Spotting the behavioural signs of predatory offending in the night-time economy

Predatory behaviour exhibited by men towards women in public spaces is often dismissed as mere ‘street harassment’ (Fileborn and O’Neill, 2021). This term, seemingly benign, can obscure the precursor signs of more sinister crimes, including sexual assault and rape. 

This article discusses findings from a proactive policing strategy in the night-time economy (NTE) of Thames Valley Police (TVP), emphasising the need to upskill police officers in proactive tactics and bystanders in recognising predatory behaviours amounting to street harassment. 

Background and context 

In 2019, 42% of violent and serious sexual assaults in the city of Oxford were linked to the NTE. Reports of stranger rapes and a series of sexually motivated creeper burglaries involving an offender known as ‘The Prowler’ raised real concerns. It was unknown whether the rise in stranger rapes and the creeper burglar series were connected. The Prowler was suspected of identifying his victims in the NTE and stalking them as they walked home. The common factor with each victim profile was their route home.  

Our response involved deploying plain-clothes and uniformed officers along the route to identify and intercept The Prowler. With only a vague description to go on, we looked for men who were displaying stalking behaviours. The observations were startling. Numerous men were seen to be converging in the public spaces of the NTE, acting in a predatory way towards women. The predatory behaviours ranged from loitering and aggressive persistence to stalking, harassing, making sexualised comments and unwanted sexual attention.   

Each person was engaged with by uniformed police and their activity interrupted. This led to the arrest of The Prowler (subsequently convicted of his crimes) and the revelation that nearly half of the individuals stopped (19 out of 41) were known to police for sexual offences.  

A notable reduction in sexual assaults (48%) linked to the NTE during our patrols and the capture of The Prowler was deemed a success by TVP leaders. It was decided to broaden the scope of the operation across each of the four main NTE areas in TVP. This was named Project Vigilant. 

Additionally, the Serious Crime Analysis Section (SCAS), which advises on modus operandi (MO) of offenders, taught us that two-thirds of stranger rapes start with seemingly benign interactions known as ‘behavioural try-outs’. These interactions allow offenders to assess the vulnerability of potential victims and the presence of guardians, such as police or friends. Offenders use these moments to decide whether the rewards of their actions outweigh the risks involved. 

During the initial operation, we observed ‘try-outs’ where suspects pretended to innocently loiter in the NTE, often standing alone and not entering venues. They used dark doorways or bus stops as staging grounds, watching and following women, sometimes engaging them in conversation. This behaviour typically targeted individuals perceived as vulnerable. 

Literature review 

TVP’s approach to proactively policing the NTE to tackle sexual violence fills a gap in existing research. By synthesising insights from different studies through ‘theory knitting’ (Kalmer and Sternberg, 1988), our programme builds on our knowledge about the predatory nature of sex offenders (Lisak, 1989; Lisak and Miller, 2002; Rossmo, 2000; Beauregard and others, 2007) with core criminological theories, ‘rational choice theory’ and ‘routine activities theory’. These provide context around the environmental factors that facilitate street harassment.  

Rational choice theory (Clarke and Felson, 1993) suggests that offenders evaluate risks and rewards before engaging in predatory behaviour. Although factors such as alcohol and substance abuse are often thought to influence such behaviours (Beauregard and Leclerc, 2007), our findings may challenge this assumption, particularly in the context of street harassment.  

Routine activities theory outlines three prerequisites for crime: a motivated offender, a suitable target and the absence of a capable guardian (Cohen and Felson, 1979; Felson, 2016). However, our observations of predatory offences occurring in the presence of ‘capable guardians’ prompt a re-evaluation of what ‘capable’ truly means in these settings.

Data and methodology 

The study describes 532 instances of predatory behaviour in the four main NTEs of TVP: Milton Keynes, Oxford, Reading and the Royal Borough of Windsor and Maidenhead (RBWM). Each have unique characteristics but share several key factors in their NTE, such as peak NTE weekends, month-ends and bank holidays, influenced by the presence of active university campuses.  

Milton Keynes offers distinct policing challenges, with areas such as The Hub and the Theatre District. Oxford’s vibrant NTE thrives with a mix of students and tourists. RBWM, being close to London, attracts out of towners and Reading, with its bustling town centre and the contrasting Oxford Road, both attract diverse populations. These commonalities are crucial in understanding the NTE dynamics, despite the physical and social contrasts that define each location.  

Each NTE operates under the guidance of a locally based operational lead. Patrols across the different NTEs are not uniformly scheduled. Instead, they are tailored based on the unique needs and demands of each area. For this study, the NTE consists of the public spaces near to bars and venues. This includes late night eateries, taxi ranks and bus depots, covering the hours of 20:00 to 04:00. It is important to note that the operational scope of Project Vigilant does not include the inside of venues, clubs or pubs for reasons explained in the recommendations.  

The primary data source for the study derives from standardised debrief forms spanning May 2021 to September 2023. The forms are used to document all ‘stop and accounts’. The form, not dissimilar to a stop and search form, provides a structured framework for officers, guiding them through the process of recording details from these interactions. Key elements on the form include the perpetrator’s personal details, aiding officers in identity verification, behavioural risk factors and the suspect’s account of their behaviour. The form forms the basis of a risk prioritisation matrix, which is used to build a risk profile of each person stopped.

Data pre-COVID (May 2019 to February 2020) was used to assess the efficacy of the initial proactive operation to catch The Prowler and develop it for wider implementation but was excluded from the main study. The impact of COVID restrictions on the reopening of the NTE is largely unknown. Data to ascertain the number of people going out before lockdown and post-lockdown was inconsistent and unavailable when the study began.

The risk prioritisation matrix and a selection of vignettes from each risk category were introduced to the study to illustrate the issues at hand. However, this article will focus specifically on the data around predatory offending in the NTE, including identified behaviours, MO and relationship to the victim.

Limitations

One of the key limitations in the data is the method by which it is recorded. In every case, it is police-recorded data based on police observations. Aside from the glaringly obvious potential for bias, it has been reported that police-recorded crime data is not considered reliable (Office for National Statistics, 2023). To address this and mitigate a claim of bias, the debrief form was devised to ensure officers followed strict protocols for obtaining pertinent details by recording the circumstances that led to the stop and account conversation. 

All 200 officers involved in the study completed criminal behavioural detection training before deploying on a dedicated Project Vigilant patrol. The training blended theory and practice in the live NTE for authenticity and covered the strict operational protocols they needed to follow. Despite these measures, it is accepted that this cannot entirely mitigate bias.

Findings 

There is no correlation with university term times and the identification of predatory suspects. They correlate with weekends, the end of the month, bank holidays and large events. There is a predominance of predatory behaviours among men aged 24 to 34.  

Loitering was the most observed behaviour, often associated with unwanted attention and sexualised comments. Fewer than 10% of stops were influenced by external factors such as alcohol or drugs, emphasising the calculated nature of suspects’ actions.