had repeatedly acknowledged to be the absolute minimum to prevent dangerous climate change. It is precisely the function of the judiciary to demand and herewith allow the policy maker to offer (additional) justification for policies that interfere with human rights and that allegedly 230 are unjustified. This is a confirmation of a working system of separation of powers.

Conclusions
Leonard Besselink argued already in 2005 that the triangle of the three branches, with the legislature on top, formulating generally binding laws, and the executive and the judiciary on the bottom, executing and applying these laws, has been toppled in the 20 th century. 83 Europeanisation has strengthened the executive and the judiciary, which are now on top, and weakened the legislature and moved it to the bottom of the triangle (now on its head).
In first instance, the Urgenda case concerned primarily national tort law. Both in second and final instances, it rested on a human rights argument under the ECHR, for which a similar argument could be made with regard to the judiciary.
Human rights litigation is a context in which often controversial social issues come to the fore and many interests play a role, while legal standards are open-textured and offer little guidance for the decision in the specific case. In such cases, it is wise for the courts to act cautiously. However, this does not mean that judges, in a system of separated powers, play no role in such cases to protect the rule of law and democracy.
Separation of powers cannot be assessed in abstract. It is always a matter of contextual practice in specific historical and jurisdictional circumstances. Within the Dutch legal system, it is possible for organizations that protect a common interest, such as Urgenda, to bring judicial actions. Openness to such challenges can -depending on circumstances -successfully activate citizens and help them to realise their right to justification.
The judges in the Urgenda case did not simply balance all involved interests. This is the role of the executive and the judiciary. The 83 Besselink, 'De invloed van Europeanisering op de constitutionele verhoudingen in Nederland', supra n. 2, p. 47. judges did not substitute a political decision with their own decision, nor did they undermine the separation of powers in any other way.
What they did is offer effective legal protection by requiring further justification in light of overwhelming science and political commitments and further requiring, in the absence of it, an absolute minimum of action. They did so, soundly-reasoned as required of them in light of an exceptionally dangerous and life-threatening situation. This is the role of the judiciary in a constitutional democracy.
In my opinion, this confirms a working system of separation of powers, rather than undermines it. It strengthens the process of reason-giving that results in a legitimate exercise of public power. ❦