BeCause the World is Round

How to assassinate the rule of law and dance on its grave

What we lack is knowledge and the law deals with lack of knowledge by the concept of the burden of proof.

-Lord Hoffmann, in Gregg v Scott [2005] UKHL 2

A controversy about legal causation is not, on the face of it, a matter likely to cause rioting in the streets. The case I am going to make in this article is that it should. By the end of it I intend to make you as mad as hell and unable to take it anymore. Especially if you are Dutch. You have been warned.

I am going to win you around first by talking about the small subject of the metaphysics of legal causation. I am then going to on to discuss recent developments in what I will call, with my tongue firmly in my cheek, climate change ‘jurisprudence’. But before that, I need to briefly explain what legal causation is and why it matters.

So: imagine I shoot you in the head with a gun. Few of us would dispute that as a matter of common sense my action will have caused you harm. But the point of law is largely to intervene in human affairs in circumstances where common sense is not good enough, and you will be able to imagine situations in which causation is a matter of dispute.

The classic example of this is Weld-Blundell v Stephens [1920] AC 956. The claimant had employed the defendant, an accountant, to look at the books of a company to which the former had lent money. A letter the claimant wrote to the defendant contained libellous statements about various officers and former officers of that company (he called one of them an ‘ingenious thief’). Through the defendant’s negligence the letter ended up in the hands of those who had been libelled, and they successfully sued the claimant for defamation. He then sued the defendant for breach of a confidentiality obligation.

Nobody really disputed the confidentiality obligation had been breached. Nor did anybody dispute that the claimant had suffered a loss (he’d had to pay a big sum in damages to the people he had libelled). The question was whether the former had caused the latter.

In their famous - to lawyers - book Causation in the Law (1959), HLA Hart and Tony Honoré drew a distinction between cause and conditions. Imagine a scenario in which a man is walking past a petrol station and casts aside a lit cigarette, which happens to land in a trickle of spilled petrol. It ignites and explodes and the entire petrol station goes up in flames. There are all sorts of conditions which resulted in this scenario in the final outcome of conflagration. There was the spilled petrol. There was the thrown cigarette. There was the oxygen in the air. There were the synapses which fired in the man’s brain reulting in him making the decision to cast aside his cigarette at that particular moment. There was the fact that a shopkeeper had sold the man a pack of cigarettes. There was the fact that he had made the decision to take up smoking at the age of 13 at the instigation of his first girlfriend. And so on. All these things were necessary to have happened in order for the disaster to be the result or consequence. But the law is interested in one particular condition that it wants to single out as the cause so that it can attribute liability. Which should it pick?

In Weld-Blundell v Stephens it was a toss-up between whether it was the defendant’s breach of confidentiality which had caused the loss, or the fact that the letter had contained defamatory statements in the first place. And in the end the court decided that the loss had been caused by the latter. The fact that the defendant had negligently allowed the letter to get into the hands of the wrong people had created an opportunity for the loss to occur - it was a ‘condition’. But it was not the cause. The cause of the loss was the letter’s contents, penned by the claimant. And the reason the court decided this seems to have been that it was the causa causans or primer mover. The loss really stemmed from the fact that the claimant had been badmouthing people. The buck ultimately had to stop with him.

Now, the important thing to say about this is that, as Hart and Honoré observed about all questions of legal causation - drawing ultimately from the ideas of the philosopher RG Collingwood - the court did not make its decision because the libellous content of the letter took metaphysical precedence over the breach of confidentiality. Both of these conditions ‘mattered’ as much as each other in that sense. Rather, the court was making a normative judgment. It was identifying cause on the basis of which of the two parties it felt ought to be responsible. The claimant should have been liable for his own loss because he was more to blame, and therefore he had caused it.

Collingwood, you see, in his Essay on Metaphysics, made the case that questions about causation are never to do with what is metaphysically more salient, but always to do with the fact that a particular agent is making a decision in a particular context, and performing a particular function with a particular view of the facts. What caused Marilyn Monroe’s death? Well, lots of things did. To the toxicologist who examined her, it was barbiturate poisoning. But a psychiatrist might say something about suicidal ideation. Clearly, these both ‘caused’ her death in the sense of being conditions that resulted in her dying. But which one is singled out as ‘the cause’ is contingent, depending on what job is being performed and why the analysis is taking place.

Weld-Blundell v Stephens was a court case, and so the court was making its decision on cause on the basis of where liability justly ought to be assigned. In its view, what was just was ultimately for the claimant to bear responsibility for his own conduct. Another way of putting this is that a third party’s negligence should not absolve a tortfeasor of the consequences of his civil wrongdoing in the absence of some other justification. No doubt if the contract beween claimant and defendant had stipulated that the defendant would be liable for consequential losses if found to have acted negligently the court would have decided otherwise. But by default the claimant had to own his own conduct.

As you can see from this, on the one hand legal causation is far from an exact science. This is because courts are doing the best they can to grapple with what are ultimately extremely fraught philosophical questions. Even if, like Laplace’s Demon, we were able to establish the exact position and trajectory of every particle in the universe so as to trace everything back to the Big Bang, we would still have to make normative decisions about which events are causal of others in order to make sense of anything at all. And in the end courts are always required to decide that something is somebody’s fault (or not). This would be impossible without lines being drawn on somewhat arbitrary grounds. If every murderer could point to not being hugged by his mother as the cause of his psychopathy; if every act of negligence could be attributed to a hormonal imbalance in the brain; if every road traffic accident could be said to have been caused by a distracting, intrusive thought, then nobody could be made responsible for anything. Courts have to muddle through as best they can.

But on the other hand, clearly the decisions courts make are not completely arbitrary, and we want them to base their decisions about causation on justifiable grounds. And this is I think best exemplified by the reasoning of the House of Lords in the important causation case of Gregg v Scott [2005] UKHL 2 - most notably in the judgment of Baroness Hale. Brenda Hale has come in for some criticism on this substack over the years. But in this case she gave a lucid and thoughtful judgment, to which I will come in a moment. First, though, the facts.

Mr Gregg developed a lump under his arm and went to see his GP, Mr Scott. Scott told him the lump was benign. A year later Gregg moved house and registered with a new GP, who took one look at the lump and said, ‘My God man, you have cancer!’ or words to that effect, and sent him to hospital. By that time the tumour had spread into Gregg’s chest and he went through various stages of treatment to no avail. He was told that his chance of survival over the next 10 years was less than 50%. He sued Scott for professional negligence, basing his claim on a statistical framing that if he had been diagnosed promptly he would have had a 42% chance of survival over 10 years, but his actual chance of survival, due to the delayed diagnosis, was only 25%.

Nobody disputed that Scott had acted negligently. The problem was causation. It turned out that Gregg survived for at least the next 10 years - his prognosis had been given in 1995 and the case before the Supreme Court (litigation can take a long time unless you’re trying to prevent Brexit) was heard in 2005, at which point he was still alive - apparently as fit as a flea. He had no doubt suffered a wrong, and he had suffered an elevated risk as a result. But it was a risk that had apparently not materialised. How could anybody possibly measure or determine, then, that the negligence of Scott had caused Gregg harm? How was it even possible to retrospectively determine that the declaration that he had had a 25% chance of survival had been accurate?

This relates to a problem which people who are interested in such matters call ‘uncertainy about uncertainty’. Let’s say a weather forecaster tells you there is a 99% chance of rain tomorrow. Tomorrow, it doesn’t rain. Well, was the forecast accurate? Who knows? Maybe it was inaccurate. But maybe it was accurate and it just turned out that the 1% chance of not-raining came to fruition. You can’t tell. All you can really say is that across time it may be the case that the weather tends to produce generally accurate forecasts in the sense that when it predicts a greater than 50% chance of rain it tends to rain, and so on. The model can be assessed in aggregate but not at the level of the individual day.

So it was here. Is the reason Scott had not died by 2005 because he had not actually suffered an elevated risk of death? Or had he suffered an elevated risk of death which had nevertheless still not manifested?

Hale, supported by a majority in the House of Lords, got out her metaphorical oujia board and summoned the spirit of Collingwood. Courts, she decided, just shouldn’t touch this kind of issue with a barge pole. Courts can say, when somebody has suffered a harm, that it has been caused by x, y or z. It’s their job to do that (remember, identification of cause derives from a particular agent performing a particular function). But they cannot as a matter of justice get involved in discussions about whether negligence has caused risk alone. Partly this was for practical reasons (courts have to think about the consequences for insurers, the NHS, and so on). But mostly it was because courts are not medical prognosticators. There is something qualitatively different about applying rules to facts (which is what courts do) and forecasting about possible future events: there are no facts about the future. And applying law to possible future events is not judicial. It is another type of activity, one which courts should not engage in.

There is an important phenomonological point to be drawn from this but I will not tease it out here in the interests of space. You will I think intuit the point that doing justice means applying law to things that have happened rather than to things that may happen. And now, having established this - that legal causation is the identification of cause by courts performing a particular function and performed not as a metaphysical exercise in establishing precedence of causal factors but as a judicial exercise designed to ‘do justice’ in some sense - we can finally turn to the issue that ought to get you in the mood for a riot, which is climate change jurisprudence.

Rewind two years. Roughly around this time in 2024, you may recall, the European Court of Human Rights ruled that the Swiss were being naughty because they were not doing enough to mitigate climate change by reducing greenhouse gas emissions (GHGs). The tortured reasoning, in the case of KlimaSeniorinnen v Switzerland, was as follows: the European Convention on Human Rights (ECHR) contains a right to private and family life (Article 8). This was originally designed to prohibit arbitrary interference in private correspondence, and to prevent states overriding the wishes of parents in respect of educational choices. But the Court decided that it also included a right to be protected from the effects of climate change, because climate change may result in heatwaves, which may shorten the lives of old people, which may impact on their family lives.

You will immediately see the first problem here, which is that the claimants in the case - a group of old women and a charity representing older women who were concerned about climate change - had not actually suffered any harmful effects from climate change at all. They made some noises about heatwaves affecting their ability to go swimming and so on. But their case rested on potential harm in the sense of elevated risk that might manifest itself at some notional point in the future. Worse - while this is a rather macabre point to make, they were all already of advanced age (in their late 70s and 80s). And they were therefore likely to be already dead by the time the predicted negative consequences of climate change would really begin to bite (i.e. in the 2030s, 2040s and 2050s). How then could it possibly affect their family lives?

And you may also have seen the second problem, which is that these women were suing their own homeland of Switzerland, which with the best will in the world is hardly the source of the problem when it come to GHG emissions, even if one concedes that all the worst-case scenarios about climate change are true. Switzerland is not exactly renowned as a hotbed of industry or large scale pollutor. It has some cows, which presumably produce a degree of methane. And there is a fair bit of yodelling that goes on, which must result in a bit of hot air floating around. But what the Swiss mostly do is make cuckoo clocks, eat toblerones, and count their gold. The country therefore produces comfortably less than 1% of global emissions. Even if, in other words, the case could be made that the old women litigating the case were suffering harm as a result of climate change, how could it justly be said that this was caused by the Swiss state’s failure to reduce its emissions?

The Court dealt with this issue in exactly the opposite way to how the House of Lords had decided matters in Gregg v Scott. And it did it in a pretty flagrant way. Yes, it conceded, it was essentially impossible to make a coherent, logical argument that the claimants had even suffered a harm, let alone that such a harm had been caused by the Swiss government. But that didn’t matter. This was because, and I directly quote the judgment:

The adverse effects on and risks for specific individuals or groups of individuals living in a given place arise from aggregate GHG emissions globally, and the emissions originating from a given jurisdiction make up only part of the causes of the harm. Accordingly, the causal link between the acts or omissions on the part of State authorities in one country, and the harm, or risk of harm, arising there, is necessarily…tenuous and indirect…

Furthermore, from the perspective of human rights, the essence of the relevant State duties in the context of climate change relates to the reduction of the risks of harm for individuals. Conversely, failures in the performance of those duties entail an aggravation of the risks involved, although the individual exposures to such risks will vary in terms of type, severity and imminence, depending on a range of circumstances. Accordingly, in this context, issues of individual victim status or the specific content of State obligations cannot be determined on the basis of a strict conditio sine qua non requirement.

In other words, it’s just really important that states reduce their GHG emissions. Yes, the individuals in question couldn’t prove they were harmed and yes, the causal link between the acts or omissions of Switzerland was tenuous and indirect. But we shouldn’t really care about pedantic, fussy lawyers making objections about silly, archaic, outmoded concepts like the rule of law. What is the law for other than to make people do things? And the Swiss should be made to reduce their GHG emissions. Case closed. Switzerland caused the claimants harm because we said so.

I wish that things were reasoned out more carefully than that in the judgment and that the barest fig leaf of legal niceties were placed in front of it, but the truth of the matter is that things were really not put a great deal less bluntly than I have put them here. Law, to repeat, in the ECtHR’s eyes means just getting people to do things. If people ought to do something, human rights law must make them do it. Rules don’t matter. What matters is outcomes and that’s that.

What you will have sensed from this description of the KlimaSeniorinnen case is that, at least when it come to climate change, there is no longer even a good faith attempt being made to adhere to the notion that courts are judicial bodies and that law is a distinct phenomenon in its own right. Instead, the court is doing something else - it is simply being used as a strong-arm mechanism for forcing conformity with the demands of a global political project. And this is becoming nakedly evident in the field of climate change ‘jurisprudence’ - where the requirement to establish even a rudimentary case for causation is now more or less routinely ignored in order to operationalise compliance.

My favourite example, up until recently, of the sheer drivel that is spouted in this regard is actually not strictly speaking from a court case, but what in the UN is called (unironically) a ‘quasi-judicial process’ - a committee of jurists and experts who hear individual complaints against States and issue non-binding judgments. In 2021 the UN Committee for the Rights of the Child, in Sacchi v Argentina, came up the following gem when considering a similar claim to that of KlimaSeniorinnen but presented by a group of youths who alleged that climate change would destroy their futures:

[T]he Committee finds that the collective nature of the causation of climate change does not absolve the State party of its individual responsibility that may derive from the harm that the emissions originating within its territory may cause to children, whatever their location.

The Committee was not ashamed by the sheer barefacedness of its reasoning on this point or the ludicrous outcome implicit in it, wherein ‘whatever their location’ children may sue any state on the basis that it may have caused them harm which they may suffer in future; it is difficult to imagine a more meretricious approach to what one still laughably must call ‘legal reasoning’ than this.

But this is where the Dutch come in. Earlier this year, in the case of Greenpeace v State of the Netherlands, the district court of The Hague was asked to decide whether the Netherlands’ failure to control and mitigate the effects of GHG emissions was impacting on the right of the people of the Caribbean island of Bonaire, represented by Greenpeace, to ‘experience and practice their culture’ (since Bonaire, as a low-lying island, is purportedly more at risk from rising sea levels and flooding, has delicate ecosystems, relies on fishing, etc.).

Here, as might be expected, the Dutch government made the case, as the Swiss government had done in KlimaSeniorinnen, that even if the harms alleged by Greenpeace manifested themselves at some future date, it would not really make sense to make the Dutch state liable, since its acts or omissions could hardly be said to have been a significant causal factor in generating climate change. The Netherlands has done rather a lot to reduce GHG emissions - more than most countries - and in any case is responsible for a tiny percentage of the whole. Why on earth should it, then, be the one in the firing line as the ‘cause’ of the elevated risks to the people of Bonaire, assuming for the sake of argument that the risk was indeed elevated.

What the district court of the Hague said in response to this was naked: when it comes to climate change, we inhabit the Bizarro World of Htrae, wherein everything is reversed. Normally, the burden of proof for legal causation is on the claimant. This is for the very good reason that by default the burden of proof in law is always on the claimant (or the prosecutor in a criminal trial) - a principle dating back at least to the time of the Romans. Ei incumbit probatio qui dicit, non qui negat - which, unless my Latin fails me, means that proof lies with he who asserts, rather than he who denies. Why? Well, it is hard for a defendant to prove a negative, for one thing. For another, the claimant is the one who is instigating the litigation, and therefore has had all the time he desires or needs to amass evidence and make out his claim, whereas the defendant is on, well, the defensive. For yet another, if the claimant is going to be relying on the coercive power of the state to extract money from the defendant, he ought to have a jolly good reason for it, and the onus should therefore be on him to prove his claim is a good one.

It seems that in climate change cases, however, or at least those which pertain to human rights, the opposite is now true. Here is the Hague district court:

[F]or the application of the ECHR, the courts assume in climate cases that there is a causal link between a member state causing or failing to tackle climate change and certain types of consequences that individuals are known to experience as a result of climate change. This means that complainants in climate cases do not have to argue or prove these elements…In the context of climate change, it is therefore up to the member state to argue (and, if necessary, prove) that there is no causal link [emphasis in original] between the conduct specifically complained of and the consequences known to affect individuals as a result of climate change.

In short: when an individual brings a claim against a state for failing to tackle climate change on human rights grounds, he does not need to prove that he has suffered the ‘certain types of consequences that individuals are known to experience as a result of climate change’, and the state has to prove that it did not cause those consequences in regard to him. That is to say, the state has to prove that it did not cause something that has not happened and may never happen. We all know that proving a negative is tough. But the state in a climate change-related human rights case apparently has to prove a negative about a not as-yet existing event in order not to be found liable.

You will have seen that describing this as ‘law’ is like describing Kampuchea as ‘Democratic’. This is not law and the process is not judicial. It is simply a system designed for the enactment of show trials whereby states can be continually and relentlessly browbeaten by motivated litigants into being found in breach of human rights obligations for failing to be ever more focused on net zero fanaticism. It is not adjudicatory and the entire process is predetermined. It does not apply rules to facts; it applies a politicised imperative to governments to simply do more and more at whatever cost.

It is grimly amusing in a kafkaesque way, of course, but the essence of the matter is genuinely both bleak and disturbing. What it suggests is that commitment to the rule of law - and commitment indeed to the idea of judicial process as such - is becoming critically weakened in what is supposed to be its heartlands. We live in a time during which intelligent, educated people have convinced themselves that, if the cause they believe in is ‘right’, then the law simply ought to realise it. The most basic elements of law recognised since classical antiquity can be turned on their head or ignored in the interest of realising desired outcomes. And courts are thereby reduced to simply parroting dogma and ordering its implementation.

And the prospects are worrying when we consider that climate change is far from the only issue that those with wealth and status in our societies care about. You will have your own lists of such issues in mind; others will no doubt come out of the blue in the coming years (one suspects that suppression of ‘populism’ and the ‘far-right’ will increase in salience in this regard). The signs are all pointing in one direction: further weaponisation of courts to achieve political ends. My advice is to get ready for it.