Long read: sexual risk orders as a tactic to counter sexual violence against women and girls
Sexual risk orders (SROs)
Published on
Written by Detective Chief Inspector Dan Whitten, Metropolitan Police Service
Going equipped
14 mins read
At the time I began this research, I was running a criminal investigations department (CID) that had responsibility for investigating serious sexual offences. It was a constant concern that a number of offenders clearly posed an ongoing sexual risk to women and girls, but prosecution was often not possible. This felt like an area with considerable unmanaged risk.
When reviewing what was available to try and mitigate this risk, sexual risk orders (SROs) were often mentioned as an option, however, they appeared to be little used and little understood.
This piqued my interest. I wanted to understand whether SROs might be the answer to preventing sexual offending when prosecution was not possible.
What are SROs?
- SROs are granted by a magistrates’ court following application from the police.
- They apply negative prohibitions, such as having no more than one device, and positive requirements – for example, the recipient must engage with treatment services.
- There is no need for a criminal conviction. The court must only be satisfied that the order is necessary to prevent harm and that the recipient has committed an act of a ‘sexual nature’. This is a term used in the legislation and is not defined and it is for the court to determine whether the relevant act satisfied this criterion.
- The court need only be satisfied “on the balance of probabilities” and hearsay is admissible.
- SROs can apply to anyone from the age of 10 years.
- The minimum duration of an SRO is two years, with no maximum period.
- Breach of an SRO is a criminal offence, with a maximum sentence of five years’ imprisonment.
Contention
SROs are intended to be preventative, and proponents insist that they are. However, critics have argued that there is no evidence that they are effective at preventing harm (Bowers, 2023).
Critics also claim that their use punishes the legally innocent with severe restrictions, as SROs do not require the usual safeguards applied to criminal proceedings. Such safeguards would include the burden of proof being “beyond reasonable doubt” and significant limitations on the admissibility of hearsay evidence.
Literature
Behaviour orders have increased considerably in the UK recently (Kelly, 2019). However, despite this growth and legal controversy, levels of research on behaviour orders are negligible (Brown, 2020). There have been very few methodologically robust evaluations (Rubin and others, 2006). Where research does exist, it focuses disproportionately on the now-obsolete anti-social behaviour order, or ASBO (Home Office, 2012).
Research tends to avoid an empirical assessment of whether orders prevent harm. Instead, it usually concentrates on:
- the perception of individuals involved (Kirby and Edmondson, 2012)
- the views of practitioners (Wain and Burney, 2007)
- small samples of self-reported criminality (Matthews, 2007)
Isal (2006) highlights that the data collection regime does not enable assessment of the equity of behaviour orders between communities. This appears remiss, given the well-versed disproportionality in a range of law-enforcement tactics (Quinton, 2015).
A study by Carr and others (2017) represents a lone and notable effort to assess the efficacy of a behaviour order (in this case, gang injunctions) empirically. The authors concluded that there was evidence that orders were effective at reducing crime-harm, with little indication of their use ‘backfiring’.
Literature pertaining to sexual behaviour orders is consistent with the broader behaviour order literature. Much of the work only includes the author’s commentary on:
- whether these orders are just (Kingston and Thomas, 2018; Thomas and Thompson, 2014)
- whether they are punitive (Kelly, 2019)
- how they interact with established legal principles (Kelly and Picton, 2020; Bowers, 2023)
Analysis of how SROs are used in practice is extremely limited. SROs remain absent from formalised longitudinal data collection (Ministry of Justice, 2022) and ad hoc efforts have suffered from reporting issues (College of Policing, 2022).
In 2021, His Majesty's Inspectorate of Constabulary and Fire & Rescue Services (HMICFRS) reported that SRO training had limited uptake and poor reviews from frontline officers. This appears indicative of limited understanding and use among practitioners.
There is a repeated insistence that the preventative potential of behaviour orders must be better understood (Home Affairs Committee, 2005; Isal, 2006). However, this has not been realised. This “absence of robust empirical research means that much of what is written is dominated by anecdote, conjecture and rhetoric” (Nixon, 2005, page 22). Although this quotation was written 20 years ago, it still applies in the present. As a result, the significant legal contention surrounding SROs cannot be settled.
Similarly, police leadership requires greater understanding of how SROs are currently used, who they are used against and what the cost-benefit is. Without this, leadership is not in a position to give informed consideration, if any, to how best to use SROs.
Research questions
My research aims to answer three questions:
- how are SROs currently used?
- do SROs effectively prevent sexual harm (in other words, harm caused by criminal offending that is sexual)?
- are SROs targeted effectively against the most harmful eligible offenders?
Method
This study drew on data from the following three areas.
- Metropolitan Police Service (MPS) Department of Legal Services (DLS) - this was used to construct a counterfactual population by identifying subjects for whom SROs had been considered but were not subsequently obtained. Similarly, legal cost data was obtained for SROs.
- Police National Computer (PNC) - all live SROs in England and Wales were identified through PNC. This data set was used to construct two SRO populations. The first population included all MPS SROs granted between August 2021 and August 2023 (n = 50). A second population included SROs granted nationally between March 2015 and March 2018 (n = 55). PNC was also used to interrogate the records of those within the populations. This was to establish arrests, breach data, the context in which SROs were sought, SRO conditions and personal characteristics of SRO subjects.
- MPS VAWG100 (Sweetland, 2023) - a process that rank-orders violence against women and girls (VAWG) offenders based on harm caused, from the most to least harmful. To assess whether SROs are targeted against the highest-harm offenders, the VAWG100 search was altered and built upon in a manner conducive to targeting SROs.
Descriptive statistics were used to assess the:
- difference in SRO use between and within forces
- personal characteristics of SRO recipients
- catalyst for SRO application
- duration of orders
- breach rate
- number of conditions
The Cambridge Crime Harm Index (CCHI) was used to quantify sexual harm. This calculates harm based on offences committed by using sentencing guidelines. Harm scores were calculated for all SRO subjects in the decade before SRO and for all years after. Tests of comparison, correlation and an analytical technique called interrupted time series were used to assess whether SROs prevent sexual harm.
An altered VAWG100 stack was built of offenders whose offending was relevant to SROs. For example, non-sexual VAWG offending and domestic abuse offending was removed. The stack ranked offenders from most to least harmful. This was compared to subjects for whom SROs had been obtained, to determine whether the most harmful were subject to an SRO.
An assessment of offenders was carried out to determine SRO eligibility. Consequently, a calculation was undertaken to project the harm that could be prevented by targeting the highest-harm offenders eligible for an SRO. This was then used to undertake a cost-benefit analysis of SROs, in conjunction with cost calculations for obtaining an SRO and the Home Office economic and social cost of crime calculations.
Findings
How are SROs currently used?
In total, 307 SROs were obtained nationally between 2021 and 2023, with the majority of forces obtaining fewer than 10. There is considerable variance in SRO use between and within police forces. The average age of SRO recipients is mid- to late-thirties, and they are overwhelmingly male. Racial disproportionality is not apparent (relevant to the resident population of London). SROs are predominantly sought following failed prosecutions. On average, three failed prosecutions precede SRO application, but this can be as many as 10.
The courts grant SROs on average for more than double the minimum two-year term. Approximately 42% of those with SROs are subsequently arrested for breach. On average, the first breach arrest occurs just over 500 days after the SRO is granted. Of those who breach their SRO, the majority do so more than once, and almost a fifth breach at least five times. The average number of conditions attached to an SRO is five. The level of police supervision and condition severity do not appear to be associated with degree of harm reduction.
Do SROs prevent sexual harm?
Subjects of SROs committed approximately 85% less sexual harm per month following receiving an SRO, when compared to before receiving an SRO. For higher-harm offenders who receive SROs of greater duration, this reduction increases to approximately 93%. This is after accounting for any time incarcerated. Conversely, a counterfactual population – subjects for whom SROs were discussed but not subsequently obtained – saw sexual harm increase by 68%.
The interrupted time series chart below details the sexual harm caused by SRO subjects in the five years before they were given an SRO, versus the five years afterwards. The timepoint at which SRO was obtained is denoted by the vertical dashed red line, while the sexual harm caused is represented by the solid blue lines.
The chart shows that sexual harm reduces significantly immediately following SRO and a sustained downward trend in sexual harm is then maintained in the years that follow. The shaded grey area shows a prediction of the sexual harm caused in the years after SRO, had a SRO not been obtained.
Are SROs targeted against the most harmful offenders?
SROs are not targeted against the highest-harm eligible offenders. A negligible proportion of the highest-harm offenders subsequently receive SROs. There is considered to be a viable means to target SROs against those who cause the greatest VAWG harm, using a rank-ordering of offenders to enable systematic targeting. If such an approach were adopted, considerable harm prevention is projected. SRO cost-benefit analysis suggests that benefit comfortably outstrips cost by a margin as high as 27:1.
Discussion
This study suggests that SROs are a promising tactic to prevent sexual harm. It is therefore advocated that SROs are used more often and as a tactic of early preventative intervention, rather than as a last resort. There are, of course, limitations and additional research is required to provide greater confidence in the robustness of the results.
SROs are not systematically targeted but rather rely on the motivation, knowledge and capacity of individual officers. In particular, findings support increased SRO use, specifically targeted against the most harmful offenders. This should serve to optimise harm reduction, maximise value for money and bolster confidence in the policing response to VAWG.
Tracking SROs and their impacts is also advocated to facilitate understanding of use. In turn, this enables ongoing public dialogue and improving use. This is essential to deepen the practical understanding of sexual behaviour orders, and to provide continued assurance of efficacy and value for money.
Practical advice for forces
Given the suggestion of effectiveness for VAWG harm prevention when prosecution is not viable, forces may want to consider:
- actively monitoring how they currently use SROs, including the volume, in what context, against whom and the outcome on offending or harm
- increasing SRO usage
- SRO usage as a tactic of early intervention, rather than one of last resort after multiple failed prosecutions
- incorporating SROs as a prominent tactic within pre-existing tasking and offender management processes in the areas of VAWG and sexual offending
- implementing a data-driven approach to identifying the highest-harm eligible offenders, rather than reliance purely on the professional judgement of individual officers
- raising awareness of SROs as a tactic
- considering how the application process can be partially automated
Broader policy considerations
There is limited understanding of SRO use, in large part because the national data collection regime is poor. It is difficult for forces to identify how many SROs have been obtained, in what circumstances and what the impact is. It may be prudent to consider improving the data collection regime at force and national level.
Policing tactics need to be assessed to determine effectiveness. This is particularly important where there is considerable infringement on individual rights. SROs were not assessed for the first eight years of their use. Future behaviour orders may benefit from a robust approach to assessment being built in from the outset.
Once tactics are identified as effective, it is important to gauge officer knowledge and to review the process used to implement the tactic, with a view to optimising its use. With SROs, it is clear that officer knowledge is limited and the process for acquiring them is time-consuming. Given indications of efficacy, it may be apt to consider means to increase the profile of SROs, improve the training offer in this area, and embed SROs into national VAWG workstreams and initiatives.
- This article was peer reviewed by Stephanie Devon, Detective Inspector, Kent Police
Further information
- Full research article: Whitten D. (2024). Tracking, testing and targeting sexual risk orders: A tactic to counter violence against women and girls. MSc thesis, University of Cambridge
- Further information: Whitten D, Neyroud E and Neyroud P. (2025). Sexual risk orders as a tactic to counter sexual violence against women and girls. Cambridge Journal of Evidence-Based Policing, volume 9(1), pages 1 to 19
References
- Bowers R. (2023). Sexual risk orders: Lowering the standard. Doughty Street Chambers. (Viewed on 2 October 2025)
- Brown KJ. (2020). ‘Punitive reform and the cultural life of punishment: Moving from the ASBO to its successors’. Punishment & Society, volume 22, pages 90 to 107
- Carr R, Slothower M and Parkinson J. (2017). Do gang injunctions reduce violent crime? Four tests in Merseyside, UK. Cambridge Journal of Evidence-Based Policing, volume 1, pages 195 to 210
- College of Policing, HMICFRS and Independent Office for Police Conduct (IOPC). (2021). A duty to protect. Police use of protective measures in cases involving violence against women and girls. (Viewed on 2 October 2025)
- College of Policing and National Police Chiefs’ Council (NPCC). (2022). Tackling violence against women and girls – Policing performance and insights publication. Statistical bulletin: Police performance, England and Wales. Reporting period: 1 October 2021 – 31 March 2022. (Viewed on 2 October 2025)
- Home Affairs Committee. (2005). Select Committee on Home Affairs fifth report. (Viewed on 2 October 2025)
- Home Office. (2012). Putting victims first: More effective responses to anti-social behaviour. London: Home Office
- Isal S. (2006). Equal respect – ASBOs and race equality. London: Runnymede Trust
- Kelly R. (2019). Behaviour orders: Preventive and/or punitive measures?. DPhil dissertation, University of Oxford
- Kelly R and Picton HHJM. (2020). Sexual harm prevention orders and necessity. Criminal Law Review, volume 5, pages 411 to 428
- Kingston S and Thomas T. (2018). The sexual risk order and the sexual harm prevention order: The first two years. Probation Journal, volume 65, pages 77 to 88
- Kirby S and Edmondson A. (2012). The effectiveness of the ASBO – a practitioner perspective. Safer Communities, volume 11, pages 96 to 104
- Matthews R, Easton H, Briggs D and Pease K. (2007). ‘Assessing the Use and Impact of Anti-Social Behaviour Orders’. Bristol: The Policy Press.
- Ministry of Justice and HM Prison and Probation Service. (2022). Multi-agency public protection arrangements (MAPPA) annual report: 2021 to 2022. (Viewed on 2 October 2025)
- Nixon J. (2005). ASBOs: More questions than answers. Criminal Justice Matters, volume 62, pages 22 to 37
- Quinton P. (2015). ‘Race disproportionality and officer decision-making’. In Delsol R and Shiner M, ‘Stop and Search: The Anatomy of a Police Power’. Basingstoke: Palgrave Macmillan, pages 57 to 78
- Rubin J, Rabinovich L, Hallsworth M and Nason E. (2006). Interventions to reduce anti-social behaviour and crime: A review of effectiveness and costs. Santa Monica, CA: RAND Corporation
- Sweetland J. (2023). Prof Sherman: ‘Who are the most important people to go after, and how much VAWG harm can we prevent?’. Policing Insight. (Viewed on 2 October 2025)
- Thomas T and Thompson D. (2014). New civil orders to contain sexually harmful behaviour in the community. British Journal of Community Justice, volume 12, pages 19 to 33
- Wain W and Burney E. (2007). ‘The ASBO: Wrong turning, dead end’. London: Howard League for Penal Reform
