Indeed in most cases reasonable prudence is in fact common prudence;
but strictly it is never its measure … Courts must in the end say what is
required; there are precautions so imperative that even their universal
disregard will not excuse their omission.
Judge Learned Hand in The Tj Hooper, 60 F.2d 737 (2d Cir. 1932)
This invited response commentary considers Benoit Mayer's analysis of the decision of the District Court of The Hague (The Netherlands) of 26 May 2021 in Milieudefensie v. Royal Dutch Shell, published in this issue of Transnational Environmental Law.Footnote 1 The Dutch environmental non-governmental organization (NGO) Milieudefensie,Footnote 2 six other NGOs, and 17,379 individual claimantsFootnote 3 brought this climate case against the oil giant Royal Dutch Shell (Shell). The Hague District Court was the first court worldwide to order a private company to realize a specific greenhouse gas (GHG) emissions reduction target: at least 45% by 2030, compared with the year 2019.Footnote 4 This is a landmark ruling in many respects. Responding to Mayer's case comment allows me to discuss – albeit in a short and thus necessarily limited manner – fundamental questions regarding the normativity of tort law and the legitimacy of qualifying corporate plans to increase GHG emissions as tortious in the 21st century.Footnote 5
While Mayer welcomes the establishment of a corporate duty to mitigate climate change, he criticizes the content of this duty as established by the District Court, and suggests an alternative approach for courts tasked with deciding comparable cases. In his reading, the Court interprets a political and global GHG reduction goal as ‘a matter of scientific necessity’, after which it relies ‘exclusively’ on ‘descending’ reasoning, ordering Shell to align with this goal as well.Footnote 6 Mayer suggests that the Court instead should have engaged in a mix of descending and ascending reasoning, as this combination would have forced the Court to take into account sectoral practices. Though ‘[a]dmittedly, ascending reasoning would not define a … very ambitious [standard]’,Footnote 7 Mayer concludes that this alternative approach would lead to a ‘more consistent’ and ‘more convincing’ legal analysis.Footnote 8
I respectfully disagree with this evaluation on two levels. On a general level, I would like to challenge its take on the normativity of tort law (Section 2). Mayer borrows the terminology of ascending versus descending reasoning from Martti Koskenniemi's 1989 classic From Apology to Utopia.Footnote 9 Accordingly, I explore key ideas from this book in relation to national tort law, in particular the standard of care that was used by the District Court to formulate the injunction against Shell. I suggest that problems identified with legal reasoning in international law are less relevant in the context of national tort law, but that Koskenniemi's book nevertheless provides useful guidelines for legal reasoning in this context too. Legitimacy of legal reasoning lies less in substantive outcomes than in an open and legitimate procedure, which is actualizing in rules of civil procedure in the context of tort law cases.
On a more specific level, regarding the case against Shell, I would like to offer an alternative reading of the Hague District Court's considerations (Section 3). I examine the judgment in the light of Koskenniemi's work on international legal reasoning. I submit that defining corporate climate obligations by referring to sectoral practices is dangerously apologetic and risks hollowing out the duty of care required under tort law, which, in the case of climate litigation, potentially leads to the dystopia that climate scientists have been warning the world about for decades.
2. NORMATIVITY AND CONCRETENESS IN NATIONAL TORT LAW
Koskenniemi's From Apology to Utopia provides a critical assessment of the supposedly objective character of international law. It argues convincingly that the ‘objectivity’ of international law relies on two mutually contradicting concepts: on the one hand, the idea of ‘concreteness’, based on actual state practice, will or interest; on the other, the idea of ‘normativity’, based on certain abstract ideas or rules. Both are visible in, for example, Article 38 of the Statute of the International Court of Justice (ICJ), which defines international customary law as ‘evidence of a general practice accepted as law’.Footnote 10 The idea of concreteness is matched by the inductive argumentative style that Koskenniemi calls ‘ascending’, while the idea of normativity is matched with a ‘descending’ argumentative style.Footnote 11 From the perspective of each of these styles, the other is subjective:
From the ascending perspective, the descending model falls into subjectivism as it cannot demonstrate the content of its aprioristic norms in a reliable manner (i.e. it is vulnerable to the objection of utopianism). From the descending perspective, the ascending model seems subjective as it privileges state will or interest over objectively binding norms (i.e. it is vulnerable to the charge of apologism).Footnote 12
Moreover, in international (customary) law, the problems attached to the utopianism of normativity cannot be fixed by saying that a state has bound itself to international law in general, because that would violate the principle of sovereign equality.Footnote 13 This is essential in the context of international law: the relevant actors – states – are formally equals, which makes it impossible to determine the ‘ought’ independently from their actual practice. At the same time, if the norms of international law are only another way of expressing what states are doing, the law merely functions as an apology for the status quo and loses its regulatory potential.Footnote 14
A similar mechanism seems to be in play with regard to the norm that forms the legal basis of Shell's GHG emissions reduction duty, namely Article 6:162(2) of the Burgerlijk Wetboek (BW) (Dutch Civil Code).Footnote 15 This provision defines what constitutes tortious behaviour, which includes (negligent) behaviour that violates unwritten law regarding what is betamelijk – that is, seemly/becoming/desirable in societal interrelationships.Footnote 16 The central concept in this provision – betamelijk – is laden with normativity. However, what counts as betamelijk should be inferred from ‘unwritten law in societal interrelationships’, suggesting a more ‘concrete’ approach.
Mayer's recommendation to take into account sectoral practices therefore may seem to present a doctrinal question about the interpretation of Article 6:162(2) BW: should what is deemed betamelijk (seemly/becoming/desirable) in the sense of this provision be interpreted by reference to how we think we should behave, or should this standard also be interpreted by reference to how people in society actually behave? This is not an easy question, because the two can overlap: if the majority of people refuse to wear face masks in public transport, this might point to the fact that there is little normative support for wearing masks. That said, they need not necessarily overlap: dog owners might actively look the other way when their furry friend does its business on a sidewalk, even if they believe that, in general, it behoves dog owners to do their part to keep the street clean.
Reading the provision more closely, however, the question of Article 6:162(2) BW is ultimately of a normative, not a sociological, nature.Footnote 17 Nevertheless, this normativity should be inferred from concrete normative views on unwritten laws in society (not concrete behaviour in society). It is not even necessary that the whole of society supports a certain norm – normative acceptance in relevant specialized circles of society can be sufficient.Footnote 18 Hence, in principle, Mayer is right in suggesting that courts in interpreting this provision should engage in some mix of ascending and descending reasoning: in determining what is the norm, they ascend from concrete ideas, which are normative indeed. Thus, judges do not infer the norm from their personal opinion.
The key question then becomes: whose views should be taken into account when interpreting Article 6:162(2) BW? It is here that Mayer's ‘apologetic’ suggestion to take into account sectoral practices falls prey to utopianism itself, because there is an implicit ‘utopian’ procedural norm behind this claim: namely, that sectoral practices are normatively relevant. In other words, whereas Mayer alleges that the District Court's reasoning is too utopian, a reader of his case comment might reproach him for supporting a different utopia in which the sectoral practices of big oil should be the ones to (co-)determine what is betamelijk in societal interrelationships.Footnote 19
Can we ever escape these mind-bending circles of apologism and utopianism? Well, yes. The answer, however, cannot be found in substantive norms or in a static concept of law, but in an ongoing and open procedure through which subjects can exchange views as equals on how to act.Footnote 20 In Koskenniemi's words (referring to Habermas's Legitimation Crisis Footnote 21):
The legitimacy of critical solutions does not lie in the intrinsic character of the solution but in the openness of the process of conversation and evaluation through which it has been chosen and the way it accepts the possibility of revision—in the authenticity of the participants’ will to agree.Footnote 22
Now, in the context of national tort law this compromise is much easier to reach than in the context of international law, thanks to the procedural guarantees to facilitate legitimate collective decision making at the level of a national constitutional democracy. In international law, states are the lawmakers, subjects of the law and its enforcers, and the jurisdiction of the ICJ is relatively limited.Footnote 23 By contrast, parties in national tort proceedings are subject to the democratically legitimized laws enacted by the legislature, applied by the judiciary, and enforced by the executive.Footnote 24 These parties might have played a role in the legislative process by voting or influencing democratic decision making from their position in the public sphere,Footnote 25 but otherwise this role is not comparable with that of states in the realm of international law. Indeed, applying national law against a party in tort law proceedings cannot be seen as a violation of its sovereign equality, unless one rejects the rule of law.Footnote 26
Contract law – especially when a conflict arises in the interpretation of a contract – is closer to international law in this respect; that is, it focuses more on ‘concreteness’ in the sense that it leans on the parties’ intentions. Contractual relations come into existence at the behest of the parties. By contrast, tort law relations can arise regardless of the intention of the parties. The emphasis in tort law is on normativity – its primary goal is to correct relational injustices that were not prevented by other areas of law,Footnote 27 such as criminal and administrative law, or even contract law.Footnote 28 Notwithstanding the academic debate on which notion of justice underlies tort law,Footnote 29 the normative power of tort law can hardly be denied.Footnote 30
Indeed, the emphasis in Article 6:162(2) BW on normativity helps to achieve necessary normative progress. To illustrate this, consider the case law on the usage of asbestos in construction and the liability of employers when their employees developed lung cancer as a result.Footnote 31 It was generally known for a long time that asbestos had detrimental health effects, yet the government failed to draw up legislation regarding its use and the responsibility for its adverse effects. It was therefore necessary, in these liability cases, to resort to the standard of what according to unwritten law was betamelijk (a standard that has been codified in the BW by the democratically elected legislature). If the standard had consisted of a reference to sectoral practices, it would have been impossible to hold employers liable, as at the relevant time the majority of construction companies used asbestos. Betamelijk instead invites a search for concrete indicators of normative consensus and, thus interpreted, relief for the employees could be reached. In short, it is doubtful to what extent sectoral practices are relevant as evidence of normative consensus when interpreting Article 6:162(2) BW.Footnote 32
In this vein, the American Judge Learned Hand famously noted in the Tj Hooper case that an omission may be tortiously negligent even if it is common:
Indeed in most cases reasonable prudence is in fact common prudence; but strictly it is never its measure … Courts must in the end say what is required; there are precautions so imperative that even their universal disregard will not excuse their omission.Footnote 33
In sum, while judicial interpretation of unwritten tort law might require hovering between normativity and concreteness in a way that is similar to international law reasoning, an emphasis on normativity is more legitimate in tort law thanks to the national democratically legitimized constitutional legal order. In national tort law proceedings, this order actualizesFootnote 34 in rules of civil procedure – that is to say, in these proceedings, the open process for reasonable exchange between legal equals is secured by the rules of civil procedure.Footnote 35 In the next section I consider the Court's considerations in the light of these rules.
3. NO ‘MISCHARACTERIZATIONS’ OR ‘MISCONSTRUCTIONS’
The central question in this section is how the District Court determined by how much Shell should reduce its GHG emissions. Why 45% by 2030 compared with 2019 levels? Mayer posits that the Court mischaracterizes a global target as a ‘matter of scientific necessity’ and applied this to Shell, thereby relying ‘exclusively’ on descending reasoning.Footnote 36 In this vein, Mayer alleges that the judgment is ‘plagued by inconsistencies’, mostly arising from ‘four consecutive mischaracterizations and misconstructions’. I address these in Sections 3.1 to 3.4, respectively, and offer an alternative reading: the Court's order to Shell is the result of the debate between the parties to the action and of the rules of evidence, and its reasoning displays a balance between ascending and descending reasoning. I summarize the Court's reasoning in Section 3.5, reflecting on its use of human rights.
3.1. Political rather than Scientific Consensus
As a first alleged ‘mischaracterization’, the Court would see temperature targets as scientific truths rather than as a matter of political agreement. The Court writes:
The goals of the Paris Agreement are derived from the [Intergovernmental Panel on Climate Change] IPCC reports. … [They] represent the best available scientific findings in climate science, which is supported by widespread international consensus. The non-binding goals of the Paris Agreement represent a universally endorsed and accepted standard that protects the common interest of preventing dangerous climate change.Footnote 37
The references to the Paris Agreement,Footnote 38 the ‘widespread international consensus’, and the ‘universally endorsed and accepted standard’ make it clear, in my reading, that the Court understands very well that temperature targets are laid down in legal agreements and, as such, represent the result of a political process informed by a number of factors, including climate science. The Court is looking for concrete signs of normative consensus on these goals.
3.2. Justification of Chosen Reduction Pathway
The second alleged ‘misconstruction’ concerns the Court's choice for an IPCC reduction pathway yielding a 50% chance of limiting global warming to 1.5°C and an 85% chance of limiting it to 2°C. Mayer submits that another, less ambitious pathway with a 66% chance of limiting global warming to 2°C would be an equally plausible interpretation of the Paris Agreement. The latter pathway would translate into a 25% GHG reduction target for the year 2030 rather than a 45% target. However, the Court's choice is far from a misconstruction, for two reasons.
Firstly, the Court is not interpreting the Paris Agreement, which would be out of place in this case between non-state litigants, considering that it is an international agreement with states as its primary addressees. Instead, the Court is finding an answer to what would be seemly/becoming/desirable (betamelijk) behaviour for Shell in societal interrelationships, by interpreting Article 6:162(2) BW. The Court observes that a large consensus exists that the desirable global reduction path should be set at 45% by 2030 compared with 2010 – the European Union (EU) and the Dutch state also use this pathway as their guide.Footnote 39 As the Court emphasizes, this does not directly translate into a legally binding norm,Footnote 40 and global reduction paths do not directly translate into an obligation for Shell as a non-state actor.Footnote 41 Rather, the Court is concretely reconstructing what is the normative consensus.
I understand the order imposed on Shell to reduce by 45% by 2030 to be the result of a question of evidence. Shell was unable to prove why it should deviate from the normative consensus on what is globally required. Surely, if the rest of society would be on its way to reduce significantly by more than 45%, Shell could perhaps continue to emit more, though it would still need to prove that it should, normatively speaking. Yet, sadly, overall emissions are still rising to an alarming extent.Footnote 42
Secondly, I disagree with Mayer's contention that a reduction path with a 66% chance of limiting global warming to 2°C could be an equally plausible interpretation of the (indirect horizontal effect of the) Paris Agreement, which stipulates the aim of keeping global warming ‘well below’ 2°C in Article 2(1)(a). It would be great if we could afford to reduce emissions with only 25% rather than 45% by 2030, but we cannot. Perhaps the Court should have pointed out in more detail what will happen should the planet warm by 2°C.Footnote 43 The ‘inconvenient truth’ is that 2°C of global warming is already devastating, not least for the Netherlands, which lies largely below sea level. Even Shell itself writes in its summons to the Court of Appeal: ‘Everyone agrees’ that a 45% reduction by 2030 is necessary to realize a maximum global warming of 1.5°C.Footnote 44
Today, the climate has already warmed by approximately 1°C, and the resulting heatwaves are causing a considerable number of deaths.Footnote 45 Climate scientists agree that ‘[e]ven without the potential effects of climate change on waves and storm surge, sea-level rise alone is expected to lead to increases in coastal flooding and/or erosion’, including in Western Europe.Footnote 46 Even applying low-emission scenarios, it is predicted that European glaciers will melt by 75 to 88% in volume during the coming century, leading to deadly, devastating floods and droughts.Footnote 47
In short, 2°C of warming, and certainly everything above it, already poses an existential threat to inhabitants of the Netherlands. Against this background, even an 85% chance of limiting global warming to 2°C seems rather risky – indeed, this percentage is not a matter of scientific necessity, but rather a watered-down political translation thereof, which the Court deems relevant in reconstructing normative consensus. A 66% chance to stay below 2°C would not be readily accepted by most people when presented with the facts; few people would want to step onto an aeroplane if they heard that one-third of all flights are crashing.
3.3. Distribution of Mitigation Efforts across Various Sectors
The third issue is the most specifically related to the alignment of the order to Shell with globally required targets. Mayer points out that ‘emissions reductions inevitably unfold differently in various segments of the global economy’.Footnote 48 This remark is entirely correct. That is not to say, however, that the Court order is a ‘misconstruction’, again for reasons of evidence. The claimants have posited and proven that a 45% reduction would be in line with Shell's duty of care – Shell has not been able to dispute this convincingly.Footnote 49 Hence, the Court had to accept 45%, following the rules of Dutch civil procedure.Footnote 50 This means that it is, in principle, possible that a different reduction pathway may be the result of the debate among the parties when the case is decided by the Court of Appeal.Footnote 51 However, it does not mean that the District Court has not properly motivated its decision for 45% in 2030.
It goes without saying that legislatures are better placed than civil courts to determine who should reduce by how much, especially because of the equity considerations also noted by Mayer. The judgment could also be read as a call to the Dutch legislature to (finally) come up with more detailed guidelines for various actors in society at large, so as to ensure that the government's own target is achieved:Footnote 52 to reduce by 49% by 2030 compared with 1990 levels.Footnote 53 Indeed, the judgment is seen as a reaction to the legislature's ‘failure to regulate’.Footnote 54 It goes beyond the scope of this brief response commentary to delve extensively into the legitimacy questions related hereto.Footnote 55 It is nevertheless worth noting that the Court had to make a decision, as refusal to render a decision can even lead to the prosecution of judges under Dutch legislation dating from 1823.Footnote 56 Without question, the Court could have decided to reject the claim, but even Mayer calls the Court's establishment of a corporate duty to mitigate climate change ‘convincing’, as such.Footnote 57
3.4. Baseline Year
The final alleged misconstruction, according to Mayer, is the choice of 2019 rather than 2010 as the base year to calculate by how much Shell should reduce GHG emissions. It is understandable that this choice comes across as surprising, but it is not inconsistent when appreciating the rules of Dutch civil procedure. That is, this was a choice made by the claimants,Footnote 58 and the Court cannot order more than what is requested by the claimants.Footnote 59 At the same time, this choice re-emphasizes that, indeed, the Court does not regard the IPCC reduction path of 45% compared with 2010 as ‘a matter of scientific necessity’, as discussed above in Section 3.1.
3.5. The Role of Human Rights
When interpreting the open norm in Article 6:162(2) BW, Dutch civil courts typically account for all the circumstances of the particular case to establish whether the norm has been violated, which boils down to an attempt to balance the interests involved. The Hague District Court took into account 14 such circumstances, including the UN Guiding Principles on Business and Human Rights (UNGPs)Footnote 60 and international climate law.Footnote 61
Mayer contends that the reference to human rights treaties as one of the 14 circumstances taken into account by the Court is ‘purely ornamental’.Footnote 62 I would disagree. Slightly simplified, the reasoning in the judgment may be summarized as follows: firstly, we know that global warming can violate human rights if it exceeds 1.5 to 2°C,Footnote 63 as had already been established in the Urgenda case;Footnote 64 secondly, as is clear from, inter alia, the UNGPs, there is an international consensus on the duty of corporations to respect human rights – that is, to actively refrain from activities that violate human rights; hence, thirdly, Shell has an obligation to refrain from activities that cause the planet to warm by more than 1.5–2°C. Accordingly, the Court recognized the so-called indirect horizontal effect of human rights in Dutch private law,Footnote 65 a practice also referred to as ‘consistent’ or ‘harmonious interpretation’.Footnote 66 In sum, the reference to human rights and the UNGPs is important for establishing both the existence and the content of Shell's climate obligation.
Am I unequivocally apologetic about the Court's decision? As a matter of fact, I am critical of some aspects of the judgment, including the denial of standing to the NGO ActionAid, which was also problematized by Mayer.Footnote 67 I also agree with Mayer's observation that, given that an English translation of the judgment was published immediately on the website of the Dutch judiciary,Footnote 68 the Court probably anticipated a global audience and could have engaged in more extensive motivation at specific places.Footnote 69
On balance, however, I find the judgment rather convincing. I hope, with this response commentary, to have offered additional insights into the reasoning of the Hague District Court in supporting the injunction against Shell to reduce 45% of its GHG emissions by 2030, compared with 2019 levels. As Article 6:162(2) BW dictates, the Court sought concrete indications of a normative consensus regarding both the existence as well as the substance of a corporate obligation to mitigate climate change. It is for this reason that it included international climate law, human rights treaties, soft law principles in the UNGPs, as well as political assessments of climate science in its considerations. Rather than making non-binding norms ‘binding’ and using these sources as norms,Footnote 70 the Court is evaluating them as concrete indications of normative consensus.
The Court thus used descending as well as ascending reasoning. Relying more heavily on sectoral practices as a strategy to determine Shell's duty of care, as Mayer proposes, would hollow out this duty in a way that is too apologetic for the normativity-laden legal domain of national tort law; a normativity that is, as I have argued, legitimized by the national constitutional democratic process. Indeed, given the warnings we hear from climate scientists, such an apology could lead to a true dystopia.