6.1 Introduction
Arguments from failure not only play a role in courts when it comes to structural reform litigation. Start by considering two examples from my home jurisdiction – Germany.
Germany has a comparatively well-developed public health system. Within this system, private companies provide both private and public healthcare services and medication according to guidelines developed by the Joint Federal Committee (G-BA). The G-BA consists of representatives of public health providers and German hospital associations. Individual doctors can prescribe for patients within the public system treatments and medication that fall within these guidelines, but not those outside of it. However, in a famous line of jurisprudence, the German Federal Social Court has granted individual patients a right to certain treatments or medications outside of these schedules in exceptional situations: specifically, where there were signs of a systemic deficiency or systemic failure on the part of the G-BA.Footnote 1 Such systemic deficiencies are understood to exist, for example, if new treatments, in particular treatments or medications accepted in other states, have not been assessed within a reasonable time for the purposes of adding them to the German catalogue. System failure (Systemversagen) is thus a doctrinal concept in German social law, providing a basis for courts to step in where they would ordinarily defer to other institutions.Footnote 2
Compare this to the famous Soraya judgment of the German Constitutional Court.Footnote 3 The case dealt with a decision by German civil courts to accord personal injury compensation (Schmerzensgeld) to the second wife of the Iranian Shah because the German newspaper Die Welt had published an entirely fictive interview with her. The case was a hard case not because of the facts involved, but because there existed no explicit textual basis for this kind of compensation in German law at the time. The respective courts decided, however, to create a right to such a compensation by adopting a very broad purposive approach to the existing statutory rules. In response, Die Welt filed a constitutional complaint against the judgment, claiming that by granting personal injury compensation the civil courts had overstepped their role as judges and turned into lawmakers, violating the principle of the separation of powers and Die Welt’s constitutional rights. Analyzing the law in question, the German Constitutional Court agreed with the previous courts that there had indeed been a gap in the law because existing rules did not provide for personal injury compensation in such cases. The Constitutional Court then proceeded to ask whether courts should close this gap with their own means (i.e. by building analogies) or whether in a case like this they had to wait for the legislature to act. Most scholars, the Court noted, supported the former. It then went on to make an argument that is worth quoting directly:
The alternative, to wait for the legislature to address the issue, is in light of all circumstances not considered constitutionally required. For while the government attempted twice to find a statutory solution for the problem of the protection of the right to privacy with regard to third parties, its draft bills from 1959 and 1967 were abandoned in the early stages of the drafting process – without this suggesting that the legislature had actually wanted to leave matters as they were. The judge confronted with the issue can therefore not be blamed if he decides that he cannot observe the (merely) formal legal rules in anticipation for an entirely uncertain legislative intervention in the future at the price of a very unjust result in the individual case at hand.Footnote 4
With this argument, the Court tied together ideas of injustice with an assessment of legislative performance, ultimately justifying judicial innovation. Though few German judgments discuss the role of courts as explicitly as the German Constitutional Court in Soraya, assessments of governmental performance are central to any form of public law litigation. Yet, sometimes such assessments of failure become central to justifying particularly innovative judicial action, not merely when it comes to remedies or judicial procedure, but to legal interpretation more broadly. Both two judgments rely on arguments from failure to overcome a previous practice of judicial deference to other institutions, using them as a sort of judicial trump card. But the two cases are also different. In the first type of case, courts step in where an administrative body has failed to act. In the second, we are confronted with a presumed legislative failure, leading to more permanent legal innovations which raise questions about the role of failure in justifying judicial ‘law-making’ more broadly.
In what follows, I begin by considering the relationship between arguments from failure and deference in judicial interpretation and their normative stakes as arguments for interpretive innovation. I then turn to the question if and when courts can justifiably rely on arguments from failure, given that failure in both cases serves as a basis for finding a rights violation in the first place, rather than a tool to justify unconventional remedies by identifying specific kinds of rights violations. This matters because it complicates our twin tasks of treating arguments from failure as a legal rather than political concept and of distinguishing failures from disagreements. Finally, in the last part of this chapter (Section 6.4.), I consider some examples from the field of climate change litigation to explore the connection between newness, innovation and failure in these and other cases.
6.2 Deference and Failure
If we are to understand how arguments from failure operate to override more standard arguments for judicial deference, it makes sense to begin by briefly recalling why and when courts exercise deference in the first place.
6.2.1 Theories of Deference
Courts defer to other institutions for many reasons and in different ways. If previously courts tended to pursue bright-line distinctions between legal and political questions, as famously set out by the US Supreme Court in Baker v. Carr,Footnote 5 more recently political question doctrines have become unfashionable.Footnote 6 Of course, this doesn’t mean that courts no longer exercise deference. Rather, what we see in many courts today is a more nuanced, contextualized – some would say arbitrary – approach to deference where courts do not declare certain questions to be political per se, but instead tailor their scrutiny to the particular issue and context in question.Footnote 7
Existing scholarship distinguishes mainly between epistemic and constitutional reasons for judicial deference.Footnote 8 Epistemic reasons concern the limited capacity of courts to deal with certain questions because they may be polycentric and thus not ideal for judicial resolution in a trial between two parties only, or because they require non-legal expertise to resolve, which judges do not typically possess.Footnote 9 Constitutional deference, in contrast, implies that certain kinds of issues or questions should ideally be reserved for parliaments or other institutions. Thus, foreign policy questions have often been considered a matter of governmental prerogative, and issues that involve significant budgetary commitments in turn have been considered a parliamentary matter.Footnote 10 Some courts will also consider how and to what degree an issue has been extensively deliberated in other institutions as a reason for exercising more, or less, restraint.Footnote 11 Existing scholarship also often treats the question how clear a legal text is and how settled an issue is as a matter of existing legal doctrine as relevant to the degree of scrutiny judges should exercise.Footnote 12 In Baker v. Carr, the US Supreme Court declared as non-justiciable political questions those questions for which there existed no justiciable standards by which to resolve them. Similar qualifiers can be found in many theories of review that are designed to provide a template for dealing with those questions on which there is ‘no clear law’ as with John Hart Ely’s theory of representation reinforcing review, a consideration also taken up in Rosalind Dixon’s responsive theory of review.Footnote 13
6.2.2 Judicial Law-Making
The existence of clear legal standards is, however, not only a relevant criterion for assessing constitutional deference. It lies at the heart of the familiar distinction between law-making and legal interpretation. Though that distinction is – as existing scholarship showsFootnote 14 – not tenable as a distinction between courts and other institutions such as parliaments (because all of them ‘make’ law), courts typically do so differently, and not least in a much more ‘piecemeal and incremental’ fashion.Footnote 15 Accordingly, we tend to think that judges should not usually adopt interpretations that significantly diverge from previous precedents and/or the consensus existing in their respective community of interpreters.Footnote 16
That courts engage in innovation is not a secret, of course. It is to some degree accepted as a part of routine judicial activity, sometimes explicitly so in theories of interpretation, such as living constitutionalism or dynamic interpretation. On this view, courts may modify and adapt existing arguments to fit new challenges. Over time therefore, the law evolves. But sometimes courts introduce more radical changes. Such more abrupt and obvious changes typically pose challenges for courts since law reform on a more substantial scale is generally considered a matter for legislatures rather than judges. Major innovations are therefore almost certain to trigger critical questions. They may be seen to compromise the court’s authority as the reliable and final interpreter of law, endangering the principles of legal certainty and fairness as consistency. It is thus important for judges to marshal good arguments to justify more dramatic changes in their interpretive practices.
This is where institutional failure arguments often come in – as a tool to justify intervention when there are otherwise good epistemic or constitutional reasons to exercise restraint. Arguments from failure may provide a trump card in such cases, setting aside arguments for non-intervention. Sometimes, they help judges overcome deference in light of administrative malfunctioning; in other cases, such as Soraya, they may also furnish an argument for courts to get involved in the context of legislative inertia.
As I have argued previously, however, it matters which institutions are involved. Where courts step in to address legislative failure, they take prospective decisions not confined to individual cases. They are thus stepping much further outside the traditional judicial role than if they substitute for an administrative failure in a concrete case. At the same time, it is important to recall that the difference compared to the more ordinary case of incremental adaptation and updating the law is one of degrees; unsurprisingly therefore, the line between the two will often be blurry, as we will see later in the climate change decisions.
6.3 Evaluating Arguments from Failure
Can we justify judicial innovation in these cases? Overall, I argue here that the framework developed in this book still applies to such cases. The broader legal and political context and the existing regulatory framework, as well as the question what other institutions might step in, remain relevant here. The same is largely true with regard to considerations of functionality and cost, as I argue below. Yet, there are also some new problems, because we often can no longer presuppose a rights violation in such cases. Rather, the presumed failure is a necessary condition for identifying a rights violation in the first place and for assessing the necessity of intervening.
6.3.1 Functionality and Costs
Start with the more familiar questions about the functionality and costs of judicial intervention. Especially where arguments from failure are employed to overcome epistemic deference, functionality can become a problem. Given the reasons for epistemic deference in the first place, it may seem doubtful whether courts will have a realistic chance to address the problem, for example because the problem will be polycentric or because they lack the necessary expertise to address the issue.Footnote 17 Ultimately, their chances of success will depend both on existing procedural rules and judicial willingness to adjust and develop those rules to deal with cases of institutional failure. If the problem consists of a lack of expertise, courts can bring in external experts, but the rules for doing so matter. Thus, in some systems, like the United States, it is usually the parties who bring in their own experts to make the case for their own side. In other systems, such as Germany, the process is judge-managed rather than party-driven. Courts have a list of external experts they might draw on for specific issues, who are not associated with either of the parties.Footnote 18 The challenge of polycentricity may be harder to overcome, but here too we have seen in previous chapters how courts transform judicial trials in some cases and bring in other stakeholders rather than relying only on the parties involved in the original case.
Costs to consider may be short-term or longer-term, with the latter harder to assess. Short-term costs may involve political backlash, but also the risk that courts may get things wrong and make the situation worse rather than better. This goes back to the problems of functionality discussed above: courts may not have been able to make up for their own lack of expertise in other ways and therefore intervene in a manner that ends up being counterproductive to the aims pursued. There may also be shorter- or longer-term consequences for related issues and persons who are not parties to the case at hand if the case involves polycentric questions. For example, courts deciding on what medications and treatments will be funded may – if this happens sufficiently frequently – decrease the share of the health budget available to others, in particular those with less ability to access courts. This is not only a theoretical possibility as studies on Brazilian health care litigation show, where this is precisely what happened.Footnote 19 Other longer-term consequences may be that judicial intervention may actually discourage responsible politics. In such cases, we may encounter what Mark Tushnet calls ‘democratic debilitation’ – which may occur because legislators know that the court will take care of things.Footnote 20
6.3.2 Necessity and Breach of Law
If functionality and costs are familiar concerns, things become more complicated when it comes to assessing the necessity of judicial interventions. As I argued in Part I, if we want to minimize the risk of failure arguments becoming politically instrumentalized and abused, we need to identify some breach of law as a threshold requirement. Yet, this can no longer be taken for granted here. A breach of law, or violation of rights, is no longer something we can presuppose, but rather the thing we are trying to establish in the first place. Recall our two examples: Failure was a prerequisite in the healthcare case to give plaintiffs an enforceable right to a specific medication or treatment, and in the Soraya judgment, the (presumed) legislative failure provided an argument for granting individual plaintiffs new unwritten rights to compensation, which they did not have before. Can we still therefore rely on failure arguments in this context?
6.3.2.1 Administrative Failure
Begin with our German example from above. Why and when should we assume that there is a case of Systemversagen (system failure)? The Joint Federal Committee has not usually violated any specific laws in cases of Systemversagen. The only thing we can say is that it has not acted in accordance with its function or role, as expressed in its statutory mandate. However, as lawyers know, such general mandates often are very vague. Nor are there typically hard benchmarks as to what constitutes adequate performance, though there are exceptions. Reasonable people will therefore often disagree as to what constitutes adequate institutional performance. How then do German courts reason in such cases? Broadly speaking, they ask if the institutions responsible for amending these schedules have failed to initiate the relevant procedure assessing the need for inclusion, either completely or in a timely manner, or have failed to conduct them in accordance with existing rules.Footnote 21 Overall, German courts have been fairly cautious in assuming failure. The mere registration of a treatment in other countries, for example, does not automatically imply that not including the same treatment in German catalogues constitutes a failure, according to the German Federal Social Court.Footnote 22 Rather, the courts will assess the internal procedures in detail, taking into account the different opinions of medical experts, the time needed to conduct an analysis and other actions of the different participants involved in the process.Footnote 23 Only where these different factors taken together suggest that there were no good reasons for non-inclusion in the case at hand are we confronted with Systemversagen.
This is a start, but it is not a comprehensive answer. The easiest scenario is clearly one where we can identify some breach of law: for example, when the institution in question has not properly followed procedural rules and/or its institutional mandate. Not every violation of procedural rules, however, should be considered a failure that warrants judicial intervention where deference is otherwise more appropriate. The rules in question may sometimes only be in the organization’s own rules or in administrative guidelines, instead of being a matter of statutory rules. In other cases, there may be no written standards at all, but merely a set of ‘best practices’ usually followed by the institution in question. Given the risks of abuse, it seems clear that the less we have clear statutory rules, and the more courts need to rely on broad institutional mandates or best practices, the more they should exercise caution when it comes to diagnosing institutional failure. But this still leaves many questions open.
6.3.2.2 Legislative Failure
Difficulties are compounded where courts confront straightforward legislative failure. Unlike administrative agencies, legislatures have no specific mandate in the sense of being charged with one particular task. There are of course constitutional constraints and duties and in some systems directive principles, but such provisions nevertheless typically leave legislatures large discretion, as we should expect it in a democracy. Where legislatures do not act on a particular issue therefore, this is not necessarily problematic, let alone constitutes a failure. Legislators may have other priorities or indeed believe that there is no need to do something.
There are only two ways in which we might be able to get a handle on the idea of legislative failure. One is substantive, as a matter of the legislature failing to protect and guarantee existing (quasi-)constitutional or human rights. The second is procedural, that is where the processes in question may in some ways be considered deficient.
6.3.2.2.1 Procedural Approaches
Procedural standards are partly tied to specific substantive human or constitutional rights, but they can also represent independent procedural rights. Such rules are sometimes part not of constitutional or human rights law but of administrative law. This includes, for example, principles such as transparency and participation requirements, reason-giving and the possibility of review.Footnote 24 Such violations of procedure in turn might indicate that we are dealing with a case of capture of the political process by special interestsFootnote 25 or indeed legislative inertia triggered by coalition-building problems or legislative blindspots.Footnote 26 Yet, absent a specific constitutional duty to act such cases will typically be hard for judges to identify.
Moreover, judges will need to be very careful to distinguish between cases of failure and cases of reasonable disagreement, in particular where the situation is one of legislative inaction. Because the failure in question is not tied to established rights violations, this is a harder challenge than when it comes to structural reform litigation. Several factors are relevant in this regard. First, we cannot assume failure where inaction is a deliberate choice reached after a full and inclusive debate on the issue. Deliberation particularly matters in cases where the failure in question appears as a failure to follow through on previous commitments. Again, however, we need to be careful to distinguish failure from the deliberate change of a political position. Not every inconsistency will indicate governmental failure, but it may express simply a legitimate change of opinion. To determine if this is the case, courts will have to know more about the reason for the shifting positions – and potentially look to governmental records as well as procedures to find out more. Such an analysis, however, should not be undertaken lightly. As commentators of this kind of procedural scrutiny have pointed out, there is a risk of treating legislatures and governments like administrative offices.Footnote 27 Doing so may ignore the nature of democratic decision-making which will often be subject to the need for political compromises and thus cannot follow the standards of consistency we expect of administrative actions. In a multiparty democracy, and in particular when governments are formed as coalitions between several parties, political deals and compromises may be necessary to enable effective government. In scrutinizing consistency, courts must thus be sure not to adopt a purely administrative perspective and be wary of holding democratic processes to unrealistic ideals.Footnote 28
6.3.2.2.2 Substantive Standards
In theory, things are easier if the legislature’s inaction violates certain constitutional duties or positive rights. For we do, of course, have a range of criteria to assess governmental performance when it comes to adjudicating violations of positive rights, dependent on case and jurisdiction. The ECtHR, for example, requires states to fulfil their protective duties to provide an adequate regulatory framework to prevent rights violations and to establish corresponding institutional rules and safeguards and measures to implement the said framework, as well as procedures to investigate and sanction violations ex post.Footnote 29 The German Constitutional Court has emphasized in its jurisprudence the need for consistency within regulatory frameworks.Footnote 30 Perhaps the most interesting and internationally influential approach has been that of the South African Constitutional Court. Rather than following the International Committee on Economic, Social and Cultural Rights’ suggested minimum core approach,Footnote 31 the South African Constitutional Court in the Grootboom case famously adopted a reasonableness standard for the right to housing under the South African Constitution.Footnote 32 Assessed in those terms, the judges considered the failure to develop a policy dealing with short-term housing needs and thus provide ‘relief for those in desperate need’ to be unreasonable – and thus to violate the South African Constitution. Thus, reasonableness essentially provides a measure for what we may expect of the government in a particular context, which is why the Constitutional Court has used it generally to assess positive rights obligations.Footnote 33 And so on.
These standards are, however, part of the ordinary arsenal of constitutional and human rights review – and this begs the question what, if any, role they might play in our context here as a tool to justify judicial innovation.Footnote 34 The answer to that question must be that failure only really comes into the picture in these cases if we are dealing with a rights claim that involves significant newness – and thus where the court is going beyond established precedent, in a way that requires special justification because it involves more than incremental legal change. If so, however, the standards we find in the broader jurisprudence for assessing violations of positive rights are the best yardstick for assessing the existence of failure in other contexts, where judicial intervention is not usually accepted.
6.3.2.2.3 Soraya Revisited
Let’s return for a moment to the Soraya case. The German Constitutional Court in that case provided no real reasoning or justification as to why we might be dealing with a case of failure. To start with, the Court did not argue in terms of a constitutional obligation to protect constitutional rights, such as a right to privacy in that case. We merely learnt that several attempts had been made to change the law in question and that they had faltered, but it was not clear that legislators had actually changed their minds on the issue. The only way of treating this as a failure would presumably have been to show that the legislators’ inaction in this case was not due to disagreement or something like that, but instead indicative of dysfunctionalities within the democratic process. Again, however, this was not clear from the German Court’s reasoning. In order to know therefore whether we were indeed dealing with a case of failure, we would simply need to know a lot more. Without such information, the argument for judicial intervention in this case of legislative inertia remains unconvincing.
6.4 Climate Change Litigation
To better understand the way arguments from failure work and should work in the context of courts developing the law in bold ways, consider recent litigation on climate change. Climate change litigation is of course a broad field and no longer an entirely new phenomenon. Nevertheless, it remains contested if and to what extent the struggle to respond to and contain climate change should involve courts.
Though courts can certainly assess the state of the scientific consensus, this in itself is generally not sufficient to create legal obligations to act. Responding to the threat climate change poses requires policy changes that come with trade-offs and costs, in particular with regard to economic development. Balancing the competing imperatives and interests requires making choices that democratically elected leaders supported by institutions with significant expertise in the area are generally better suited to make than judges.
Nevertheless, in recent years climate change has increasingly been framed as a human or constitutional rights issue. From a legal perspective, such approaches face a number of challenges. To start with, climate change emission targets are typically not laid down in a way that grants individuals justiciable rights to hold governments to those targets. To frame the issue in terms of human rights hence involves not just judicial creativity, using existing rights as a basis for creating new obligations in changing circumstances, but also poses several legal and strategic challenges. It is often difficult to establish causal links between specific acts or omissions and harm to individual rights, in particular where such harm is only likely to materialize much later.Footnote 35 These problems are exacerbated by the fact that we are dealing with a situation where the resulting harm may often be transboundary, thus raising questions as to the extra-territorial effect of human rights obligations.Footnote 36 Besides these more straightforward legal challenges, there are functional ones that may also be framed in terms of the separation of powers and the judicial role. Climate change represents a paradigmatically complex and polycentric problem, requiring specific scientific and other expertise to address it. All of this makes courts less than ideal fora and opens them up to broader criticisms of judicial overreach if they decide to intervene.Footnote 37
Nevertheless, courts have increasingly begun to engage climate change in their jurisprudence. Much of this jurisprudence has been framed – by judges themselves or by scholars writing on climate change – as a response to the failure of other institutions to do enough to tackle the issue. Early litigation in particular has often been understood as an attempt to fill a ‘governance gap’, resulting from the unwillingness of governments to agree on binding international standards or put forward credible domestic legislation.Footnote 38 Yet, existing accounts have shown that the governance gap explanation is not entirely accurate as a descriptive matter. We frequently see more litigation and judicial action in those countries where significant legislation already exists, as opposed to those where it is largely absent.Footnote 39 Other commentators have defended judicial action in these cases more broadly on the basis of an Elyian argument for the need to protect future generations not represented in the political process. It thus responds to a standard deficit, or indeed ‘failure’, of democratic processes.Footnote 40 But there are also strategic reasons for drawing on the language of failure in these cases. Given the lack of legally binding emission reduction rates in the Paris Agreements, which some have argued sets up governments for failure, ‘[o]ther than public shaming, [litigation is] the only way to hold nations accountable’.Footnote 41 Thus, courts have increasingly started to hold governments to their public commitments and sought to secure compliance with international agreements, such as the Paris Agreement, with its deliberate integration of legally binding and non-legally binding elements.Footnote 42
Governmental failure has certainly been one important driver for judicial intervention, including structural remedies. An example is the now famous decision of a Pakistani court in response to a public interest petition in Leghari v. Federation of Pakistan. The judgment framed climate change justice as a matter of constitutional rights, putting forward a range of expansive orders which included the establishment of new bodies to tackle the issue: ‘In the present case, the delay and lethargy of the State in implementing the [previously developed] Framework offends the fundamental rights of the citizens which need to be safeguarded’, the Court argued.Footnote 43 Though in Leghari, governmental failures are not used to justify the existence of a rights violation in the first place – because the Court rather skates over the legal issues at stake – they play a role in justifying the need for more robust judicial intervention.Footnote 44
The Dutch Urgenda decisions offer a more standard example of the role of failure arguments in climate change litigation, and it is to them that I now turn. As we will see, some of the analysis conducted by Dutch courts essentially represents an assessment of the performance of the Dutch government, which overlaps with an argument from failure, albeit not being explicitly cast in those terms.
6.4.1 The Urgenda Decisions
Urgenda is a Dutch environmental NGO. It decided to bring a case against the Dutch government with the aim of legally obliging it to reduce, by the end of 2020, greenhouse gas emissions by at least 25% as compared to the levels in 1990. A Dutch District Court, a Court of Appeals and, finally, the Dutch Supreme Court (the Hoge Raad) agreed, with the Court of Appeals and the Supreme Court relying on a human rights framework, in particular Arts. 2 (right to life) and 8 (right to privacy, family life) of the European Convention on Human Rights, with some references by the Supreme Court also to Art. 13 (right to remedies).Footnote 45 While acknowledging the need to combat climate change, the Dutch government opposed the decisions. In the Supreme Court, state representatives argued in particular that the earlier decision of the Court of Appeals should be overturned because of the absence of legally binding international obligations (3.4.) and given that states enjoy a margin of appreciation in fulfilling their positive obligations under the ECHR (3.2.). Moreover, courts – whose role in the Netherlands has traditionally been restrained – were not allowed under Dutch law to issue orders to create legislation, as they had essentially done here (3.5).
What are the key steps in the Dutch Supreme Court’s reasoning? I leave aside some of the more technical details, such as the question of Urgenda’s standing, and focus instead on the arguments with respect to Arts. 2 and 8 ECHR. First, the Dutch Supreme Court confirmed that Arts. 2 and 8 ECHR generally applied to climate change and obliged the Netherlands to do its part to counter the dangers resulting from it (para. 5.2.). It seems hard to fault this reasoning in general. The real challenge came in the next step as the Court asked what followed from this conclusion, and in particular what concrete obligations the Netherlands had. It noted that as a matter of international law, there were no binding obligations to reduce emissions to a certain level. The Dutch Supreme Court first gestured to the role of the Dutch government and legislature in deciding how to fulfil its obligations and concluded that ‘[i]t is clear, for example […] that the State cannot at any rate do nothing at all and that the courts can rule that the State is in breach of its obligation […] if it does nothing’ (para. 6.3.). The Court next set out to determine what the ‘minimum obligations’ of the Dutch government were with respect to climate change. The judges first pointed to the international developments with regard to combating climate change, drawing in particular on international declarations and soft law, arguing that there was indeed a consensus that emissions needed to be reduced by 25–40% by 2020. This – perhaps the key step in the Court’s argument – also applied to the Netherlands, notwithstanding the fact that the existing reduction rate of 20% complied with the Netherlands’ obligations under EU law. According to the Court, that did not mean that the Netherlands had no further-reaching obligations under human rights law, however. Not only had the Netherlands long been on the list of countries with the highest emissions per capita, but several other EU Member States had also adopted higher targets than the 20% reduction rate, and the European Union overall envisaged cuts of 26–27% by 2020. Finally, the Court added that a higher rate also corresponded to the previous policy of the Dutch government itself, which had originally planned for a reduction of 30% by 2020 (para 7.4.1.). It only deviated from this policy later, but without giving any reasons as to why it did so, and in spite of the fact that it was uncontested that emission reductions would become more costly the longer they were delayed (para 7.4.6.).
Unsurprisingly, the Urgenda decisions prompted an extensive debate in the legal community, about deriving concrete obligations in terms of percentages for emission reductions from the European Convention, and about what the role of courts with regard to climate change should more generally be.Footnote 46 Some Dutch critics have argued that the Urgenda line of reasoning might justify expansive judicial interventions in a wide range of other policy areas, which would be contrary to both existing Dutch law and the separation of powers. A standard comparison seems to be making the wearing of bike helmets obligatory.Footnote 47
The key question, of course, is if the critics are right. The answer is both yes and no. It is yes insofar as the Supreme Court’s analysis is worrying with regard to the separation of powers in a scenario where the Court uses soft law standards to give content to binding human rights obligations. Given that states routinely choose to rely on soft law when they want to avoid creating legally binding obligations, this move is problematic at least from the perspective of a more traditional consent-based understanding of international law. The Dutch Supreme Court’s analysis skates over this problem very quickly and without sufficiently engaging with the question whether we can use human rights as a backdoor to hold states to commitments that were not meant to be legally binding in the first place.
The answer is no, insofar as there are other more convincing arguments in the Urgenda decisions, both in the Appeals Court and in the Supreme Court, which frame the question in terms of the comparative performance of the Dutch government – and, I would argue, at least implicitly in terms of institutional failure.
First, in analyzing the content of its obligation with regard to reducing emissions, the Supreme Court pointed out that the Netherlands had long been on the list of high per capita emitters. The details are, however, only spelled out in the decision of the Court of Appeals, which noted that the Netherlands at the time ranked ‘34th of 208 countries. Of the 33 countries with even higher emissions, only nine have a higher per capita emission, and not a single one is an EU Member State.’Footnote 48 More particularly, the Court subsequently argued, the Netherlands’ reduction efforts were lagging far behind states such as Germany, the United Kingdom, Denmark, Sweden and France (para. 56). Second, and more importantly still, both courts emphasized the lack of convincing reasons for the lowering of the previous 30% reduction rate to 20%. The Court of Appeals noted that
‘a substantiation based on climate science was never given, while it is an established fact that postponing (higher) interim reductions will cause continued emissions of CO2, which in turn contributes to further global warming. More specifically, the State failed to give reasons why a reduction of only 20% by 2020 (at the EU level) should currently be regarded as credible, for instance by presenting a scenario which proves how – in concert with the efforts of other countries – the currently proposed postponed reduction could still lead to achieving the 2º C target.’
In the proceedings itself, the government subsequently argued that it preferred to adopt higher reduction rates after 2020, but the Supreme Court dismissed that argument. Given that it could not be ascertained that a delay would achieve the same effects, it emphasized again the lack of convincing reasons for adopting only a 20% target:
‘The 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’.Footnote 49
These arguments provide a comparative institutional perspective. First, they compare Dutch efforts with those of other similar countries, and find them insufficient. Second, they assess the institutional performance through the lens of procedure, here the requirement of reason-giving familiar from administrative law. This kind of perspective is also precisely what we would expect when we are addressing the question whether to intervene in response to legislative failure. The comparison with other states and indeed with the government’s own previous policy provides a particularly important – and I believe necessary – counterweight in this case to the move from soft-law standards and scientific expertise to legal obligations, which some have viewed as the weak point in the Court’s reasoning.Footnote 50 By drawing on a procedural argument – here the lack of reasons for switching from a 30% to a 20% reduction target in 2020 – the Supreme Court and Appeals Court recognize the need to accord the government discretion with regard to the details of its environmental policy. However, this discretion is not unlimited. By insisting that the government provide convincing reasons for abandoning its previous policy, courts exercise some scrutiny.
6.4.2 Comparison and Analysis
Yet, not every court arguing in terms of governmental performance or even applying a standard of consistency is putting forward an argument from failure. This is true even when it comes to an issue like climate change, though it can sometimes be hard to identify the core of the argument.
Compare Urgenda to the German Constitutional Court’s climate change decision as a case in point. In its decision, the German court argued that the Basic Law’s Art. 20a on the duty to protect the environment required the government to lower emissions, without, however, granting individuals a right to enforce this provision (Art. 20a) directly. At the same time, the Court argued that since the German government had chosen to pursue the goal of limiting temperature increases to well below 2°C and preferably to 1.5°C above pre-industrial levels in line with Art. 20a, it had to do so in a proportional manner, requiring that the ‘reduction burdens are not unevenly distributed over time and between generations to the detriment of the future’.Footnote 51 Setting out clear emission reduction rates only up to 2030 but not beyond, however, did not ensure a fair distribution of such burdens. It placed an excessive share of the burden on younger people. Invoking ideas of intergenerational consistency and ultimately fairness, the Court argued that this violated the plaintiffs’ constitutional rights.
The difference between these two cases is instructive of what it means to present an argument from failure to justify judicial intervention. At first glance, the reasoning is quite similar in both cases. Like the Dutch courts, the German Constitutional Court seeks to measure governmental performance. Its focus in doing so is different, however. Rather than focusing on what the government has done wrong – as the Dutch court does – the German Constitutional Court places the emphasis on elaborating what the government’s duties are in this case, in order to then conclude that it did not live up to them. Though this difference may appear insignificant at first, it matters in practice. The Dutch court’s approach is ultimately more deferential, essentially grounding its analysis in an institutional comparison of the Netherlands with other states, thus taking state practice more clearly into account rather than developing its own standard for good performance. The German Constitutional Court, by contrast, focuses on developing the government’s duties in constitutional terms and ultimately in terms of independently developed standards of good governance, albeit standards derived inter alia from existing German legislation. As a result, there is no argument from failure here in a meaningful sense. The German decision needs to be evaluated on its own terms, as a matter of its doctrinal persuasiveness and without recourse to governmental failure.
If we compare the three cases mentioned – Leghari, Urgenda and the German case – we see that they are on a continuum when it comes to evaluating performance and diagnosing failure. In Leghari, the court diagnoses a complete failure to deal with the issue and accordingly intervenes in robust ways, albeit that the combination of developing the law and adopting structural remedies is hard to justify. In Urgenda, the court finds a comparative failure to take the issue sufficiently seriously, drawing on state practice elsewhere and the government’s insufficiently explained change of policy. In the German case, the court essentially finds the government’s policy insufficient for distributing higher burdens on the younger generation, but there is no longer a diagnosis of failure. This is a more conventionally ‘legal’ analysis. This is not surprising, insofar as the development of rights and aspects of rights has long been a feature of the German Constitutional Court’s jurisprudence, in contrast to the Netherlands, where only review with regard to the compatibility with international law is allowed and usually practised relatively cautiously.Footnote 52 Accordingly, the need for a failure argument was more pressing in this context.Footnote 53 Framing an issue in terms of failure is thus also an indication as to how revolutionary courts believe themselves to be, and a rhetorical device to overcome resistance to judicial involvement.
As more and more courts have begun to treat climate change as a fundamental or human rights issue, we have also begun to consider it more normal for courts to deal with the issue. As Fowkes points out, this is because the essential question for judges to decide whether to intervene or not depends not so much on the polycentricity or complexity of an issue but its newness: Courts, he argues, tackle plenty of highly complex questions, but they typically do so in a context where there already exist institutional or legislative frameworks that they may seek to improve or other actors who expect judicial interventions and can adjust to them.Footnote 54 Thus, being less of a new issue, there will also necessarily be more space for judicial intervention in climate change cases. This may appear ironic, given that governments now are mostly doing something with regard to the issue, and thus, judicial intervention may seem less necessary than it was a few years ago. Yet, it speaks to the role of courts as institutions making law incrementally only. It also means that we are less likely to encounter arguments from failure in the future, as the barriers for judicial intervention have – with good reasons – been lowered. On a normative level, therefore, the role of failure arguments in climate change litigation not only depends on the existing legal resources and the newness of the issue but also on other existing reasons for intervention and the judicial role within a given system. The better our general reasons for intervention, the less need there is for an argument from failure.
6.5 Conclusion
Failure drives innovation – and as this chapter shows, not just in the realm of structural remedies but when it comes to legal interpretation more broadly. Courts may decide to intervene in cases of administrative failure and sometimes even in cases of legislative failure in cases where they would ordinarily exercise restraint and defer to other institutions. And while judicial law-making is in many ways a routine and standard matter, sometimes, courts make law in bolder and less incremental ways than we would expect. If and when they do so, arguments from failure provide a way of justifying their interventions.
Open access