The statutory tort of harassment as set out in the Protection from Harassment Act 1997 was considered by the Supreme Court in Hayes v Willoughby ( UKSC 17). The issue before the Court concerned the scope of one of the defences to the tort (section 1(3)): that the conduct complained of was for the purpose of preventing or detecting crime. The Supreme Court held that this defence can succeed only if the defendant rationally believed that his or her activity was for this purpose.
The claimant (Mr Hayes) owned a number of companies one of which the defendant had worked for prior to his dismissal. The claimant had accused the defendant of wrong-doing in the course of his employment (specifically of an attempt, with others, to forcing one of the claimant’s companies into liquidation with a view to buying back the business for themselves). Litigation in both the High Court and employment tribunals formed the background to the conduct complained of: “an unpleasant and obsessive personal vendetta” by the defendant against the claimant () in the course of which allegations of fraud, embezzlement and tax evasion were made to the authorities. The defendant wrote hundreds of letters and the claimant was investigated by the Official Receiver, the DTI and the police, all of whom concluded that the allegations were baseless. He also obtained confidential information from the claimant’s ex-wife about his mental and emotional ill-health which he “gratuitously passed … to third parties in order to generate prejudice” against the claimant, alleged to the claimant’s GP that he (the claimant) had forged the GP’s signature on sick notes and had alleged to the claimant’s landlord in the US that the claimant was about to go bankrupt.
The judge at first instance found that the defendant’s activities constituted harassment within the meaning of the Protection from Harassment Act 1997 (“the 1997 Act”), amounting to a course of conduct linked by a common purpose and subject-matter, which was calculated to, and did, cause alarm, distress and anxiety to the claimant. The fact that the defendant did not directly communicate with the claimant did not prevent a finding of harassment, a matter not appealed to the Supreme Court.
It was argued for the defendant that his conduct had been “for the purpose of preventing or detecting crime.” The judge at first instance had found that the defendant’s conduct was animated by “mixed motives, including personal animosity to [the claimant] (in fairness based largely on the same suspicions) and a sort of intellectual curiosity”; further that the defendant “has at all times sincerely believed that [the claimant] had stolen large sums from his companies in the United Kingdom and committed a variety of offences in the course of doing so” and that the conduct was “subjectively directed at the prevention or detection of crime. ” Although that belief at first had a reasonable basis this ceased to be the case by June 2007 after the reports of the various investigations undertaken into the claimant at which stage the defendant’s persistence “exceeded even the widest limits of reasonableness and became unreasonable and obsessive”.
The Judge had ruled in the defendant’s favour on the basis that the defence on which he relied imposed an entirely subjective test, and it was accepted that the defendant believed that his actions were concerned with the prevention or detection of crime. The Court of Appeal ( EWCA Civ 1541) allowed the appeal on the basis that the purpose of the defendant’s conduct ceased to be reasonably or rationally connected to the prevention or detection of crime after June 2007, also that this purpose did not in any event apply to the defendant’s intrusions into the claimant’s personal life. It was these findings that the defendant appealed to the Supreme Court.
The leading judgment was given by Lord Sumption with whom Lords Neuberger and Wilson agreed, Lord Mance concurring and Lord Reed dissenting. Lord Sumption took the view that the defence was triggered where the prevention or detection of crime was the harasser’s dominant (as distinct from sole) motive (). He disagreed with the various authorities which had adopted either an objective or a subjective approach to the defence.
The former () was “not consistent with either the language or the purpose of the Act”, which provided the defence (s1(3)) where “the person who pursued [the course of conduct complained of] shows – (a) that it was pursued for the purpose of preventing or detecting crime … or (c) that in the particular circumstances the pursuit of the course of conduct was reasonable.”
The latter () was problematic because conduct which otherwise amounted to harassment had to pass a threshold of gravity which merited criminal sanction (Majrowski v Guy’s and St. Thomas’s NHS Trust  1 AC 224 ) and because, however well-intentioned, individuals who engaged in such conduct required to be controlled. “[V]igilantism can easily and imperceptibly merge into unlawful harassment” (), and the tort of harassment was the primary mechanism for controlling obsessive and oppressive conduct which did not involve, for example, assault or criminal damage. For Lord Sumption: “
“It cannot be the case that the mere existence of a belief, however absurd, in the mind of the harasser that he is detecting or preventing a possibly non-existent crime, will justify him in persisting in a course of conduct which the law characterises as oppressive”. 
Having decided () that “Some control mechanism is required” Lord Sumption found it in the concept of rationality (). He stressed that this was not to be confused with reasonableness (even with the “broader categories of Wednesbury unreasonableness, a legal construct referring to a decision lying beyond the furthest reaches of objective reasonableness”):
“Reasonableness is an external, objective standard applied to the outcome of a person’s thoughts or intentions… A test of rationality, by comparison, applies a minimum objective standard to the relevant person’s mental processes. It imports a requirement of good faith, a requirement that there should be some logical connection between the evidence and the ostensible reasons for the decision, and (which will usually amount to the same thing) an absence of arbitrariness, of capriciousness or of reasoning so outrageous in its defiance of logic as to be perverse” .
Rationality, Lord Sumption stated, required that the harasser had “thought rationally about the material suggesting the possibility of criminality and formed the view that the conduct said to constitute harassment was appropriate for the purpose of preventing or detecting it” (). Failure to follow these steps would mean that someone who engaged in a course of conduct amounting (subject to any defence) to harassment would not be regarded as having taken the relevant steps to form a relevant purpose, and that the necessary causal connection between purpose and conduct would be lacking. Lord Mance agreed, suggesting () that “Mere unreasonableness is not the limit. But the law recognises looser control mechanisms such as complete irrationality, perversity, abusiveness or, indeed, in some contexts gross negligence”.
Lord Reed dissented, finding himself “unable to agree that Parliament”, which had not expressly required reasonableness, or indeed rationality, as an element of the defence, “may nevertheless have intended to impose a requirement that the pursuit of the course of conduct should have been rational”. As he pointed out (), the statutory prohibitions on harassment involved a “significant exten[sion of] the reach of the criminal and civil law in controversial circumstances”, “care [being] taken to identify expressly occasions when conduct was to be judged by an objective standard”.
The defence provided by §1(3) applied to the criminal offence of harassment as well as to the tort and, as Lord Reed pointed out,
“It is trite that a statute is not normally to be construed as extending criminal liability beyond the limits which Parliament itself made clear in its enactment”.
Further, the approach taken by the majority would have the effect that juries would be required to determine the “boundary separating the unreasonable from the irrational”.
“I have to confess that I am not sure that I understand the distinction drawn at para 14 between on the one hand “rationality [as] a familiar concept in public law”, which “is not the same as reasonableness”, and on the other hand “the broader categories of Wednesbury unreasonableness”; or the statement that there should be “an absence of arbitrariness, of capriciousness or of reasoning so outrageous in its defiance of logic as to be perverse”, but that the court is not referring to “a decision lying beyond the furthest reaches of objective reasonableness”; or how that test is related to the causal connection between the purpose and the conduct, discussed in para 15; or whether it is the same test as is reflected in the various standards, ranging from gross negligence to “complete irrationality”, mentioned by Lord Mance. In any event, a meaningful jury trial requires not merely that the jury is given a legally accurate direction, but that it is one which they can make sense of in practice and apply with confidence to the evidence they have heard. I am not convinced that Parliament can have intended that a jury should be expected to understand and apply the sophisticated distinctions which Lord Sumption seeks to draw” .
Lord Reed was also troubled by the potential impact of the decision on “numerous public agencies with responsibilities relating to the prevention or detection of crime” and on “other activities of other persons such as investigative journalists”, citing the activities of the late Robert Maxwell and the fact that the 1997 Act had already been used in the context of journalism: Thomas v News Group Newspapers Ltd  EMLR 4). He did “not demur from the view that it may be desirable that the courts should be able to restrain the activities of a person who causes real distress through his irrational behavior … But that is not in my view a sufficient reason for extending the scope of the Act beyond what Parliament intended”.
The decision of the majority in this case is superficially attractive given the outrageous behavior of the defendant. It is instructive to reflect, however, that the 1997 Act was primarily directed at classic stalking behavior. One of the specific examples which was an influential trigger for the legislation was the acquittal of Charles Wilson who had stalked a woman for more than two years, spying on her with binoculars, photographing her and sending her letters which referred to kidnapping. He regularly turned up at her workplace, the supermarket she used and her mother’s home, drove a van fitted with spying equipment and had been described as “a danger” to the woman at a previous hearing. As was pointed out in the Introduction to the 2011 Consultation Paper on stalking (which preceded amendments to the 1997 Act to introduce specific provisions on stalking), the 1997 Act had “followed a series of high profile stalking cases in 1995/1996” and a consultation paper which was “focused on ‘stalking’ and [in which] the perpetrator was referred throughout as a ‘stalker’” (“Consultation on Stalking”, November 2011, Introduction, referring to the Home Office Consultation Paper “Stalking – The Solutions”, of July 1996). The Consultation Paper had stated that
“The Government proposes to deal with the menace of stalking through a combination of civil and criminal measures’ (§3.4)”.
But because an earlier Private Members Bill which attempted to define stalking
“was deemed flawed precisely because its scope would have been too narrow [t]he then government felt that the solution was instead to focus on the harm caused to the victim rather than the constitutive elements of ‘stalking’. The Bill was therefore drafted to cover ‘many incidents of harassment, as well as its primary aim of dealing with stalking’” (emphasis added).
The application of the 1997 Act to classic stalking proved inadequate and the Act was amended in 2012 to include a specific prohibition on “stalking” (s2A). Meanwhile it had been applied to the conduct of British Gas in pursuing a debt wrongly believed to be owed by a former customer (Ferguson v British Gas Trading Ltd  All ER (D) 80 (Feb)); that of a local authority which had brought a number of separate proceedings in respect of rent arrears (Allen v Southwark London Borough Council  All ER (D) 113 (Nov)) and to alleged bullying at work (Majrowski v Guy’s and St Thomas’s NHS Trust  1 AC 224). This all being the case, it is difficult to avoid the conclusion that the decision of the Supreme Court is the latest example of the stretch of the 1997 Act significantly beyond its intended targets.