IN Tindall v Chief Constable of Thames Valley Police [2024] UKSC 33, [2024] 3 W.L.R. 822, the claimant’s husband, Mr. Tindall, died following a collision with another motorist who had lost control on a patch of ice. Shortly before this fatal crash, another driver, Mr. Kendall had lost control on the same patch of ice, crashing without serious injury. He had attempted to flag down traffic to warn other motorists of the danger of ice on the road, before phoning the police. He spoke to a civilian call handler, whom he informed about the danger but whom he did not inform about his attempts to warn other motorists. On arrival, the police spoke to Mr. Kendall and placed him in the care of an ambulance crew who removed him from the scene. They then left the scene as it was before the crash, without putting in place any further warning. Soon after, the fatal collision involving Mr. Tindall occurred. It was an agreed fact that but for the actions of the police Mr. Kendall would have continued his attempts to warn other motorists.

The high court rejected the police’s strike out application, holding that whether the police made matters worse should be settled at trial. The Court of Appeal (Stuart-Smith, Thirlwall and Davies L.JJ., noted [2022] C.L.J. 245) allowed the police’s appeal, concluding that the police had not made matters worse, but rather, had failed to confer a benefit ([2022] EWCA Cic 25 at [66]–[69], C.A.). None of the defined exceptions to the general rule that there is no liability for omissions applied and the court thus held that the police owed no duty of care to Mr. Tindall.

The Supreme Court, dismissing the claimant’s appeal, agreed that the police owed no duty of care, but for different reasons than those given by the Court of Appeal. The Supreme Court held that the police’s actions had the effect of stopping Mr. Kendall from attempting to warn other motorists (at [60]). However, the police owed no duty of care because they did not know and could not reasonably have known, that Mr. Kendall had been flagging down traffic to warn other motorists of the danger of ice (at [58]–[68]).

There are many noteworthy features of the judgment: the endorsement of the “interference principle” (at [58]); the affirmation that the law on negligent omissions is no longer “in a state of flux” after a flurry of recent cases at the highest level, including Michael, Robinson, Poole and HXA (at [68]); the restatement of a “uniform”, as opposed to “policy” approach to public authority liability, reaffirming a central theme of this flurry of cases (at [1]); and the reassertion that liability for the acts of third parties is governed by the same rules as all omissions cases (at [44(iv)], [56]).

This note focuses on the lack of clarity in the Court’s judgment about the precise normative significance of the fact that the police could not reasonably have foreseen that they were making matters worse. The question is whether this fact entirely extinguished any duty not to make matters worse for Mr. Tindall or whether it merely defeated an existent prima facie duty.

One possible analysis is that there was in principle a duty on the part of the police (a “notional duty”) because they did make matters worse for Mr. Tindall, but no factual duty in this particular case because they did not know and could not reasonably have been expected to know that they had made matters worse (“no factual duty”). An alternative, importantly different, analysis is that because the police did not know and could not reasonably have been expected to know that they made matters worse, there was no notional duty in the first place (the “no notional duty” view). If the “no factual duty” analysis is adopted, then the notional duty arises regardless of the extent of police knowledge; whereas if the “no notional duty” analysis is adopted, then the existence even of the notional duty itself is dependent on the epistemic position of the police. Unfortunately, both of these incompatible analyses find some support in the Supreme Court’s judgment.

There are various indications that the Supreme Court endorsed the “no factual duty” analysis. Crucially, the court was clear that the police had caused Mr. Kendall to desist from his attempts to warn other motorists, that it was irrelevant that they did nothing specific to encourage him to desist and that so long as “the activity of the police as a whole created a danger” this was enough to give rise to a prima facie duty of care (at [60], [61]). Moreover, the requirement for the defendant to reasonably foresee the risk was said to originate simply from “the well-established approach to establishing any duty of care” (at [58]). And the Court was clear that the lack of foreseeability in this case was a “factual difficulty” (at [59], [68]). If we attend to these features of the Court’s judgment, we are driven to the conclusion that there was a notional duty and that it was merely the absence of reasonable foresight that negated the liability of the police.

However, there are also some passages of the judgment which suggest that the court was tempted by the alternative “no notional duty” analysis. Most importantly, the court endorsed McBride and Bagshaw’s formulation of the “interference principle” (at [50], [58]): “[I]f A knows or ought to know that B is in need of help to avoid some harm, and A knows or ought to know that he has done something to put off or prevent someone else helping B, then A will owe B a duty to take reasonable steps to give B the help she needs” (emphasis added). See N. McBride and R. Bagshaw, Tort Law, 6th ed Harlow (2018) 213-217.

By endorsing this form of the principle, the court seems to suggest that existence of even the notional duty is dependent on the defendant’s state of knowledge. This reading of the judgment is reinforced by the Court’s observation that where the interference principle applies, it is simply “a particular way of making matters worse” (at [49]). If one combines the two premises – (1) that the interference principle applies only when the defendant knows that they are interfering and (2) that whether or not the interference principle applies is just another way of saying whether or not the defendant made matters worse – one could erroneously be led to the conclusion (3) that if the defendant did not know they were interfering, they did not make matters worse.

These conceptual problems we highlight could have been avoided if the issue of whether the defendant had made matters worse had been clearly identified as a matter of notional duty and the issue of whether the defendant reasonably foresaw that they were making matters worse had been clearly identified as a matter of factual duty. Whether the “interference principle” applied would then clearly have been determined by whether the defendant had performed a particular form of act – preventing another person from protecting the claimant – and the state of the defendant’s knowledge would clearly have been irrelevant to this determination. If the judgment had formulated the interference principle in this clear way (as Tofaris and Steel formulate in [2016] C.L.J. 128) rather than adopting McBride and Bagshaw’s confused formulation, then the court would have found that there was a notional duty independently of the epistemic position of the police and would then naturally have moved on to a conceptually distinct, second-stage inquiry into whether the police knew or ought to have known, that their interference would create a risk for other motorists.

It might be tempting to see all of this as philosophical nitpicking – given that treating (or half-treating) reasonable foreseeability as a necessary condition of notional duty did not in practice affect the outcome of the Supreme Court’s judgment. However, as many academic commentators have argued, conflating the different elements of the duty of care may in the long term foster judicial confusion that has practical consequences (see D. Nolan, “Deconstructing the Duty of Care” (2013) 129 L.Q.R. 559, 581). For example, this conflation might lead a court to see foreseeability as not merely one necessary condition of a duty of care, but as a factor influencing the entire duty enquiry. This in turn might lead a court to the false conclusion that the more foreseeable the risk to the claimant, the more likely it is that the defendant owes a duty of care. More pertinently, a court might be led to the idea that the more foreseeable the risk to the claimant, the more the defendant can properly be described as making matters worse. Indeed, there is already confusion amongst commentators over whether the Supreme Court in Tindall held that the police made matters worse or not (see N.R. Langen, “Acts v. Omissions”, LRB Blog, 28 November 2024). This general confusion about what exactly constitutes “making matters worse” can only serve to encourage errors of exactly the sort made by both the high court and Court of Appeal in this case.