The Value of Crime Scripting : Deconstructing Internal Child Sex Trafficking

This article demonstrates how Crime Scripting is a viable and cost-effective tool in supporting strategic policing without requiring additional data, software, or training. This study shows how a script can deconstruct a complex crime into its component parts and create a set of outcome-focused recommendations informed by the principles of Situational Crime Prevention. Scripting offers an effective framework for collating and condensing voluminous data in order to establish a clear sequence of actions and decisions crucial to a given crime. This practical introduction to Scripting uses the example of Internal Child Sex Trafficking (ICST), a little-understood crime which has increasingly attracted police and government attention. Key findings from the offender-focused script highlight areas for harm-reduction interventions which go beyond traditional enforcement to include detection, disruption, and awareness. This article concludes by exploring the results’ application to diverse areas including policing, legal strategies, policy and research, and youth work and education.


Introduction
The police are increasingly expected to play preventative, disruptive, and educative roles, to collaborate with external agencies and to address a larger range of criminal and antisocial behaviours than ever before (Ratcliffe, 2004).Problem-oriented Policing (POP) continues to gain ground as an alternative to traditional enforcement-led policing styles (Goldstein, 1979).Rather than simply respond to crimes as and when they happen, POP attempts to discover and address the immediate causes of crime, so that preventative strategies can be implemented (Braga, 2002).New tools can offer invaluable support to analytical policing styles, such as POP, but they must be both accessible and affordable.With overall police budget cuts set at £125 million (Herbert, 2010) for the 2010 financial year, there will be even greater scrutiny on spending.Now is the time to explore ways to enhance the capabilities of existing resources to inform practical harm-reduction techniques.

Crime Science and situational crime prevention
Crime Science draws on tools and techniques from a multitude of disciplines and the only prerequisite is that they are based firmly on the scientific method.Empirical evidence forms the basis for theories, which are subject to robust testing (Cohen and Nagel, 1934) and an open discussion of any methodological or data limitations.As opposed to addressing 'crime' as an abstract whole, Crime Science focuses on combating specific crimes (Cornish and Clarke, 1987).This article, one of a pair exploring the merits of Crime Science when tackling Internal Child Sex Trafficking (ICST), will focus on Crime Scripting.Its complement investigates the offender and victim networks using Social Network Analysis (Cockbain et al., 2011).
It is important to briefly discuss Situational Crime Prevention (SCP), since the application of Crime Scripting draws heavily on its principles.SCP argues that opportunist offenders may be easily deterred by making conditions unfavourable to crime, i.e. by removing opportunities and incentives to offend.SCP's main contribution is a framework to guide stakeholders in developing crime-reduction strategies.Better known as the 'Five Key Pillars' (Clarke, 1997;Clarke and Eck, 2003), this framework is based on five ways to deter offenders: Increase effort; Increase risk; Reduce benefit; Reduce provocation; and Remove excuses.
Interventions need not be complex or expensive, indeed Crime Scientists argue that it is enough to change 'the situation in the narrow sense, not the environment in the broad sense' (Laycock, 2003).

The crime script
A script is simply a sequence of actions which make up an event.Initially developed in Psychology, scripts are now used in fields as diverse as Artificial Intelligence and Management Consulting.The classic example from Psychology is the 'Restaurant Script'.This expresses the typical sequence of actions involved in dining out: finding a seat; reading the menu; ordering food; eating; paying the bill; and leaving.In the context of crime, Cornish (1994) introduced Crime Scripting as a way of deconstructing an offence into its basic component actions.
With no commonly accepted rules on creating scripts, it is unsurprising that they have varied greatly in form and content.Some simply list sequential actions and never actually draw a diagram, which contrasts with a graphical representation depicting a series of action and decision points.
When presented graphically, scripts are commonly drawn as a series of boxes linked by arrows, indicating direction of flow.The boxes represent actions or decisions to be taken by the offender.Since the same crime can be committed in different ways, numerous routes, known as tracks, can co-exist on a single script.
Scripts can be used to deconstruct a wide variety of crimes, and are particularly useful for new or complex crimes.Scripts can be built using different types of data, including material routinely collected by the police.A benefit of scripts is that they can be created from partial or incomplete data and later amended as more information becomes available.Scripting requires no specialist software: our script was created in Microsoft Visio.Scripting techniques are currently used by police under a variety of names and guises, such as Flowcharting or Criminal Business Profiling.
Although Crime Scripting has traditionally dealt with high-frequency crime (JDI, 2010), recent work has demonstrated its wider relevance to organized crime, including sex trafficking (Bullock et al., 2010 andBouloukos et al., 2003).Part of a wider exploratory project, this article represents the first application of Crime Scripting to ICST.

Scripting ICST
Recent years have seen increased awareness of systematic grooming and abuse of UK children: yet ICST remains, in the words of the UK Human Trafficking Centre (UKHTC), a 'knowledge black hole'.1 Our term, ICST, will exclude the internal movement of international trafficking victims since their abuse has a very different profile.In line with government child protection guidance, 'child' will refer to a person under-18 (NSPCC, 2008), rather than below 16-the age of sexual consent in the UK.
Internal sex trafficking, of adults or children, is a crime under Sections 58 of the Sexual Offences Act 2003 (OPSI, 2003).This crime carries a maximum custodial sentence of 14 years (SGC, 2007).It is an offence to arrange, or facilitate, the movement of someone within the UK, for the purpose of sexual exploitation.Crucially, there is no minimum movement requirement, so movements within a city or even a single street qualify.
According to the Child Exploitation and Online Protection Centre (CEOP), British children constitute a fifth of all child trafficking cases in the UK (CEOP, 2009).Since there is no evidence of UK minors being trafficked other than for sex, it can be assumed that these cases represent ICST.Although the scale and geographical distribution of ICST remains unclear, fieldworkers describe a latent but widespread and ingrained problem.ICST offences can encompass numerous crimes, including rape, making and distributing child abuse images, and forced prostitution: due to low awareness levels of ICST and greater understanding and provisions for its component crimes, ICST is rarely reported or recorded as a trafficking offence.Consequently, existing ICST statistics are likely to be severely under-representative (CRIN, 2001).
Once one of the first major operation into ICST in the UK concludes in late 2010 (see Data below), police anticipate nationwide pressure to pursue ICST under the Home Office's harm-reduction agenda (Home Office, 2004).Police will be expected to develop investigation and intervention strategies quickly, with little time for additional research and analysis.In helping police understand ICST, our script could support rapid but wellinformed police responses.

Data
Our crime script was developed using combined data from two UK police operations into ICST, referred to here as Operations X and Y.By amalgamating the data, we were able to create a single script covering both operations which included 25 offenders and 36 victims.Although this appears to be a small sample size, the prevalence of repeat victimization, often over weeks or months, resulted in a much higher number of individual instances of abuse being reported.
The data used were typical of a police investigation and included: Victim Records of Video Interviews (ROVIs); Offender ROVIs; MG5 case summaries; Text messages and video footage from mobile telephones; and Formal charge list (Operation X only).
To complement police data, we gathered material from in-depth interviews with Government agencies, NGOs, and academics.Although this material did not feature directly in our analysis, it provided vital context and helped guide our interpretation of police data.Our data featured only male perpetrators and female victims but other perpetrator/ victim gender configurations may exist.

Crime script methodology
There are various levels of scripts and selection depends on the script's intended application.For example, a high-level script might be appropriate for an introduction to broad topics such as 'human trafficking' or 'child sex abuse'.These scripts, however, are often too vague to be of practical use.In order for our script to be of use to the police, we used the most specific level of scripting, one which deals with individual actions.This level is called the 'track' and follows an action sequence from start to finish for a particular version of the crime, e.g. for credit card fraud, possible tracks might be: card stolen from person, card cloned in shop, etc.The overall script is made up of individual tracks and shows how the same crime can be committed in a number of different ways.Our ICST script features multiple tracks, all based on the actions and decisions of offenders.
With no set script creation rules, we created our own typology of symbols (Table 1).
We grouped similar actions which fulfilled the same basic function under umbrella terms which covered all known variants of an overarching action.For example, 'attempt to communicate with girl' covered the subgroups 'call', 'text', 'wait outside house', and 'pass message via friend'.Using umbrella terms enables simplification and flexibility: the script is left open enough to accommodate subtle changes in approach (e.g.'send message via Facebook').Nonetheless, police starting a new investigation might wish to explain how they interpret key umbrella terms on the basis of known cases, to ensure that everyone has the same frame of reference.This is particularly important for terms such as groom which can cover a huge range of actions, some of which may be operation-specific.Figure 1 below shows our ICST Crime Script.
We only included actions in our script which led to an offence.The only exception to this criterion for inclusion was when an offender decided a girl was no longer worth pursuing and released her from the script.We felt this should be included as it represents an active decision by the offender (see Decision Points for further details).

Tracks
We identified five different starting points for ICST based on an offender's intentions and behaviour.These were: Cruise.In this track, the offender looks for a completely new girl to abuse.The term cruise was chosen because the majority of offenders drove around looking for girls near schools, shopping centres, parks, under-18 discos, and other 'targetrich' environments.Offenders who had no access to cars engaged in similar behaviour on foot.
Convert.Here the offender targets a girl who he has not previously abused but who he, or a co-offender, already knows.A common part of the grooming process was for offenders to socialize with girls prior to abuse.Since offenders regularly shared girls' phone numbers with each other, many girls were called by complete strangers who knew their name, age, and contact details.
Recruit via girl.'girlfriend' figures.'Girlfriends' were victims who were regularly abused by men they perceived as their 'boyfriends', a myth deliberately fuelled by some offenders.These girls played an important role in bringing their own friends to abuse situations.
Re-abuse.Many girls were re-abused on at least one further occasion.A small but significant number of girls were abused by many men over a long period of time: the most extreme case was a victim abused by around 150 men over 18 months.
Pimp.Although the majority of the abuse appeared to be motivated by non-commercial reasons, some girls were pimped out to clients.These clients included Bangladeshi restaurant staff who abused a girl in their workplace and a group of Polish drug dealers who abused a girl in their home.Girls were pimped for money but may also have been exchanged for other commodities, used as a 'sweetener' in drug deals, or used to repay debts.Once all individual tracks were established, we analysed them as a group to look for common actions where the tracks overlapped.We found that all tracks converged at grooming after which they proceeded through the same sequence of actions leading up to abuse.

Decision points
We identified four offender decision points: here an offender's decision determined whether he proceeded to the next action or chose to exit the script.Three of the four decision points occur before any actual abuse takes place: if offenders could be persuaded to exit the script here, more girls would escape unharmed.At present, little is known about why offenders choose to give up on certain girls.If those girls who escaped unharmed shared certain traits or actions, these insights might inform interventions designed to safeguard potential victims.Moreover, a clearer sense of how many girls are targeted for every one abused would give insights into the victim selection process and its efficacy.An alternative source of such information would be post-trial interviews with convicted offenders.
Two of the decision points prior to abuse have so-called 'feedback loops': when an offender decides that a communication attempt has not been successful or a girl is not 'ripe' for abuse, he asks himself if it is worth persisting.If the answer is yes, he goes back to the prior action and repeats it, as many times as necessary before he can move forward in the script or gets tired and gives up.It was common, for example, for an offender to call the same girl repeatedly to break down initial resistance to meeting up.After what could be dozens of calls over a short period, many girls decided it was easier to pick up and agree to meet than face continuing bombardment.
The final decision point differs slightly in that its feedback loop leads directly into the Re-abuse track.Some girls are victimized over weeks, months, or even years.When such sustained abuse goes unnoticed or unpunished, this may fuel a sense of impunity among offenders.

Stages
We merged the five tracks, where possible, to avoid duplication of actions and added the feedback loops, before dividing the script into three consecutive stages.We did this by clustering actions with the same overarching goal, which were Find, Groom, and Abuse.In the first stage, Find, actions vary according to track, but by the end of this stage the tracks converge and share the same sequence of actions across Groom and Abuse.
The actions in Abuse are illegal and constitute sex trafficking and sexual offences.In contrast, most actions in Find and Groom are technically legal.Yet, without these actions abuse cannot occur.This raises questions around the criminalization of grooming: how can children be protected from ICST without demonizing normal adult-child interactions?Although it is already illegal to meet a child after sexual grooming, this requires an offender to have met or communicated with a victim on two or more occasions before abusing them.This could take the form of physical meetings, online interactions, or telephone communication.Often offenders act so quickly that this criterion would not be met.In Operation X offenders abused 40% of victims2 on their first meeting.With its numerous possible configurations, grooming defies easy definition and vague legislation might lead to large numbers of false positives.
Stage 1: Find.The victims included in our data came from a variety of demographic and socio-economic backgrounds and were approached by offenders at any time, on any day of the week.Although some offenders were skilled in psychological manipulation, even for them it appears to have been a numbers game: approach enough girls and some will cave in.They rarely took the risk of crossing obvious barriers to reach a girl, e.g.bouncers or school gates, and instead loitered nearby.This suggests a rational evaluation of risk which in turn means they may be receptive to SCP-based interventions aimed at deterrence.
Offenders could skip the cruising stage if they already had a target.Some girls passed on their friends' details to gratify or placate offenders and/ or divert attention away from themselves.With this liberal flow of information between offenders, girls were sometimes contacted directly by unknown men, often calling from withheld numbers.
Once identified, offenders relied on unsophisticated but effective tricks to get a girl's attention, such as whistling or shouting.They then bantered, flirted, or pestered their way to extracting name, age, and telephone number.
Stage 2: Groom.This stage was characterized by huge variation in type, degree, and frequency of grooming techniques.Offenders tended to use a range of tactics, often switching abruptly between them, for example, being nice one moment and nasty the next.The extent to which they were aware of the psychological impact of their behaviour is unclear, but Table 2 shows the primary effect of some of the more common techniques.
Our data showed no fixed grooming period: offenders abused some girls within minutes and groomed others for weeks.
Stage 3: Abuse.Once offenders believed girls were suitably groomed, they moved into the Abuse stage.They invited targets to socialize, often offering to pick them up by car.They took girls to abuse locations, which were usually co-offenders' properties, cheap hotels, playing fields, or even parked cars in isolated areas.In order to qualify as sex trafficking, the victim's movement must somehow be facilitated by the offender (OPSI, 2003), e.g. by buying a train ticket for her or driving her to an abuse location.Although not every individual assault in our data set fits this criterion, every victim had been trafficked at least once.
Offenders used various physical and non-physical isolation tactics.They locked doors or took away mobile telephones.Even if girls were not physically restrained, they may have been too afraid of repercussions, or disorientated due to drink, drugs, or an unfamiliar location to risk an escape attempt.Importantly, many of these approaches are illegal and fall under the offence of false imprisonment (Slapper, 2007).
Abuse covered vaginal, oral, and anal rape, forced masturbation, and sexually inappropriate touching.In some cases it was not limited to a single action, but featured a number of offenders taking turns to abuse the victim.Offenders sometimes filmed abuse on mobile telephones and used it to scare other girls and presumably to brag to co-offenders.Once finished with a girl, offenders drove her home, arranged a taxi, or ordered her to leave.In a small number of cases, victims managed to escape.Varying levels of violence, pressure, and coercion often accompanied the abuse and deterred reporting.

Generating initiatives from the script
To assess whether our script could help tackle ICST, we used it as stimulus for a structured brainstorming session.We addressed each action systematically, considering possible intervention initiatives based on the five pillars of SCP.
This approach is illustrated in Table 3, which shows ideas for one action from each of the three stages.These suggestions span areas including policy, education and awareness, enforcement, and deterrence.ICST suspects could be targeted  for motoring offences in order to take away their driving licenses and therefore make cruising harder.Mobile phone service providers might be enlisted to bar all calls from withheld numbers on request.Awareness campaigns might encourage reporting and deter some offenders.Controversially, legalizing prostitution might reduce the client-led demand for ICST.
As expected, our offender-focused script primarily produced ideas for offender-focused interventions.When victim-focused interventions are formulated, it is crucial not to imply that the victim is to blame e.g. by berating girls for wearing revealing clothing.This is especially relevant to the 'Remove Provocation' pillar.
To maximize the impact of limited funding, interventions could be targeted at places where the tracks have converged.By focusing interventions on the Groom and Abuse stages, for example, all five tracks could be tackled simultaneously.
Although we have tried to offer realistic suggestions, our idea generation was not constrained by specific practical considerations.In reality, ideas would need to be assessed and key contenders refined and tested against criteria such as cost, resources, social acceptability, and political viability.

Applications
In addition to the script-specific interventions above, we considered the general applications of our crime script for different stakeholders in the following core areas: policing; legal strategies; policy and research; and youth work and education.

Policing
Our script might be a useful briefing document for police new to ICST.This single diagram conveys much information about the component parts of the crime and how they connect together.This information would be of use to senior officers responsible for strategy planning as well as front-line officers and analysts working directly on investigations.A clearer sense of behaviours involved in ICST might help the interpretation of intelligence around possible cases and the recording of instances as internal sex trafficking offences.
The script might provide a helpful supporting document for funding requests for new investigations or preventative or disruptive interventions since it clearly conveys two key issues central to ICST cases: human trafficking and child sex abuse.
The script might also be used in community outreach work to introduce ICST and stimulate debate around possible community initiatives.It might help engage communities in reporting suspicious behaviour and explain the thinking behind more unusual interventions.

Legal Strategies
In the UK to date, offenders have typically been charged with rape, false imprisonment, etc. rather than trafficking: an offence that can be tried without a victim's testimony (CPS, unpublished data).Giving evidence in court is strenuous and sex abuse cross-examinations are notoriously traumatic: two-thirds of legal professionals involved in child sex abuse cases would not want their own child to endure the ordeal of testifying (Eastwood, 2003).It is possible that a greater understanding and awareness of ICST by legal practitioners would have a positive impact on prosecution strategies, achieved, in part, by the use of our script.There might be a further role for scripts as an evidential aid to be presented to jurors, since they clearly visualize complex data and provide an overview of the crime.

Policy and research
Our script offers policymakers a concise overview of ICST and might inform debate around the criminalization of grooming.Further research needs include: testing our offender script's validity for male victim ICST data; creation of a victim-focused script which identifies victims' actions and decisions and possible areas for intervention; and elaboration of scripts through work on decision making.
More needs to be known about ICST clients so that appropriate and focused interventions can be created.Although we believe offenders primarily targeted young girls because they were easier to manipulate, it is possible that some clients actively preferred young girls.

Youth work and education
Adults who work with children, e.g.teachers, social workers, and youth group leaders may be privy to information regarding ICST.They are legally obliged to report child protection issues, but concerns may be dismissed due to lack of hard evidence.Anecdotal data from such sources could, however, add depth and context to the script.

Script considerations
There are a number of important things to remember when using the script, including: Scripts are just one of many tools police can use when tackling ICST.They can be an invaluable support but their interpretation and application requires thought and should not become automated.
Our script has been created from two operations and is by no means exhaustive.It remains a work-in-progress, to be amended as new data become available.
Our script accommodated the material we received from NGOs as well as fitting the police data.As new data become available, our script can be tested, and if necessary modified.
Both sets of police data may suffer from biases such as omission, exaggeration, understatement, and/or fabrication on the part of the victims and offenders.This is an unfortunate but inevitable reality of statements given to the police.
Our script covers the sequence of actions in any one instance of ICST, but it does not indicate the extent of offenders' ICST behaviour: i.e. how many girls they may have on various tracks at any one time.Neither does it cover other forms of crime committed by ICST offenders, such as drug dealing, tax evasion, or attempting to facilitate illegal immigration.Including such material would have made the script too complex for practical use.

Conclusion
Police officers are increasingly under pressure to play a role in developing harm-reduction interventions.As policing continues to move towards a more proactive, as opposed to reactive, approach, it is important to find tools which complement current capabilities and open up new avenues for intervention.
Crime Scripting allows specific crimes to be logically deconstructed into their component parts, even from a relatively small data set.This tool enables an apparently complex crime such as ICST to be mapped out in a clear way.The resulting diagram can help inform focused intervention based on previous experience and knowledge of typical features of a given crime.
Scripts alone cannot solve crimes, but if used in a thoughtful and logical manner they can generate novel ideas for intervention strategies.They can ensure a common language and frame of reference for different stakeholders and therefore support multi-agency collaborations.

Table 1 :
Crime script typology CategoryDescription SymbolStarting pointThe way the offender identifies and selects his target (i.e.new girl or previously abused girl) determines the actions in the initial phase of the crime Action Action performed by the offender Decision point Offender's decision, based on assessment of situation, determines next action Flow line Flow lines direct progression through the script Exit point Point at which offender gives up on the girl in questionScript findingsOur script yielded a number of key insights into ICST that were common to offenders and victims in both operations, three of which are discussed below, namely:

Table 2 :
Grooming behaviours -Insult or undermine friends, family, or other support networks -Take girl to unfamiliar or difficult to access places -Separate girl from her friends if they are also present -Lock door and block girl's exit

Table 3 :
Ideas for interventions