warwick.ac.uk/lib-publications Original citation: Sorell, Tom. (2016) Online grooming and preventive justice. Criminal Law and Philosophy. Permanent WRAP URL: http://wrap.warwick.ac.uk/79478 Copyright and reuse: The Warwick Research Archive Portal (WRAP) makes this work by researchers of the University of Warwick available open access under the following conditions. Copyright © and all moral rights to the version of the paper presented here belong to the individual author(s) and/or other copyright owners. To the extent reasonable and practicable the material made available in WRAP has been checked for eligibility before being made available. Copies of full items can be used for personal research or study, educational, or not-for profit purposes without prior permission or charge. Provided that the authors, title and full bibliographic details are credited, a hyperlink and/or URL is given for the original metadata page and the content is not changed in any way. Publisher’s statement: “The final publication is available at Springer via http://dx.doi.org/10.1007/s11572-016- 9401-x “ A note on versions: The version presented here may differ from the published version or, version of record, if you wish to cite this item you are advised to consult the publisher’s version. Please see the ‘permanent WRAP URL’ above for details on accessing the published version and note that access may require a subscription. For more information, please contact the WRAP Team at: wrap@warwick.ac.uk http://go.warwick.ac.uk/lib-publications http://go.warwick.ac.uk/lib-publications http://wrap.warwick.ac.uk/79478 http://dx.doi.org/10.1007/s11572-016-9401-x http://dx.doi.org/10.1007/s11572-016-9401-x mailto:wrap@warwick.ac.uk 1 On-Line Grooming and Preventive Justice Tom Sorell Interdisciplinary Ethics Research Group Department of Politics and International Studies University of Warwick Coventry CV4 7AL +44 (0) 2476 528031 T.Sorell@warwick.ac.uk Abstract: In England and Wales, Section 15 of the Sexual Offences Act (2003) criminalizes the act of meeting a child –someone under 16—after grooming. The question to be pursued in this paper is whether grooming –I confine myself to online grooming—is justly criminalized. I shall argue that it is. One line of thought will be indirect. I shall first try to rebut a general argument against the criminalization of acts that are preparatory to the commission of serious offences. Grooming is one such act, but there are others, sometimes associated with terrorism. According to me, the general argument misapplies certain considerations about autonomy that are alleged to be in force in other areas of criminal law. Contrary to that general argument, criminalization of preparatory acts does not, in general, bypass the agency of citizens. Moreover, the criminalization of preparatory acts can disrupt activity that would have led to very serious crime, and with relatively low costs to the perpetrators, costs that reflect the non-occurrence of the more serious crime. There is evidence that grooming is harmful in itself, and so another point against the general argument is its assumption that preparatory offences are often harmless or at least victimless. There are objections to some of the undercover policing techniques that lead to a Section 15 prosecution, but these objections are not all weighty. Keywords: Anti-Paedophile policing Preparatory offences Sexual Grooming Sexual Offences Act (2003) Undercover policing Preventive justice mailto:T.Sorell@warwick.ac.uk 2 Grooming occurs when an adult prepares someone –usually a child—for sexual contact. This behaviour is criminalized in many jurisdictions, and it is one of a range of offences that have been called into question by critics of preventive justice.1 In England and Wales, Section 15 of the Sexual Offences Act (2003) criminalizes the act of meeting a child –someone under 16—after grooming. The offender must be over 18, and must have met or communicated with the child at least twice. In addition, the offender must have intended to meet the child and travelled to the meeting intending to commit one of a range of offences defined in Part One of the Act. These range from outright rape to a variety of sexual assaults. Part One distinguishes between, on the one hand, rape and assaults against children under 13, and cases where the victims are older but under 16. Section 15 has some distinctive features. For example, it applies to international travel, some part of which is within the domestic jurisdiction. This means that if someone in London who was engaged in the online grooming of a child in e.g. Thailand or America flew out from England or Wales to meet that person, intending to engage in a sort of sexual contact outlawed in England and Wales, that traveller would still be liable to prosecution on returning to England and Wales. Offences under section 15 are punishable by a maximum of 10 years imprisonment on indictment. The question to be pursued in this paper is whether grooming –I confine myself to online grooming— is justly criminalized as in Section 15. I shall argue that it is. One line of thought will be indirect. I shall first try to rebut a general argument against the criminalization of acts that are preparatory to the commission of serious offences. Grooming is one such act, but there are others, sometimes associated with terrorism. According to me, the general argument misapplies certain considerations about autonomy that are alleged to be in force in other areas of criminal law. Contrary to that general argument, criminalization of preparatory acts does not, in general, bypass the agency of citizens. Moreover, the criminalization of preparatory acts can disrupt activity that would have led to very serious crime, and with relatively low costs to the perpetrators, costs that reflect the non-occurrence of the more serious crime. There is evidence that grooming is harmful in itself, and so another point against the general argument is its assumption that preparatory offences are often harmless or at least victimless. The rest of this paper falls into four parts. In the first, I introduce the Section 15 offence and indicate the number of prosecutions that have been brought under it since 2006. Next, I expound and criticize the general argument against criminalizing preparatory offences. I go on in the third section to give reasons why the approach to grooming in section 15 of the Sexual Offences Act in particular is justifiable. First, it effectively distinguishes between “contact” groomers and “fantasy” groomers, 2 removing doubt about grooming offences based on the possible remoteness of grooming behaviour from actual sexual contact. Second, s.15 distinguishes between, on the one hand, the sexualization of online behaviour between children, and, on the other, grooming by forensically aware adult paedophiles who sometimes pose as sexualized children. In the final section, I discuss the justifiability of two kinds of undercover operations that are used to collect evidence for prosecutions of Section 15 offences. I argue that while the two kinds of operations are morally different, and one is normally more objectionable than the other, both are often morally justifiable on balance. 1 A Ashworth and L Zedner, Preventive Justice (Oxford: Oxford University Press, 2014), p.99. Daniel Ohana, ‘Responding to acts preparatory to the commission of a crime: Criminalization or prevention?’ Criminal Justice Ethics 25 (2006) pp. 23- 39. See also Petter Asp, ‘Preventionism and Criminalization of Nonconsummate Offences’ in A Ashworth, L Zedner, and P. Tomlin, eds. Prevention and the Limits of Criminal Law (Oxford: Oxford University Press, 2013), p.23; p. 25; p. 28. 2 Stephen Webster et al, European Online Grooming Project, Final Report, 2012: http://www.europeanonlinegroomingproject.com/media/2076/european-online-grooming-project-final-report.pdf 3 1.Meeting a Child after Grooming Section 15 is in keeping with, if not motivated by, the following line of thought about serious crime: The more serious a crime –understood in relation to a scale of harm and culpability—the more urgent it is to prevent it. Prevention can sometimes take the form of criminalizing component offences, or criminalizing steps that, while apparently harmless in isolation, characteristically lead to the intentional commission of the serious crime. The penalty for the commission of the component or preparatory offences is of course lighter, sometimes much lighter, than the penalty for the full blown offence (2 years in prison is the starting point for a section 15 offence against a child over 13 but under 16), so that defendants convicted of them do not suffer disproportionately. More importantly, the availability of these offences allows charges to be brought against suspects before serious harm can be done. Even if the proceedings arising from these charges produce no conviction, bringing charges can itself disrupt a course of action that might have led to serious harm and that would have been more difficult to disrupt if it had been allowed to run on. Here ‘serious crime’ refers to a Part One offence, such as rape or assault of someone under 13 or 16: if the criminalization of grooming prevents this sort of crime, then, according to the line of thought just articulated, it is justified. When associated with the internet, Section 15 prosecutions usually involve a record of online chat between the suspect and a child. Online chat rooms can be visited speculatively by groomers, or visits there can follow contact on social networking websites which contain very detailed profiles and sometimes photos of children who become grooming targets. Chat room contact typically begins with introductions and an indication of ages and gender. Names –usually pseudonyms—will be exchanged. Then a process will begin in which the suspect will gather information about the child and show interest or admiration. Along the way, mobile telephone numbers (or even email and home addresses) will have been exchanged. Online chat is orchestrated by the groomer in such a way that it eventually turns to the discussion of sex. The groomer will often send pornography to the child, and prompt sexualized text or chat room messages. Sometimes the suspect will be interested in the sexual experience of the child, and will offer to teach the child sexual techniques. At other times the suspect will induce the child to produce sexualized webcam or smartphone images of herself. (The vast majority of grooming targets are female.) Then a meeting may be proposed, sometimes at a place that the child suggests. The chat record may contain evidence of an agreement to engage in sexual contact, in which case the makings of a Section 15 prosecution are in place, other things being equal. The chat record may also reveal evidence of the suspect’s awareness of the risks of arrest should the plans for the meeting be discovered by parents or come to the attention of the authorities. Arrests on Section 15 charges are sometimes made after parents discover that their child is engaged in sexualized online conversation. The police are informed, and sometimes, with the appropriate authorizations under the Regulation of Investigatory Powers Act (2000) (RIPA), undercover officers will assume the online identity of the child until the arrangement to meet is proposed by the groomer. Police will then arrest the suspect at the proposed meeting place. While the suspect is being questioned, the police will carry out an authorized search of the suspect’s home address and seize computers. Subsequent investigation sometimes reveals downloaded indecent images of children, and 4 sometimes even evidence of participation in online paedophile groups. In such cases further charges can be brought. Section 15 arrests can also be made on the basis of the action of online vigilantes. These are civilians who pose as children on social media sites and cultivate approaches for sex from groomers. Like undercover police, they will carry on a protracted online conversation with a suspect, agree to a meeting, and then, when the suspect arrives, confront him on camera. The film is subsequently posted online as an act of exposing a paedophile. The police in England and Wales sometimes act on information provided by the vigilantes, but the guidance from the Department of Public Prosecutions (DPP) on the use of this material points to difficulties with vigilante operations in relation to RIPA,3 and a Memorandum of Understanding between the Crown Prosecution Service and the Association of Chief Police Officers is critical of vigilante operations, and excludes certain defences by vigilantes against charges for making indecent images, even if that was done to gain the trust of a suspect in the course of a “sting” operation. McGuire and Dowling have summarized proceedings under Section 15 between 2006 and 2012:4 Table 1. Section 15 proceedings by year and gender of victim, 2006-2012 Table 1 is not confined to cases of proceedings launched against suspects after online grooming alone, and many more grooming offences were recorded by police than were proceeded against:5 3 http://www.cps.gov.uk/legal/p_to_r/rape_and_sexual_offences/soa_2003_and_soa_1956/#a29 4 M McGuire and S Dowling, Cyber-crime: A Review of the Evidence (London: HMSO, 2013), ch. 3. This table makes confusing reading. It will quickly be noticed that the number of those sentenced in a given year can exceed the number proceeded against in that year. This because cases can carry over from one year to the next, with sentencing in a year later than proceedings started. 5 McGuire and Dowling, op. cit. p. 8 5 Table 2. Number of recorded grooming offences by year in England and Wales, 2004-2013. Tables 1 and 2 show that the number of grooming offences is relatively low,6 that the number of prosecutions is lower still, but that the conviction rate for those prosecuted is high. There are much larger numbers of recorded offences of possessing, making and distributing indecent images compared to grooming offences. For example, charges of making an indecent photograph of a child were brought in over 14,000 cases in 2012-13, 7 dwarfing the numbers of Section 15 proceedings. 2. Section 15 and general objections against preparatory offences What, if anything, is morally wrong with Section 15? It would certainly be criticizable if it incorporated stereotypes about grooming that are not borne out by relevant empirical research. It would equally be open to criticism if prosecutions under it regularly failed. I shall consider the criticism that it carries risks of injustice associated with the criminalization of preparatory offences. Ashworth and Zedner have developed this sort of criticism.8 They include section 15 among preparatory or pre-inchoate offences, one of 9 categories of preventive offences they recognize. Preparatory and pre-inchoate offences include heavily criticized provisions of the UK Terrorism Acts. Ashworth and Zedner argue against the criminalization of acts in this category from two directions: they (a) try to undercut some of the general defences of their criminalization on the basis of the harm principle, and (b) introduce and justify “restraining principles” for preparatory offences. I argue that neither approach is entirely successful against the whole range of preparatory offences and that, in particular, section 15 survives both kinds of criticism. The reasoning behind the criminalization of preparatory acts might be put as follows. “If a type of act A is very harmful, say because it produces fatal or life-threatening injury on a large scale, then the prevention of a type-A act is justified, other things being equal. The prevention of a type-A act is a matter of the prevention or discouragement of a type-B act, where a type-B act characteristically precedes and facilitates a type-A act. Where type-A is a co-ordinated suicide-bomb attack, type-B acts might include online or offline meetings to recruit possible attackers; online or offline meetings to plan an attack; purchases of bomb- making materials; communications with people known by the police to have bomb-making expertise; and so on. To the extent that there is evidence that a Type-B act is likely to contribute to a Type-A act, the harm caused by the Type-A act justifies the criminalization of the Type-B act.” 9 6 But the corresponding American figures are much higher. In 2006 there were more than 6,300 reports of online “enticement” of children, up from 2600 the previous year. See K Raymond Choo, Online Child Grooming: A Literature Review on the Misuse of Social Networking Sites for Grooming Children for Sexual Offences grooming( Canberra: Australian Institute of Criminology, 2007) p.25. 7 Under Protection of Children Act 1978, Ss.1(1)(a) and 6. See McGuire and Dowling, p.15 8 A Ashworth and L Zedner Preventive Justice (Oxford: Oxford University Press, 2014) p. 110 9 This reasoning is in keeping with what Simester and von Hirsch (Crimes, Harms and Wrongs (Oxford: Hart, 2011) call the Standard Harm Analysis, even though instances of Type A and Type B acts may be temporally remote from one another. 6 The most obvious problem with this reasoning is that type-B acts can contribute not only to Type-A acts but also to perfectly harmless Type-C acts. For example, the purchase of large amounts of fertilizer can contribute to bomb-making for a terrorist attack, but it can also contribute to a perfectly harmless, or indeed beneficial, agricultural enterprise. Ashworth’s and Zedner’s argument against preparatory offences is partly to do with the fact that perfectly harmless and ordinary behaviour as well as acts leading to terrorism can be eligible for charges; in addition, Ashworth and Zedner give weight to the chilling effect: people being put off harmless and legal behaviour leading to Type-C acts for fear of being taken by the authorities to be preparing Type-A acts. In general, Ashworth and Zedner hold that preparatory offences limit the liberties of everyone in some ways, and disproportionately limit the liberties of the few –for example, those fitting discriminatory stereotypes that might be implicit in profiles of potential terrorists. 10 More generally, Ashworth and Zedner criticize arguments from harm for what they regard as their utilitarian bias. The arguments may look irresistible when rightness or wrongness itself is a function of harm and benefit, but liberty is valuable, too, even when it is potentially harmful, and the argument from harm for preparatory offences arguably fails to give liberty enough weight.11 The force of this sort of argument varies with the range of preparatory offences associated with the production of serious harm. Ashworth and Zedner do not show that it is unjust for the police ever to take an interest in anyone who buys large quantities of fertiliser. There are good reasons for the police to look into such a purchase, notwithstanding the innocent uses of large quantities of fertiliser. Similarly with other ingredients of explosives. In UK terrorism legislation, however, many things other than purchasing bomb-making equipment are taken to facilitate violent offences, including some kinds of non-disclosure, and some very vaguely specified kinds of possession. The looser the association of these with the preparation of a terrorist attack, the more the limitation of liberty involved in their criminalization seems disproportionate. Only when the association between criminalized acts and the harm of terrorism is very loose does the Ashworth-Zedner position acquire real force.12 And even in these cases their position may be overdrawn relative to the practice of police and prosecutors. For the fact that certain sections of the Terrorism Act (e.g, sections 1 and 75) are arguably too vague, or set the threshold for contributing to a terrorist act too low, needs to be taken together with the fact that that the authorities have brought few or no prosecutions under them.13 Again, the fact that a lot of behaviour criminalized under the Terrorism Act (2000) can easily turn out to be harmless means that the burden of showing in a particular case that the behaviour is harmful is not a trivial one. The burden is made even harder to discharge by the fact that people who conspire to prepare acts of terrorism are often forensically aware and, as a result, conduct their electronically 10 Ibid p. 104 11 Ibid. 12 See also Simester and von Hirsch, op.cit. ch. 13. 13 According to the most recent (2014) statistics (https://www.gov.uk/government/publications/operation-of-police-powers- under-the-terrorism-act-2000-financial-year-ending-march-2014/operation-of-police-powers-under-the-terrorism-act-2000- and-subsequent-legislation-arrests-outcomes-and-stops-and-searches-great-britain-financial ), “of the 460 persons charged under terrorism legislation since 11 September 2001 the main principal charges related to: possession of an article for terrorist purposes (16% of such charges);preparation for terrorist acts (16%);failing to comply with duty at a port or border controls (16%);collection of information useful for an act of terrorism (10%);fundraising (10%).” In this respect the independence of the Dept. of Public Prosecutions counteracts remnants in Terrorism legislation of populist provisions whose introduction was more a matter of pose or gesture than of considerations of practicability. 7 intercepted conversations in coded vocabulary. These facts translate into obstacles in the way of successful prosecutions under the Terrorism Acts, and so obstacles to the loss of liberty. Even when charges are brought against suspects for preparatory offences, clear evidence of an intention to mount a lethal attack may be hard to adduce. Does this fact show that preparatory acts should not be criminalized in the first place? No, and neither does it show that criminalization cannot help with prevention. After all, with the preparatory offence in place, the mere act of bringing charges under it may be enough to disrupt the preparation of a terrorist act by someone against whom a successful prosecution cannot be brought, and the charges can communicate to all concerned that certain people have come to police attention and that their future participation in harm-producing conspiracies would probably attract police attention, too. This message may well set back the plans of terrorist conspirators, and it can be conveyed whether or not a subsequent prosecution fails. In this way criminalization aids disruption but not necessarily disruption that deprives anyone –including anyone innocent-- of their freedom. It is not obvious, in fact, that any harm is necessary inflicted on anyone by disruption without arrest. (On the other hand, disruption that includes arrest can be harmful to the extent that arrest by itself is stigmatizing.) If the Ashworth-Zedner criticism of appeals to harm by defenders of the criminalization of preparatory offences is less than conclusive, what are we to say of their arguments for restraining principles for preparatory offences? One restraining principle is that acts that are criminalized not be too remote from the production of harm.14 But how remote is too remote? Some preparatory acts might involve no intention to produce harm oneself at all, as when one publicly approves of some past terrorist act without any encouragement to anyone to follow its example. But suppose that someone actually produces a working bomb with the intention of detonating it as an act of jihad, but then has a change of heart and abandons the jihadist project. Should the production of the bomb still attract a conviction under section 5 of the Terrorism Act (2006) if charges are brought after the change of heart? And should we say that where there is a possibility of a change of heart that criminalization is out of order and only starts to be in order when a point of no return has been reached? Ashworth and Zedner seem to answer this last question in the affirmative: It is generally accepted that punishment for mere thoughts is objectionable, since it invades the individual’s private world and allows liability to turn on ideas that the individual has not gone so far as to bring into the external world. Moreover, individuals should be treated as capable of changing their mind and conforming to the criminal law, ‘because this is what it is to respect him as a responsible agent’. This may be a sufficient reason for ruling out ‘any overt act’ as an appropriate test, and also some of the ‘substantial steps’ set out in the Model Penal Code which involve preparatory acts such as reconnoitring a location. What is wrong with those tests is that they criminalize the individual as an attempter when the distance to the substantive offence is so great that a change of mind remains a possibility.15 This passage suggests that the point of no return occurs where an agent becomes liable to an attempt charge, and it suggests further that criminalizing a preparatory act in effect pushes the point of no return earlier in time, at the cost of not treating the agent as responsible. But this line of thought is unsound. First, what is at issue is whether it is justifiable to criminalize steps in a harm-producing process, if the steps come before the attempted full blown crime. Even if 14 Ashworth and Zedner, p. 109 15 Ibid. p.110 8 the criminalization of very early steps is not justified, this is not because agents who go through with the early steps could still change their mind about the later steps.16 It is because the type of offence constituted by taking an early step is usually very far from producing harm. But what about steps later on in the process? An intermediate step –such as the act of making a working bomb—has already significantly increased the probability of harm. For example, a bomb that is not entirely competently made can become unstable and explode even if the maker does not want it to. If that happens, then even if the maker had abandoned the plan of detonating it in a terrorist attack, he had gone beyond the point of no return with respect to the threshold for harmful effects, and he is responsible for producing the unstable but as yet undetonated bomb. Arguably, he had reached the threshold for doing serious harm even earlier, when he assembled bomb-making ingredients separately. If so, then it makes sense to criminalize arrival at the lower threshold for doing significant harm in a way it does not make sense to criminalize expressions of support for past terrorist acts. Reaching the threshold for doing significant harm is what matters, notwithstanding the distance between reaching this threshold and outright attempt, or the distance between reaching this threshold and the fully realized result of a terrorist attack. Now if a threshold for significant harm can be reached by a preparatory offence like assembling bomb materials, and if an agent knows that preparatory acts have been criminalized, then the law does not treat the agent as irresponsible by treating steps earlier than an attempt as points of no return for less serious offences than attempted mass-murder. For the agent presumably has the same opportunity for changing his mind about putting together the fertilizer and peroxide as he has, later on, for changing his mind about detonating the finished bomb. The line of thought developed by Ashworth and Zedner only reaches their preferred conclusion if all steps before attempt are harmless. Again, it is quite disputable to say, as Ashworth and Zedner do, that treating someone as a responsible agent has some special connection with making allowances for last-minute changes of mind. For the most serious offences, that is, the kind of offences that are most likely to produce life- threatening injury or death, or life-threatening injury and death on a large scale, it makes sense for the law to say not only ‘Don’t commit those offences –or else!’ but ‘Don’t even go there!’ or ‘Don’t even think about it! --or else.’ Now this sort of injunction does not literally make a crime of something that goes on in an agent’s mind: it does not create a thought-crime in the strict sense. ‘Don’t even think about it!—or else’ does not mean ‘Don’t even conjure up the mental image of a terrorist attack!—or else’. It means: ‘Don’t take so much as the first step toward a terrorist attack--or else’ –precisely the message of the laws that criminalize preparatory acts. When the injunctions expressed by the criminalization of preparatory acts are understood in this way, there is still room for dispute as to whether certain criminalized acts are in fact even first steps. This is the sort of dispute that might break out over the offence of “glorifying terrorism” in s.1. of the Terrorism Act (2000). But for the bomb-making step in the process of preparing a terrorist attack (which seems to engage with s.5 of the Terrorism Act (2000)), this dispute does not seem to me to arise, and it does not seem to treat an agent as irresponsible for the law to insert into their deliberations that they face a certain term of imprisonment for so much as assembling a bomb. Similarly for meeting a child for sex after grooming, as I shall argue later. 16 Pace L. Alexander and K Ferzan, Crime and Culpability (Cambridge: Cambridge University Press, 2009), p. 197. 9 At this point, sceptics about preventive justice may have to turn to fundamental differences in “logic” between punishment and prevention.17 Punishment takes account of an agent’s past law-breaking behaviour and imposes hard treatment proportionate to that behaviour. But where a measure is preventive, a rather different logic seems to be entailed— that the type and quantum of the measure imposed must be those judged necessary to bring about a significant reduction in the risk of future harm, irrespective of whether those measures are proportionate to any past or present conduct of the subject. ... [W]here the rationale is desert, the punishment must censure the subject in a way and to an extent that respects his or her responsible agency, and the role of prevention is to supply underlying deterrence whereas the choice or quantum of the sanction should be governed by proportionality considerations. Where prevention is the rationale its logic applies without respect for whether the subject is a responsible agent or not, since the purpose is to obtain the optimal preventive outcome.18 But this line of thought is highly disputable. It implies that proportionality is at home with punishment but not prevention, and it is easy to think of counterexamples. Thus, it is customary for department stores to prevent shoplifting by CCTV surveillance, by alarm-activating devices attached to goods, and by posting unarmed security guards at exits equipped with alarms. Preventing shoplifting by shooting shoplifters at exits might be more effective, but there is nothing in the logic of prevention that entails that shoplifting be prevented at all costs, or that shooting be preferred to arrest by unarmed guards when both are effective. Preparatory offences are proportionately less severely punished than consummated offences. If the penalties reflected the alleged logic of prevention, the penalties would be whatever was sufficient to discourage taking first steps toward the consummated offence. Sufficiency might be associated with much more severe penalties than are actually threatened or applied in relation to preparatory offences. Again, preparatory offences work with police strategies of disrupting the plans of those embarked on serious crime. This is not a matter of as it were paralysing the planners of serious crime, but of bringing charges so as to induce the thought that continuing with those plans will be futile because the plans are no longer secret. The “different-logics” claim associates proportionate punishment for individuals after prosecution with respecting persons, and associates prevention with treating people indiscriminately –regardless of their past conduct—and treating them as appropriate sites for whatever discouraging pressures produce omissions of unwanted behaviour. This sharp contrast between prevention and punishment makes no allowance for the way that the threat of punishment works in laws against consummated offences. For example, the law against homicide is addressed to everyone in a jurisdiction— irrespective of their past or present behaviour. If ignoring past behaviour is objectionable in the criminalization of preparatory acts, why is it not also objectionable in the threats of punishment carried by laws whose prosecution starts after the fact? In laws against homicide, for example, the threat of hard treatment in the form of a long term of imprisonment is addressed to those who would not dream of killing anyone, as well as those who would. Why does not this indiscriminate form of address treat too many people as dangerous if preparatory offences do? Again, why is not the threat of severe treatment for homicide also open to the objection that it uses fear of imprisonment to deter people rather than thoughts about the costs to people of being killed? I think the Ashworth-Zedner argument from the different logics of prevention and punishment does imply—controversially if not 17 This paragraph and the next are drawn from my unpublished ‘The Scope of Serious Crime and the Justifiability of its Prevention’ 18 Ashworth and Zedner, p.18 10 simply incorrectly-- that the threat in the law against homicide is objectionable for both its indiscriminateness and its eliciting demotivating fear rather than demotivating reasons --thoughts about the badness of being killed. Suppose the prospect of hard treatment is the only thing that keeps a particular person K from killing someone else. In that case the agent is –in Kantian terms--only pathologically driven, not governed by practical reason. Fear of imprisonment rather than an understanding of the reasons why homicide is wrong operates to deter K from homicide. Although he is not a typical agent, K is still a suitable addressee of a law against homicide. This made sense to Kant because he thought people in general were hybrids of pathology and autonomy, and that juridical law added engagement with pathology to the moral law against killing, which only engages with reason. Typically the juridical law as Kant conceives it addresses both elements of the hybrid, providing in Kant’s terms “pathological” or inclination-based motivation in addition to moral motivation for not killing. In other words, there are reasons for not killing that most agents can understand and accept, and these can motivate alongside the threat of hard treatment, or independently of it, to produce compliance with the law against homicide. It is not as if the pathology-engaging part of the law –the “or else”--denies or cancels out the agency of most addressees, that is, agency in the form of receptiveness to moral reasons. The mere fact that a threat is made by the law does not imply that the law treats an agent as solely pathologically motivated –impervious to the other reasons against homicide. Not everything in the law has to treat its addressees as (fully) responsible, according to Kant. The Ashworth-Zedner position seems incorrectly to deny this. Writers other than Ashworth and Zedner, including Simester,19 seem to require that the law, in Waldron’s phrase, should not be “brute” at the same time as it is mandatory. It should treat its addressees as fully responsible, which, according to Simester, it can fail to do in the case of preparatory offences. I have just suggested that the law need not treat all of its addressees as fully responsible in the sense of always being open to the force of strong reasons for aborting a course of action leading to a consummated offence. But in the case of preparatory offences where what Simester calls a “nexus” condition is met –the otherwise harmless proscribed thing is done by an agent with an intention to do harm20-- the agent is responsible for the preparatory behaviour, and the behaviour is not innocent in the sense of not intended to be harmful. Simester asks how justification for the criminalization of preparatory acts can rule out the criminalization of an act of eating cereal that serves to nourish an intending murderer. I think the answer may be to develop the nexus requirement more fully. Unless eating cereal instantiates a type that is mentioned by a statement of a causal regularity leading to a type of serious harm, it does not meet a nexus requirement. To illustrate, it might be true that eating cereal is nourishing and that being nourished helps agents carry out their plans, but that is not a causal regularity that leads to harm, since being helped to carry out a plan is not a type of harm. Again, it is not always true that eating cereal helps intending murderers to carry out murderous plans just because eating cereal helps people to carry out their plans. Someone who is hungry and is planning a murder is not necessarily in need of food in order to commit the murder. He may need food anyway. The nexus requirement, suitably nuanced, may be the key to keeping the justification for preparatory offences from becoming too permissive. 19 “Prophylactic Offences” in G.R. Sullivan and I Dennis, Seeking Security: Pre-empting the Commission of Criminal Harms (Oxford: Hart, 2012), p. 70 20 Ibid. p. 67f. 11 3. The appropriate threshold for a preparatory grooming offence Ashworth and Zedner include Section 15 offences among the targets of their onslaught on the criminalization of preparatory acts. But is Section 15 comparable to problematic provisions in counter-terrorist legislation? For example, does it criminalize a type of preparatory act that occurs too early in the typical stages of the process leading from the online grooming of a child to sexual assault? There are two aspects to this question. First, it implies, perhaps objectionably, that harm is only inflicted at the stage of sexual assault, and not in the grooming. Later, I dispute this. But beforehand I consider whether the Section 15 preparatory offence justifiably focuses on the meeting stage of the process of grooming for sex. To sharpen this question, let us consider the (in England and Wales) fictional offence of sexualizing an online conversation with someone believed to be under 16. Suppose that this offence is committed if and only if (i) sexual content is introduced into an online conversation with a child; (ii) the conversation partner introducing the sexual content is over 18; and (iii) the sexual content might reasonably be believed to express an intention to meet a child after grooming. Someone whose behaviour meets these conditions might not yet have proposed a meeting in so many words, but might have behaved in a way which a reasonable person would believe was putting sex on the agenda in an ongoing online relationship with a child.21 If there were an offence of sexualizing an online conversation, then that would criminalize a step in the process leading to a sexual assault that occurred earlier than meeting a child after grooming. This might qualify as a suspect precursor offence for the same reason that some precursor offences in the case of terrorism are suspect: namely, the looseness of the connection between that kind of offence and outright sexual assault: that is, although sexualized conversation may raise the probability of a sexual assault, it is causally insufficient for such an assault. Would the fictitious offence of sexualizing an online conversation with a child be justified, then, if incorporated into the Sexual Offences Act (2003)? The looseness-of-the-connection point is one reason to think not, but there are also empirical reasons. Recent and influential empirical work on sex crime distinguishes between contact and fantasy groomers among those who are operating in online chat rooms.22 According to this work, there are some chat line groomers who are content with sexual fantasies generated by sexualized conversations. These groomers do not wish or intend to meet their conversation partners: the fantasies fuelled by conversation with children by themselves provide material for auto-erotic gratification. For these groomers, then, the sexualization of online content is not a preparation for sexual contact, let alone sexual assault. The existence of fantasy groomers constitutes an objection to a preparatory offence of sexualization of online conversation with a child. It may even count against an offence of sexualizing conversation with children, when the makeup of the group at the receiving end of these conversations is taken into account. For child targets of online grooming are themselves not all of one kind. According to the taxonomy developed by the European Online Grooming Project,23 there is a “resilient” child internet user, and this type predominates: 21 My imaginary offence of sexualizing a conversation seems to resemble section 474.27 of the Australian Criminal Code Act (1995), which carries a penalty of 12 years imprisonment. See Choo, op.cit. p. 4 22 Peter Briggs, Walter T. Simon, and Stacy Simonsen ‘An Exploratory Study of Internet-Initiated Sexual Offenses and the Chat Room Sex Offender: Has the Internet Enabled a New Typology of Sex Offender?’ Sex Abuse March 2011 23: 72-91 23 Stephen Webster et al, European Online Grooming Project, Final Report, 2012: 12 the greater proportion of young people approached [in our research study] were resilient and refused to engage when they sensed the person they were talking to was ‘weird’. For example, some young people were described as abruptly ending initial interactions by: cutting-off the online link when the groomer started a sexual conversation; blocking the groomer on the site; and/or hanging up the mobile phone.24 If a large number of children interacting online with others can recognize sexualization as a danger signal, and are able safely to discontinue contact with groomers once sexualization of conversation begins; if they are able to shrug off that online experience, then perhaps the sexualization offence we have invented is not properly regarded as an offence leading to sexual contact, or even as harmful enough to criminalize. Of course, the “resilient” child group could in the future shrink as a proportion of the total number of children using online chat rooms. If the proportion of “resilient” children in chat rooms became tiny, and the proportion of very isolated and vulnerable children was much higher than research indicates it currently is, the argument from harm for the sexualization offence might be strengthened. On balance, however, the case for a preparatory offence of sexualization of online conversation does not look strong. Do problems with that offence infect other preparatory offences, including Section 15? The answer is a clear ‘No’. First, Section 15 has a far higher threshold for charges than the sexualization offence. Section 15 requires (i) a minimum level (2 occasions at least) of contact between someone over 18 and a child; (ii) evidence of a plan on the part of the person over 18 to meet the child; and (iii) actual travel toward the agreed meeting point by the person over 18 after that person has sexualized the conversation and expressed a desire for sex with the child. Behaviour that satisfies these conditions indicates that grooming as opposed to some other behaviour is being displayed.25 Section 15, then, cannot be criticized on the same grounds as the sexualization offence. Not only does it exclude fantasy groomers, but it requires persistence in online contact, typically, sexualized contact. Nor, in my view, can Section 15 be criticized as excessive just because it is a pre-attempt offence. Here an analogy with pre-attempt terrorism offences may help. We agreed in the last section that some pre-attempt offences raise the probability of harm significantly, while others are only very loosely associated with harm, if associated with harm at all. Behaviour fitting the Section 15 offence comes closer to preparatory offences in the case of terrorism that raise the probability of harm e.g. part-assembling a bomb, than those that by themselves are only loosely connected with harm e.g. an act of “glorifying” terrorism. In other words, meeting a child after grooming comes closer to being harmful in its own right than to being only loosely connected with harm. To go back to the argument of the last section, although assembling materials for explosives is perhaps not sufficient for an attempted terrorist attack, it does reach a threshold for being a dangerous act, in that assembling explosives before detonation could itself be lead to explosion. For some groomers, physical proximity to a child will increase the probability of an assault much more than online contact, even when webcams and message-exchanges in real time bring virtual meetings much closer as experiences to physical meetings. So taking the step of achieving physical proximity can by itself be very dangerous, which creates a prima facie case for criminalization. http://www.europeanonlinegroomingproject.com/media/2076/european-online-grooming-project-final-report.pdf 24 European Online Grooming Project, p. 100 25 On arrest for a Section 15 offence, the groomer will be searched, and if he is found in possession of condoms or lubricating jelly, those items as well as the record of intercepted conversations with the child will be treated as evidence of the intention to go through with sex, and therefore (given the age of the child) sexual assault, in the course of a meeting. 13 Aiming at physical proximity distinguishes fantasy groomers from contact groomers, and some contact groomers can be, in the taxonomy of the European Grooming project, “hyper-sexualized”: The hyper-sexualised group of men were characterised by extensive indecent image collections of children and significant online contact with other sexual offenders or offender groups. Some men in this group also had significant collections of extreme adult pornography. They adopted different identities altogether, or had an identity picture that was not of their face but of their genitals. Their contacts with young people were highly sexualised and escalated very quickly. 26 This is the group that the Section 15 offence seems to fit best, especially if members of that group succeed in arranging a meeting with children who, rather than being “resilient,” are, in the taxonomy of the European Grooming Project, either “vulnerable” or “risk-taking”. Table 3 summarizes the Table 3. Characteristics of Vulnerable and Risk-Taking Victims of Online Grooming characteristics of vulnerable and risk-taking child users of internet.27 Vulnerable users tend to be more isolated and sometimes use childlike chat room avatars. Contact grooming is a process with characteristic stages, and an early stage is scanning the online environment for displays of characteristics associated with either vulnerable or risk-taking children.28 Online groomers are good at identifying children who turn to the internet for a social life and affection that elude them offline. These fit the profile of “vulnerable” children. Groomers will manipulate these children by expressing sympathy and admiration in turn. Risk-taking children, by contrast, will overestimate their ability to control or hold their own in online sexualized conversations, and may experiment with flirtatious behaviour that can attract the hyper-sexualized groomer. Section 15 would be criticizable if (i) it were appropriate only for hyper-sexualized groomers of vulnerable and risk-taking children but (ii) these turned out individually and in combination to make up only an insignificant proportion of actual groomers and victims of sexual assault. In fact, however, Section 15 fits the whole range of contact groomers, and “Hypersexualized” is only one of three classifications used in the European Online Grooming research. Contact groomers belonging to all three types have gone on to have illegal sexual contact with children after grooming them in broadly 26 Ibid. p. 85 27 Ibid. p. 15 28 Ibid. p. 8f 14 similar ways. Besides hypersexualized groomers there are “intimacy-seeking” and “adaptable” groomers. “Intimacy-seeking” groomers29 according to the European Online Grooming project, do not typically collect sexual images of children and have no prior convictions for sexual offences. They described themselves as seeking consenting sexual relationships, rather than one-off encounters. They did not exploit anonymity on the internet and were quite willing to move from online to telephone contact, again being truthful in their offline conversations about their age and identity. Intimacy-seeking groomers, then, tend to share few characteristics with hypersexualized groomers. Between those two extremes lie “adaptable” groomers, men who typically had previous convictions, collected images on a modest scale, and had contacts with other groomers and online offenders. They are closer to hypersexualized offenders than intimacy-seeking ones. Their distinguishing characteristic is an ability to tailor their style of grooming to whatever type of child seems receptive to their online overtures. These groomers, then, can adapt to the different requirements of both “vulnerable” and risk-taking online conversation-partners.30 The sample on which this typology is based was composed of mainly middle-aged make groomers whose conversation partners and eventual victims were 13 to 15 years of age.31 Section 15 seems to cohere well, then, with what the empirical research reveals to be the behaviour of contact groomers. Although there are different typologies of groomers and of the grooming process in this literature, review articles indicate that competing approaches identify many of the same things as salient. For example, grooming consists not only of a scanning or surveillance stage, but of an initial contact and getting-acquainted stage, followed by rapport-building, exchange of confidences, and joint problem solving. After sexual contact takes place there is the process of making the victim compliant in keeping it secret. 32 While online and offline grooming processes have various things in common—they may share the stages of scanning, rapport-building, the exchange of confidences and joint problem solving—they also diverge. Offline grooming seems distinctively to require the cultivation by the groomer of a caring public image. When the groomer does not target his own children or family members, he often befriends the parents of young children and gains access to other children through volunteer work in youth clubs. It is different with internet-based grooming. Online chat rooms allow groomers to bypass parents and dispense with public image-building. When chat room contact is combined with deception about the groomer’s age, it can create the online impression of a romantic approach from another teenager, an illusion that offline grooming would normally exclude. In some respects, then, online grooming can be more deceptive and more insidious than its offline counterpart. Then there is the connection between sexual abuse facilitated online and the production of pornographic records of that abuse, which in a sense immortalize it and can further traumatize the victim.33 Section 15 is not addressed to online grooming alone, but is it appropriately framed for online grooming? I think the answer is ‘Yes’. First of all, it sets an age threshold for the offence. Someone charged under section 15 must be an adult. So sexualized conversation between teenagers under 18 is 29 Ibid. pp. 81ff 30 Ibid. pp. 83ff 31 Ibid. p. 89. For a related taxonomy , see Karen J. Terry & Jennifer Tallon, ‘Child Sexual Abuse: A Review of the Literature’ (The John Jay College Research Team, 2004) http://www.usccb.org/issues-and-action/child-and-youth- protection/upload/child-sexual-abuse-literature-review-john-jay-college-2004.pdf 32 See Whittle, H. C., Hamilton-Giachritsis, C., Beech, A., & Collings, G. (2013). A Review of young people’s vulnerabilities to online grooming, Aggression and Violent Behavior, 18, 62-70, esp. §2.3 33 K Raymond Choo, Online Child Grooming: A Literature Review on the Misuse of Social Networking Sites for Grooming Children for Sexual Offences grooming( Canberra: Australian Institute of Criminology, 2007) p.xii.:( http://www.usccb.org/issues-and-action/child-and-youth-protection/upload/child-sexual-abuse-literature-review-john-jay-college-2004.pdf http://www.usccb.org/issues-and-action/child-and-youth-protection/upload/child-sexual-abuse-literature-review-john-jay-college-2004.pdf 15 excluded, even if one of them is vulnerable or isolated and is in fact being targeted by the other. The central case of an older offender pretending to be the same age as the victim in order to facilitate a meeting is catered for, reflecting the particular hazards of online anonymity and the abilities of offenders to simulate teenage online chat. In the case where the provisions of section 15 are otherwise met but the target of the grooming is in fact a persona created by an undercover police officer, the groomer can be charged with an attempt of a section 15 offence.34 How completely does the section 15 insistence on age difference between the offender and the victim exclude unwanted consequences? Various issues are broached by this question, some having to do with the international reach of section 15, and the variation internationally in the recognized age of consent to sex. Table 4 shows age of consent to sex by country as of 2007. Ignoring changes to those ages that have been decreed since, we can see that an arrangement in 2007 by an English or Welsh person to meet a consenting 13-year old in Spain for sex would have counted as an offence under section 15 once the English or Welsh person had travelled to Spain. This may seem incongruous if an arrangement to meet in Spain, entered into by an English person in Spain after grooming in Spain, would not have been an offence. Another issue arises from setting 18 years as the threshold for adulthood. Conditions for a section 15 offence could be met by travel prompted by a sexualized teenage conversation between someone who was 15 years and 364 days and someone else who was 18 years and one day. Clearly this is not the central case of the deceiving older offender posing as a very young person for the purpose of meeting for sex, and it is a question whether in this kind of case a prosecution would actually be brought, or whether it would be selected for undercover police intervention. On the other hand, in a highly publicized recent case in England, Lewis Daynes sexually assaulted and killed a 14-year old boy whom he had groomed online. Daynes was only 19. The boy’s mother had reported the grooming to the police and had been ignored. Section 15 is not well-designed for this exceptional sort of case; but this is compatible with saying that it is well designed for typical online grooming offences.35 34 http://www.cps.gov.uk/legal/p_to_r/rape_and_sexual_offences/soa_2003_and_soa_1956/#a29 The common grooming tactic of getting the victim to produce a sexualized image of herself (which may later be used for blackmail purposes or posted shared with selected others online) is also catered for in other parts of the Sexual Offences Act. 35 Something like an endorsement of section 15 within a human rights frameworkis provided by Alisdair Gillespie in early commentary on the Sexual offences Act (2003) in Childright : http://www2.essex.ac.uk/clc/hi/childright/article/204/index.html#cR204_2 http://www.cps.gov.uk/legal/p_to_r/rape_and_sexual_offences/soa_2003_and_soa_1956/#a29 16 Table 4. Age of Consent by Country in 2007. Source: Choo, op.cit, p. 49 4. Undercover operations leading to Section 15 prosecutions In many jurisdictions anti-grooming legislation is enforced in sting operations conducted by undercover police. Officers pose as children in chat rooms likely to be used by groomers, and they maintain conversational contact over periods ranging from days to months. There are two kinds of case. Either the undercover officers assume the online identity of children whose grooming has reached the point where a meeting has been proposed; or else police officers impersonate a child with characteristics likely to appeal to a contact groomer. When the groomer responds to the child persona built up by the police, a conversation develops that over time culminates in an invitation to meet from the groomer. Undercover operations often if not always involve deception, and when they are directed against individual paedophiles or paedophile rings, they carry the risks of engaging in other kinds of morally and legally dubious behaviour. For example, to maintain cover, a police officer may have to hold back from intervening in the commission of a crime. When the undercover operation involves active child 17 abuse on the part of a paedophile ring being infiltrated, the broadcast rape of children in real time is sometimes passively witnessed for the sake of collecting evidence and mounting a large-scale prosecution. Are the moral costs of these operations justified by their results, if their results include the destruction of large, international paedophile rings? Arguably, yes. On the other hand, what if an operation is focused on one or two online groomers, and they are in fact fantasy groomers rather than contact groomers? What if undercover operations in this area run the risk of legal entrapment –where offences are committed only with the strong encouragement of police masquerading as children or as willing members of paedophile rings? In considering these questions, it is important to distinguish between operations against paedophile rings, which involve active and frequent sexual assault of children, and operations against uncoordinated online groomers, who may aim at sexual assault, but who may or may not be engaging in any sexual assault at the time they are in contact with undercover officers. Deception and apparent but very short-term collusion by the police may be tolerable if it breaks up a ring of child rapists for good, but how far can officers justifiably cut corners in operations against groomers? In its most general form the question before us is that of whether the harm produced by grooming leading to sexual assault is so severe that undercover preventive policing is justified. The answer to this question could be ‘Yes’, even if particular preventive policing operations are morally and legally defective. The answer does appear to be ‘Yes’. For apart from the harm of rape involved in much sexual abuse, to which at least all contact grooming is intended to lead, there is a wide variety of harms associated with online grooming. These range from depression, sexual dysfunction and drug- dependence to difficulty making friends, uncertainty over one’s gender or sexual preferences and suicidal and other self-harming tendencies.36 When one considers that sexually explicit images of children are often produced during the online grooming process, and that groomers sometimes threaten to post these on the internet or email them to parents and friends if the child does not co- operate, the adverse psychological effects of grooming are not hard to appreciate. Other things being equal, these seem to be serious enough harms to justify preventive measures, even where the preventive measures include deception and secrecy. We can distinguish, nevertheless, between the ethics of assuming the identity of a genuine child who is already being groomed, and developing the persona of a child online to see whether a groomer will make contact and suggest a meeting. In the former case, it is not true that, but for the intervention of the police, the grooming and the arrangement to meet would not have taken place. In the other case this arguably is true. For the persona of the child is wholly manufactured by the undercover officer, and not, as in the case where the police take over the online identity of a child, a matter of replacing a real child who is being urged by the groomer to have sex. It may be true in the case of the wholly manufactured persona that if the groomer had not made contact with that persona he would have made contact with and probably harmed a real child. But this fact, where it is one, does not absolve the undercover officer of a much more active role in the cultivation of the groomer’s interest than in the case where the officer merely continues online contact already established entirely independently of police deception. The more actively the undercover officer is involved in the creation of a kind of persona likely to appeal to a sex offender, the more active the undercover officer is in the process that is causally responsible for the invitation to meet. I am not suggesting that none of the causation comes from the offender, but more comes from the offender when all the police officer does is continue a 36 Choo, p.39, summarizing Melissa Wells & Kimberly J Mitchell ‘Youth Sexual Exploitation on the Internet: DSM-IV Diagnoses and Gender Differences in Co-occurring Mental Health Issues’, Child and Adolescent Social Work Journal, Vol. 24, No. 3, June 2007, pp.235-60, http://www.unh.edu/ccrc/pdf/CV128.pdf http://www.unh.edu/ccrc/pdf/CV128.pdf 18 conversation that has probably gone on for weeks, that the groomer is in control of, and that has been skilfully steered in the direction of sex. The problems in the case of the wholly manufactured child persona are magnified if one accepts the distinction between fantasy and contact groomers, and if one accepts that fantasy and contact groomers can both find sexualized conversation and even the fantasy of meeting for sex highly gratifying. This means that the suggestion of meeting could come from either kind of offender, and that even a fantasy offender could be convicted in a jurisdiction in which the suggestion of a meeting, without steps taken to meet the offender, could suffice for an offence. Section 15, as already pointed out, is well framed for weeding out fantasy groomers, but other anti-grooming offences in other jurisdictions are not, and these may produce injustice when combined with sting operations that succeed when the evidence of an intention to meet and have sex does not include travel to a meeting.37 So far I have been suggesting that, other things being equal, operations that continue contact with a groomer are morally more acceptable than operations in which contact is initiated by the undercover officer. This position does not imply that operations against groomers in which children are never involved are never justified. Since groomers often cultivate several potential targets simultaneously, it can prevent harm to real children if one of the targets of the grooming is in fact a police officer. Again, a sting operation in which a groomer targets an undercover officer can lead to the discovery that a repeat offender is offending again, or that he is violating his parole or other court-ordered restrictions. In this sort of case the benefits of the sting operation extend beyond protecting children to acting as a check on compliance with bail or parole conditions. If the groomers are repeat-offenders, then, since many offenders are involved in online exchange of child pornography, which can in turn be produced and distributed by full-scale paedophile rings, the anti-grooming undercover operation can actually lead to an operation that prevents much worse harm than grooming much more quickly than in the absence of an undercover operation. In short, because of the interconnections between grooming offences, offences consisting of making and distributing images, and organized sexual assault, even deceptive sting operations which manufacture a child persona attractive to groomers may be amply justified by the harm they prevent or help to prevent. There is a risk of legal entrapment of a groomer if the online conversation produced by a sting operation seems to offer a special incentive to the groomer to propose a meeting. An operation is more likely to produce a successful prosecution if the idea of meeting comes from the groomer, once the age of the “child” in the sting operation has been established, and the sexualization of the conversation has been initiated by the groomer. Many online conversations with groomers actually follow this pattern. Could there be circumstances, however, in which entrapment was morally necessary, notwithstanding the resulting failure of a prosecution? The answer seems to me to be ‘Yes’. Here we can imagine a sting operation that has revealed that the groomer is a repeat offender who has been convicted of a series of serious sexual assaults. We can imagine that he is also forensically aware, and that he has already raised several times with the “child” he is grooming the point that he is concerned that either the child’s parents or the police will find out 37 Can even someone who travels to a meeting disown the intention to meet for sex? In filmed confrontations of groomers who turn up at a meeting place, undercover police are sometimes told that the groomer was only coming to warn the child of the dangers of internet relationships. schttps://www.youtube.com/watch?v=VrhEZBl0qrQ; https://www.youtube.com/watch?v=YAS6XOsEd0s These claims are unconvincing on camera, but have succeed in at least one US court case. See Choo, op.cit.p.31 https://www.youtube.com/watch?v=VrhEZBl0qrQ https://www.youtube.com/watch?v=YAS6XOsEd0s 19 about their proposed meeting. Suppose he proposes a meeting place in which there are many escape routes, and which would make an arrest difficult. In such a case, for the sake of a more certain arrest, and the likelihood of disrupting his grooming, it might be justifiable for him to be given a special reason to go to a location where arrest is easier for the police, notwithstanding the grounds the defence at trial might have for excluding some evidence. A related case is where some of the evidence that the police uses to arrest a groomer comes from an internet vigilante group that makes contact with a groomer without the knowledge or co-operation of the police. The internet group would not have received the legal authorizations that a comparable police operation would have depended upon, creating grounds for the exclusion of evidence at trial in England and Wales. In cases like these the moral case for disruption may be stronger than the moral case for meeting the legal requirements of a successful prosecution under Section 15.38 5. Conclusion Section 15 of the Sexual Offences Act (2003) seems to survive general objections to preparatory offences in general brought by critics like Ashworth and Zedner. This is partly because the online grooming process often involves significant harm in its own right, including coerced production of sexualized images and blackmail involving the threat of publicizing those images to force the production of more. Harm is not only in the offing when the groomer is in a position to commit a Part One sexual assault. Again, Section 15 justifiably focuses on meeting after grooming rather than sexualized conversation as the trigger for arrest and prosecution. Offences that gear arrest only to evidence of online sexualized conversation between children and older groomers would probably lump together fantasy with contact groomers. Many moral risks attend undercover operations in which the police manufacture a child persona rather than assume the identity of a child that has been subject to grooming. But these risks may often justifiably be run in view of the dangers posed by hypersexualized groomers and the online grooming process itself. 38 I am grateful to Chris Nathan and Jethro Butler for comments on earlier drafts of this paper. Research for this paper was funded by the ESRC project “Assuming Identities Online (Project No.ES/L003279/1).