THE popular image of the police informer, drawn from crime fiction, film and television portrayals, is of the copper's nark, the grass, the stool-pigeon, a seedy figure lurking outside legitimate policing without legal status or protection. In fact, the police informer has had a 21^st century rebranding and is now a “covert human intelligence source” (“CHIS”), whose role and relationship with the police is detailed in the Regulation of Investigatory Powers Act 2000 (“RIPA”) and the Code of Practice under it. Yet despite their length and level of detail, neither RIPA nor the Code spells out whether the police owe a duty of care to a CHIS they are “running”; the Code does provide any “public authority deploying a source should take into account the safety and welfare of that source, when carrying out actions in relation to an authorisation or tasking, and to foreseeable consequences to others of that tasking”, but, like RIPA, is silent as to whether any enforceable civil law obligation is created. This was the context of the Court of Appeal's decision An Informer v A Chief Constable [2012] EWCA Civ 197.

C was having financial dealings with a man referred to as X and learnt that X was involved in fraudulent criminal activities, and therefore informed the police. The police obtained authorisation under RIPA for the use of a C as a CHIS and thereafter C continued his relationship with X, passing valuable information to the police about X's activities. C's existence and identity was known only to his “handlers”; the team of officers investigating X was kept completely separate, to preserve C's anonymity and avoid compromising the investigation. Unfortunately, this separation of functions led to problems because, when the investigating team arrested X and charged him with serious offences, they looked into X's on-going dealings with C and promptly arrested C on suspicion of money laundering. When arrested, C did not disclose his status or role as a CHIS.

A few weeks later, the investigating officer responsible for C's arrest applied to court, without notice to C, and obtained a restraint order against him under the Proceeds of Crime Act 2002, prohibiting him from dealing with any realisable property. Case law provides that prosecutors applying for a restraint order owe a “duty of candour” to disclose all relevant matters to the court, but in C's case the judge was not told of his status as a CHIS. The restraint order was eventually lifted after several months, and C never charged with an offence. C claimed that the restraint order “put him in dire financial straits” and sued the police. He alleged that his handlers had given assurances that his and his family's personal safety, welfare and livelihood were their first priorities, which gave rise to a contractual and tortious duty of care, owed to him as a CHIS, and which extended to pure financial loss. The police conceded that they owed him a duty of care in tort to keep him reasonably safe from physical harm arising from his CHIS status, but disputed that any duty was owed to C to safeguard his economic interests.

Somewhat unusually, the question of whether the police owed a duty of care was not litigated as a preliminary point of law, but instead the claim proceeded to full trial. The trial judge (whose judgment has not been published because of the highly sensitive nature of the facts) found that the police had given no express contractual assurances to C that his livelihood or financial welfare would be protected and that no such assurances could be implied. This finding also ruled out the necessary assumption of responsibility to support a duty of care in tort for pure financial loss, leaving as the default position the police's general, public policy immunity from civil liability when engaged in frontline police duties derived from Hill v Chief Constable of West Yorkshire [1989] A.C. 53. For completeness, the judge went on to consider C's various allegations of breach. Most were rejected easily, but he found the police's conduct in relation to the restraint order – their failure to disclose C's identity and the length of time before it was lifted – “very difficult and troubling”. Nonetheless, without a relevant duty, that conduct could not be actionable and C's claim failed. C appealed to the Court of Appeal.

The Court of Appeal unanimously dismissed C's appeal. Each member of the court adopted the same basic approach, deciding the case on tortious rather than contractual principles (on the basis that if a tortious duty was rejected, there would be no prospect of implying an inconsistent contractual obligation). They accepted C's argument that an explicit “assumption of responsibility” is not a prerequisite to a duty of care for pure economic loss, since the phrase expressed a legal conclusion rather than an essential factual element. But beyond these common starting points, their reasoning was significantly different. The most straightforward approach was that of Toulson L.J., who ruled out a duty of care without obvious agonising. He relied on his own dictum from Glaister v Appleby-in-Westmoreland Town Council [2009] EWCA Civ 1325 that for a duty of care for pure financial loss to arise,

So any duty must be connected with the purpose of the CHIS relationship, the prevention and detection of crime, but for Toulson L.J., the suggested duty to safeguard the economic interests of a CHIS would, if anything, conflict with the police's responsibility to the public. In addition, since the Proceeds of Crime Act gives the “victim” of a restraint order a right to compensation only in cases of “serious default” falling short of negligence, “it would not be right for the court to create a right of damages for negligent non-disclosure”.

Arden and Pill L.JJ. found it much harder to rule out a remedy for C's treatment. Unlike Toulson L.J., both emphasised that exceptions to Hill have been recognised, including in Swinney v Chief Constable of Northumbria Police Force [1997] Q.B. 464 where a duty of care was owed to an informant (in respect of physical harm) whose identity the police had undertaken to keep confidential. And both appeared to decide that the same exceptional treatment applied in C's case, rejecting many of Toulson L.J.'s conclusions. Pill L.J. explicitly found that the police owed “a duty to C arising out of the proximity of the relationship, which … could cover his financial welfare…” but, with considerable reluctance, concluded that the police were not in breach of this duty, although had they proceeded to charge or detain him, “the position might have been different”. Arden L.J.'s reasoning is less clearcut, initially treating Swinney as an exception “displacing” the Hill immunity and aligning C's case with Swinney's, then later concluding, “the public policy underpinning the investigations immunity prevails over that of protecting the CHIS from purely financial harm. That means that the assumption of responsibility imposed on the police was displaced.” Yet her reasoning in reaching this “no duty” conclusion is really that of “no breach”– C's identity had to be kept from the investigating officers for good reason, his status was never revealed and he was physically protected: in short, then, the police had not behaved unreasonably in the circumstances.

It is not surprising to find yet another elision of breach and duty reasoning in a novel negligence case – judges will apply whichever concept serves their favoured conclusion and at least here the “no breach” conclusion was legitimately available, unlike in appeals solely on the preliminary “duty” question of law, where “no breach” reasoning must be disguised as “no duty”. English law's fascination with employing the duty concept frequently hides the fact that a nuanced approach to the standard of care (e.g. the reasonable defendant would do more to preserve physical safety than financial wellbeing) can solve many difficult problems traditionally left to the duty of care, ensuring the flexibility to give a remedy in a truly egregious case of wrongdoing without a problem of blanket immunity. And when you think about it, even Toulson L.J.'s wholly orthodox statement of when a duty of care will exist for pure financial loss contains, at its heart, the circular paradox of defining duty by reference to breach – surely the claimant can only “properly expect to be entitled to rely on the defendant to safeguard him from economic harm likely to result from want of care on the part of the defendant” (emphasis added) if there is a duty to start with.