Two Avenues for Loss and Damage

Anirudh Sridhar*

Introduction

In 2022, the Conference of the Parties (COP) to the UN Framework Convention on Climate Change (UNFCCC) announced a Loss and Damage (L&D) Fund. The Fund would be a mechanism to pool financial pledges together to assist vulnerable countries in dealing with the effects of climate change, which has been estimated to climb to USD 1.7-3.1 trillion per year by 2050. The idea that L&D should be a separate pillar of climate action (along with mitigation and adaptation) has animated least developed countries (LDCs) since the early 1990s. Their economies have and do contribute very little to the GHG emissions tally, and adaptation is but protection from further avoidable harm. Climate change is already devastating economies, cultures, and lives, and there needs to be some way to pay for these existing losses – or so their argument went. For long, the issue was politically toxic because the tenor of L&D, from its tort law origins, is accusatory, and developed countries did not wish to accept blame, even if they were ready to accept a higher share of responsibility in the overall climate fight. But pressure to address L&D mounted throughout the last decade as the G77 countries plus China banded together on this issue in a way that they haven’t on any others. Finally, at least as a means of disbanding this powerful negotiating bloc, an L&D Fund was agreed to in 2022 and set in place in 2023.

But the wording of the Fund is vague; it seems designed to dilute the notion that there are perpetrators and victims in climate change. And while such urbane restraint may have been diplomatically necessary, it also leaves out the crucial question of who will pay for the Fund. This seemingly strategic ambiguity in effect restores matters to the status quo, where developed countries can contribute to L&D, if they wish, from their magnanimity and spirit of alms. All that the Fund seems to have achieved is a psychological satisfaction among developing country activists and diplomats who can now tote the Fund as a trophy, and thereby a deflation of the animus binding the G77+China to the cause of L&D. To add to the drama, the World Bank has been elected to manage whatever charity trickles in in the next five years; the Bank’s record of deeds in the last century undoubtedly speaks to the intent behind the 2022 compromise. All told, the Fund has thus far proved a masterclass in postcolonial diplomacy.

In this essay, we will notice an alternate model of L&D to the Fund, compare their ‘philosophical’ grounding, and asses their viability. We will exhume the almost entombed question of liability and compensation and see if there are ways of reintroducing a faux-legal process to international L&D. Although the legal process has clearer ways of attributing culpability and damages than a Fund, we will see that the former is vulnerable to the circus of competing scientific models before a judge – we will suggest ways of avoiding this.

Liability And Compensation 

The question of compensation (as opposed to relief) for L&D is often ignored due to the belief that the accompanying decision of the Paris Agreement (PA) has foreclosed the possibility. The PA introduced a separate article (#8) dedicated to L&D, but in exchange, the appended decision said, ‘Article 8 of the Agreement does not involve or provide a basis for any liability and compensation’. If we read carefully, however, the text only states that Article 8 of the PA is not the basis for liability: there could be other bases. A sceptic may argue here that even if there are possibilities to introduce liability de jure, developed countries would never accept such a measure de facto. If this is so, we need not limit L&D to the UN paradigm – there are other international fora to explore. The International Court of Justice (ICJ) and the International Tribunal for the Law of the Sea are currently contemplating whether countries can bring L&D claims for climate change. The Small Island Developing States (SIDS) have recently brought a case on states’ legal duties on climate change before the ICJ. The increasing engagement of the Global South with such international judicial fora could prompt a turn to them also in the case of L&D. The question before us then is which mode, liability or the L&D Fund, agrees with the larger aims of L&D.

L&D Justice

L&D was introduced to the climate change discussion in 1991 by the Association of Small Island States (AOSIS) as a means of seeking climate justice. Their islands were drowning; it was not their fault, and they did not have the means to save themselves. They therefore asked for corrective justice from the parties that had most contributed to climate change: you caused our losses, so you must pay for them. But the means they suggested to realise this payment confused the issue. An L&D insurance fund was the most popular idea at the time: as and when there are disasters, affected countries would be compensated through the L&D insurance. Given developed countries were supposed to pay into the insurance, the theory was that the insurance would transfer, and thereby distribute, some of the risk of damages from the providing party to the provided: we may regard this as distributive justice. But there were many practical and theoretical problems with the idea. Many of the effects of climate change are gradual and insidious, like the salination of arable soil, say, and not blockbuster events like hurricanes. An insurance model is not well suited to losses that are not immediately quantifiable from isolable events. But more deeply considered, the concept of insurance is associated with risk, to events that could happen but have not yet; whereas the conceptual thrust of L&D is to demonstrate that nations are not just vulnerable but already suffering from the extant effects of climate change. Grounding L&D in insurance completely undermines that crucial narrative. Although the insurance idea never quite took off, the theory of distributive justice undergirding it has remained and taken new form in the L&D Fund.

The advantage of corrective justice is that it leaves no room for ambiguity on the crucial question of who pays whom. As in a typical lawsuit, a wrong must be established and compensation quantified according to the degree of negligence and injury – the process must establish a victim, damage, and the responsible party. In climate change, the victims would be those facing the impacts (fires, sea level rise) of climate change, the damage would be quantified from economic (crop yield, infrastructure damage) and non-economic (lives, cultures lost) losses, and the major emitters of GHGs would be the responsible parties. In other words, answering the question of who pays whom and how much, a problem that has baffled the L&D Fund, is the precise function of the liability and compensation paradigm.

Despite repeated recommendations during the Transitional Committee (tasked with fleshing out the L&D Fund) meetings, the COP decisions year after year have failed to identify developed countries as the or even the majority contributor to the L&D Fund. To save face, the Fund has ventured private companies as potential sources and floated a vague notion of cooperation between developed and developing countries in raising funds. Does this mean that developing countries will have to share in contributing to the Fund? Such ambiguity begins to erode the notion that there are perpetrators and victims – a central axiom of L&D – in climate change. And what incentives do private companies have for paying for damages in destitute countries? What is more likely is that these private contributions will be structured as loans, and the managing World Bank will revert to its historical reflex of driving poor countries into deeper wells of debt.

Models of distributive justice arise often as a compromise when the parties likely to lose from a legal process are also the more powerful. But distributive models could also simply be a consequence of fatigue in negotiations. It remains for developing country actors to resist the diplomatic and rhetorical appeal of the L&D Fund and continue seeking corrective justice elsewhere. This seems the only way to preserve L&D from becoming a tool for debt entrapment, or worse, charity.

Problems With Liability And Compensation

The mechanics of establishing liability in the case of climate change, however, are fraught with logical inconsistencies. For a victim to be owed compensation, they must have a reasonable right to expect protection. While citizens can argue (however tenuously) for such protection within the confines of sovereign states, states do not have duties to others in the same way, especially for activities that they didn’t know would be harmful at the time of carrying them out. But even if such a right was discovered from the moral architecture of the post-war world order, there is the more daunting task of establishing a causal connection between the actions of one state centuries ago and contemporary events in another on the other side of the world. And if at all it was possible to establish the connection through causal modelling and specific developed countries can be held responsible for all climate disasters, it does not follow that they can be held liable.

If we transfer the criteria of tort law to the international arena, in order for a court to find liability, the potential harm must be foreseeable, and there needs to be some degree of proximity between the defendant and claimant. Mostly, it is enough to appeal to common sense to establish foreseeability. For instance, it can be assumed that dumping toxic chemical waste into a city’s water supply will lead to disease and death – the obviousness of causality here is what allows it to become a villain’s plot in children’s cartoons. In more diffuse environmental processes, however, causal connection is typically established by science. But the harms cannot be deemed foreseeable as soon as a relevant academic paper is published. For a harm to be foreseeable, there must be some notification of potential consequences that is public enough that the actor in question can be reasonably expected to factor it into their actions. And in the case of climate change, the relevant science was established gradually from the 1960s to the 70s, and it was only towards the end of the 1980s and early 90s that states began to recognise the predictions of climate science. So, while industrialised countries are responsible for much of the emitting activity that in the last two centuries has contributed to climate change, they cannot be held liable for all of it.

The second criterion of proximity is again rather tricky to establish in the case of climate change. In more tractable cases like acid rain, action and consequence are not separated by too many degrees and actors. A claimant can argue that a nearby industrial plant’s release of sulphur dioxide into the atmosphere caused water vapour to acidify and that the damage to health and property from the resulting rain should be borne by the emitting entity. But in the case of climate change, GHG emissions are fungible. Even the most adventurous litigants do not claim that it was the specific carbon molecules escaping from factory X that caused the damage in country Y. Attribution science proceeds by calculating the overall GHGs in the atmosphere, and attributing a percentage of responsibility for a climate-related event based on the share of overall GHGs in the atmosphere contributed by a given entity – this is how national L&D cases against specific companies, for instance, are able to quantify damages.

The central question for climate adjudication – and this is in some ways the fundamental philosophical question for law-and-science in general – is how science is to be rationally incorporated into legal reasoning. Deciding the question of causation based on the testimony of a causal model could compromise the rationality and fairness of the process. The main problem is that there are multiple ways to fit the given data in a particular circumstance to models. There could be cases (and indeed have been, e.g.: Milieudefensie v. Royal Dutch Shell) where the two opposing parties nominate experts who used the same data but fed into two different algorithms, yielding different answers to the key questions of (1) if and (2) to what extent a weather event was caused by climate change and (3) what degree of responsibility lies with the accused. Even if the two competing models were placed before a scientifically literate judge, the side she picks will be arbitrary, given there are no higher criteria or principles governing correctness – that is, both models might be as legitimate as the other. We needn’t even resort to the fact that judges are seldom acquainted with the technicalities of a science to argue that the process would then be ultimately irrational.

In systemic mitigation cases, where parties accuse states or companies of insufficient future ambition in their mitigation plans, the problem of competing models is typically overcome by a crude appeal to a ‘consensus’ or ‘majority’ view. If the saving principle is simply a numerical tally, it becomes the worst kind of tyranny of the majority, for it is swaying not the legislative but the judicial process. But even if the appeal to a majority view became accepted practice, it is not a solution available to L&D cases. The attribution models there are not general predictions about climate change that every other think tank produces (thereby creating enough models for a majority to emerge) but rather bespoke models linking one entity’s actions to specific weather phenomena. Even if multiple models were produced in an L&D case to create a faux majority, it is likely that the accused country will have the resources to produce more and potentially more convincing causal models. Data is also often more extensively available in developed contexts, given they have more recording stations and have had them for longer. This means more generally that cases revolving around disasters faced in poorer countries will often have weaker models for establishing causation. The competition of causal models in an adversarial setting is thus likely to skew the stakes in favour of the polluter.

This problem exists only because climate science is a newer discipline and, unlike physics, does not yet have a mathematical paradigm against which individual models of causal explanation can be tested. If a case turns on the testimony of ballistics experts, say, the two sides will not have basic disagreements about the adventure of the bullet while using the same data because they are governed by the laws of classical mechanics. Similarly, a case turning on a question of electricity will not produce fundamental disagreements about the science because assessment will be governed by the laws of electromagnetism. In the absence of a mathematical paradigm that bridles and directs the individual assessment of causation, empirical sciences will always be prone to alternate explanations: a circumstance that threatens to undermine the legal process.

The lack of a mathematical paradigm, however, should not prevent legislators from using climate science as the best available information upon which law and policy is to be based. It is better if a political process decides which climate model is to be regarded as trustworthy rather than a judicial process, because, rather than determining culpability, laws and policies indicate a direction of travel. Once a state endorses a scenario, it becomes a clear notification to emitting entities (including the state itself) of what the expected standard of care is.

The Intergovernmental Panel on Climate Change’s (IPCC) science reports could be regarded as one such indicator of best available science in international L&D cases. Given the reports are scrutinised by an international political process, and 195 countries are members of the IPCC, the models they endorse are, in a sense, formalised and legitimised. In international L&D cases, the standard of care expected of countries can be determined based on the threat scenarios listed in IPCC reports. The damages would then be calculated in a staggered fashion, increasing in quantity as the IPCC scenarios grow surer and more dire. Adjudication of liability would then be based on what measures a country should have taken, given that they endorsed a report that predicted a certain likelihood of the disaster in question, and the damages would be in proportion to the derogation of duty determined thereby. The calculation of liability would become smoother if a body like the ICJ declares, perhaps drawing from arguments of national climate judgments, when (for instance, 1990, after the IPCC came into effect) duties of care in climate change began to take shape. In this way, any individual judgment would be saved the embarrassment of implying that the accused party has or has not, and to what extent, caused a given disaster: they would simply be affirming that a duty was notified and therefore expected, which is a mode of reasoning much more familiar to the competence of jurists and courts.       

Responsibility vs Liability

Reducing the scope of liability to activities that occurred after the advent of the IPCC reports (1990) rationalises the L&D process, but also, it can be argued, transfers the burden of liability to developing countries like China and India, who have emitted more GHGs since 1990 than most European nations. This seems unfair when the latter enjoyed more than a century of consequence-free industrialisation. It is important when considering this question not to conflate liability with responsibility. Although a government may recognise the dangers of chemical fertiliser Q, say, it might permit its use in crop R that is essential to feed the population, while limiting or banning Q in the case of other crops. While farmers of crop R would be responsible in this case for the denaturing of the soil, they cannot be held liable because they were not expected to change their behaviour, even when its harms were widely recognised and notified.

Similarly, a central principle (Article 3) of the UNFCCC is Common But Differentiated Responsibilities (CBDR). Given large developing countries like India and China had more mouths to feed, and the standards of living of their citizens was sorry compared to those of Europe and America, CBDR was introduced in the 1990s as a principle to indicate that the bulk of the obligation to reduce GHG emissions would lie with the industrialised nations. This means that vulnerable nations like the AOSIS could argue for a reasonable right to expect mitigation and adaptation finance from developed countries (where they could not expect the same from India and China): and the derogation of those duties can lead to a finding of liability. Thus, although responsibility derivatively conditions liability, they are not the same.

Conclusion

Late medieval theories of monarchy often argued that, beyond military success, economic felicity, and hereditary lineage, the king’s greatest source of authority is that he already wears the crown. The L&D Fund already exists. Would it not be better to make the best of the faulty system we have rather than reaching after an ideal? But history also shows us that not only revolution but regicide is possible and defensible when the basic conditions of a social contract are not met. The basic requirements of an L&D mechanism are that it be able to identify who should pay and how much, and that it has a formal process of demanding that sum. We say demand here and not ensure because no international body or process has the ability to ensure compliance, short of arms – at best, there are means of economic coercion, which are not much relevant in the case of L&D where the weaker are demanding justice from the stronger. Judged from both these basic expectations, the Fund has failed. It betrays no institutional logic that would lead us to believe it will charge developed countries to pay for the Fund in the future – and this seems a problem internal to the logic of distributive justice models. Add to this, the figure of the World Bank at the helm portends a grant-to-loan ratio to make American universities blush. Amid this state of affairs, given the ICJ has agreed to consider inter-state liability for L&D, corrective justice seems a more sensible pursuit: the adversarial process, as we have seen, is designed specifically to identify who pays whom and how much.

But in thus moving from distributive to corrective justice, it is easy to surrender to the temptation of leaving the difficult questions to science. In preferring the court for L&D, it becomes essential to vaccinate the court from the possibility of competing causal models – this circumstance leaves too much to chance, and that chance, too, as we have seen, is loaded in favour of the polluters. The process can be rationalised by restricting the findings of climate science to the political rather than legal realm. Once a set of climate models is endorsed and formalised, the courts can resume their traditional function of discovering duties and standards of care.

By diverting efforts from the Fund to the ICJ and tying the duty of care to the IPCC, we can thus rescue L&D from impotence on the one hand and arbitrariness on the other.  

---

*Dr Anirudh Sridhar is a Senior Visiting Fellow at the Grantham Research Institute on Climate Change and the Environment at the London School of Economics. He is also Dean of Thought Leadership at Alliance University, Bengaluru. He took a DPhil from the University of Oxford in 2020 and is currently pursuing a second bachelor’s degree in pure mathematics. He recently worked on the design of Meghalaya’s framework law on climate change and before that, on the potential design of the now seemingly aborted climate law for India. He is interested in philosophical questions surrounding law and science, particularly in the context of climate change.