Fairness, relationships and perceptions of police legitimacy in the context of integrated offender management

ABSTRACT Integrated Offender Management (IOM) involves the police in England and Wales working closely with other criminal justice agencies to reduce the criminal activities of prolific offenders. The work involves traditional policing methods, but also police officers engaging in meaningful rehabilitative work with IOM participants. However, there may be a tension between aspects of IOM policing and procedural justice theory. Research has shown strong links between citizens’ perceptions of fair treatment and their beliefs about police legitimacy, yet at the same time some IOM policing efforts emphasise hostile and disciplinary practices. These types of approaches can threaten police legitimacy if perceived by citizens as unfair. This paper examines the self-described experiences of people subject to IOM policing. Study findings revealed that, whilst participants viewed aspects of IOM policing as unfair, they broadly accepted the legitimacy of IOM policing. This would run counter to many of the core claims of procedural justice theory. However, one plausible explanation for this reaction lies in the dynamic nature of police legitimacy and its antecedents in IOM. In this context, perceptions of police legitimacy are more firmly grounded in the relational (treatment) dimensions of procedural fairness than in other more decision-orientated aspects of the concept. This reflects the multidimensional nature of procedural justice within which the balance shifts, depending on the criminal justice setting.


Introduction
Procedural justice theory (PJT) is a process-based model of regulation that states that police legitimacy and compliance/cooperation with the law are obtained not through the threat of force or risk of sanction but through the presence of procedural justice during police-citizen encounters (Tyler 1990, Paternoster et al. 1997, Sunshine and Tyler 2003).Procedural justice (or procedural fairness) is defined as 'the fairness of the processes through which the police make decisions and exercise authority' (Sunshine and Tyler 2003, p. 514).Here, 'fairness' has several key dimensions, which for present purposes can be reduced to two core issues: quality of decision-making, which denotes impartiality, objectivity, transparency, trustworthy motives, and participation in the decisionmaking process, and quality of treatment, which includes honesty, dignity, and respectful treatment (Thibaut andWalker 1975, Sunshine andTyler 2003).PJT, routinely supported by empirical evidence, has been shown to explain public cooperation with the police, compliance with police officer directives and reductions in offending behaviour (e.g.Paternoster et al., 1997, Tyler and Huo 2002, Tyler and Wakslak 2004, Trinkner et al. 2018).This is largely due to the impact of procedural justice on legitimacy judgements.Tyler (2006, p. 375) conceptualises legitimacy as the property an authority has when others feel obligated to defer to its decisions and rules voluntarily, rather than out of fear of punishment or anticipation of reward.
The problem is, we know very little about how these arguments might relate to modern forms of policing, operating outside the context of traditional patrol interventions (where much of PJT has so far been constructed).Virtually no studies have investigated PJT in the context of multi-agency offender management programmes involving police officers engaged in a mixture of rehabilitative and crime control activity.This omission is important, as the content and antecedents of legitimacy, as well as the implications for the behaviour of the parties involved, may be shaped by factors specific to this particular setting.
Integrated Offender Management (IOM) is one such multi-agency offender management programme.IOM involves police officers working closely with Prison and Probation services and Criminal Justice Intervention Teams (drugs workers) in an effort to change or control the criminal activities of acquisitive recidivists.The work is done through a combination of rehabilitative and restrictive/enforcement-orientated interventions.Within this framework, the police take a 'carrot' and 'stick' approach to the management of scheme participants.Plain-clothed officers, deployed as 'police offender managers', gather intelligence and monitor people for signs of reoffending.At the same time, these officers attempt to draw participants away from crime, working alongside the other agencies, facilitating access to drug services, education, employment and transitions into stable housing arrangementsthe 'carrot'.Where there is evidence that a person is failing to comply with licence conditions, or engage with IOM positively, traditional catch-and-convict policing methods are employed by a distinct set of uniformed patrol officersthe 'stick'.
Previous research on IOM has examined whether the changes this form of policing entailed were reflected in changes in underlying police cultures.It has come to differing conclusions.Cram (2020Cram ( , 2023) ) carried out an extensive observational study of police practice and culture at one English IOM unit.He found that rehabilitative activity carried out by police offender managers not only failed to displace orthodox police cultural attitudes towards IOM participants, it also acted as a cover for the pursuit of traditional 'catch and convict' policing goals (2020, p. 392).However, when Annison et al. (2015, p. 398) examined policy and practice, within Thames Valley IOM, they discovered that a number of police officers, working within this setting, had moved away from traditional cultural practices, instead adopting values and ways of working more akin to that of probation.Work by Sleath and Brown (2019, p. 11) also revealed a similar 'cultural and practical shift' had taken place among some IOM police officers they encountered, again 'towards probation'.
A handful of studies on IOM have also included empirical exploration of how participants perceived and experienced this sort of policing activity,particularly the police contribution to IOM.The findings reveal mixed views among IOM participants.For example, IOM participants managed under a London IOM Initiative, which involved interactions with police officers as offender managers, said this form of police involvement had had a 'positive influence' on their lives (Dawson and Stanko 2011).However, IOM participants interviewed as part of a process evaluation of 5 IOM sites, conducted by Senior et al. (2011), found relations with police offender managers challenging, particularly when working with officers who had arrested them previously.Nevertheless, those who were able to form working relationships with officers, reported that a high level of interaction with the police generated a strong motivational influence and a positive impact on their chance of desistance from crime.Moreover, their view of the police had changed (positively) as a result of working with offender managers in IOM (Senior 2011, p. 34, Criminal Justice Joint Inspection 2014, p. 45).Further research, by Sleath and Brown (2019, p. 522) on IOM, also captured a 'shifting perception' of police officers in the eyes of the IOM participants they interviewed.Here some participants had developed a more positive view of the police organisation and a greater sense of trust in police officers.
While this research helps us understand better the culture and practice of IOM police, as well as participant thinking around the broader notion of working with IOM police, it does not directly address the possible relevance of PJT to participant experiences and perceptions of IOM policing.
IOM is a policing setting in which police/citizen/suspect interactions are framed very differently to those of traditional police patrol/response work, where contact is, sporadic, and in which suspects do not necessarily have a history of offending.In the context of IOM, officers are proactively targeting persons known to be involved in criminal activities and dealing with them in a variety of ways that include and also go beyond traditional arrest and convict practices.The last decade has seen the proliferation of IOM schemes across England and Wales.As of 2020, 39 out of 43 police force areas reported implementing an iteration of the IOM model as part of their core policing offering (Home Office and Ministry of Justice 2020).By the end of the same year, the UK government had positioned IOM as central to its 'neighbourhood crime' strategy (Home Office and Ministry of Justice 2020).IOM is now ubiquitous in the policing landscape of the UK.Given the sharp focus of IOM on encouraging a voluntary predisposition towards law-abiding behaviour among participants, a particularly germane inquiry seems to be how PJT might operate within this settinga setting that differs significantly in context to that of other procedural justice studies to date.
The article proceeds as follows.In the first section, I examine prior work done on procedural justice and police legitimacy, with an emphasis on its corollaries among those that have frequent contact with the police.These ideas are positioned within Bottoms and Tankebe's (2012, 2017, 2021) dialogic concept of legitimacy.The second section describes the characteristics of the sample, the data collection and method of analysis.Next, in the third section, I present IOM participants' responses, capturing their experiences of IOM policing.The fourth section discusses the implications of the findings, whilst the fifth and sixth sections offer some concluding thoughts on the direction of future research and the limitations of the present study.

Literature review
The basic argument of PJT is that a consistency of procedural fairness (fair decision-making and respectful interactions) during police-citizen encounters helps the police strengthen legitimacy; in turn, officers are more likely to receive support and voluntary compliance from members of the public (Tyler andHuo 2002, Tyler 2006).Instrumental evaluations, such as whether the police provide a credible risk of sanction for rulebreakers (deterrence), effectively manage crime and disorder (effectiveness) and fairly distribute resources and outcomes, across individuals and communities (distributive fairness or justice), have also been found to be important predictors of police legitimacy, although, generally less influential than perceptions of procedural justice (cf.Tankebe 2009, Karakus 2017, Lee and Cho 2021).Sunshine and Tyler (2003, p. 534), for example, tested data from a survey of, and telephone interviews with, registered voters in New York City.They found that procedural justice had a 'sweeping influence' on legitimacy judgements and that legitimacy had a linear relationship with compliance.Police effectiveness and distributive fairness also had a significant effect on legitimacy judgements, but less of an impact on cooperation (2003, p. 529).Similar findings were observed in a further study using a sample of adult Californians (Tyler and Huo 2002) and later again, by Tyler (2006) in more survey data from Chicago (also, Reisig et al. 2007, Tyler and Fagan 2008, Tyler et al. 2010).

Procedural justice theory and recidivists
A small but growing body of literature offers insights into PJT in the context of encounters between the police and persons who have a history of offending and frequent prior police contact.Again, the primacy of perceptions of procedural justice over instrumental concerns is broadly confirmed in respect of pathways to police legitimacy.Significant links between procedural justice and perceptions of police legitimacy have been found across a diverse sample of recidivist offender types.
For example, robbers, burglars, drug offenders and those convicted of violent crimes (White et al. 2016), early-onset and adolescent-onset offenders (Augustyn 2015), domestic violence perpetrators (Paternoster et al. 1997), convicted adult and adolescent offending populations (Piquero et al. 2005, Fagan andPiquero 2007), active gun participants in Chicago (Papachristos et al. 2012) and young men (aged 18-26) frequently and intensively exposed to a regime of police stops in New York City (Tyler et al. 2014).Other studies (also small in number) have shifted the focus to policing settings that are qualitatively different to those in which the vast bulk of the PJT research evidence has been generated.Moreover, the work also employs ethnographic methods, unlike much existing research on PJT, which emphasises large-scale survey data and is thus in danger of 'portraying a reified social world divorced from the social psychological dynamics of encounters between the police and policed' (Radburn and Stott 2019, p. 421).This research has yielded complex and partly unexpected findings relating to PJT.Savigar-Shaw et al. (2021), for example, explored the capacity of PJT to explain regular and dynamic police-citizen interactions in police custody suites.Counter to assumptions regarding the causal ordering found in the established PJT literature, procedurally fair treatment did not constitute a straightforward mechanistic route to engendering compliance, but instead was also a 'gift' of instrumental powerholders (i.e. the police) that flowed from existing compliance and cooperation of suspects.Here the relationship between procedural justice and compliance was bidirectional.Furthermore, the data revealed an intersectionality between the four established dimensions of PJT: trust (i.e. the police explain the motives for their actions in ways that signal concern for the people they encounter), respect (i.e. the police are courteous toward people and their rights), neutrality (i.e.police do not exhibit bias and are consistent in their application of rules) and voice (i.e. they provide people an opportunity to explain their perspective on the incident and how it should be dealt with) (Tyler 2013).Notions of trust, respect, neutrality, and voice overlapped and had different meanings in context, rather than operating merely as 'stand-alone' rules to be applied mechanically to achieve a procedurally fair interaction (Savigar-Shaw et al. 2021, p. 788).
Research done by Kyprianides et al. (2021, p. 685) on whether the tenets of PJT applied to people experiencing homelessness in London, also reveals data apparently disrupting the procedural justice -> legitimacy -> compliance/cooperation pathway.Fairness was important to the homeless in their interactions with the policei.e. they wanted to be treated with dignity and respect but it did not mechanistically procure 'self-regulated', law abiding behaviour (2021, p. 865).Distributive justice and instrumental concerns mattered more to participants.Proportionality was important in respect of police treatment; deterrence (and to an extent personal morality) set the limits of serious illegality; and benefits associated with non-compliance in relation to (minor) offences (survival) were weighed against the potential costs (starvation/death).Other factors such as the socio-structural position of the homeless, their historical policing experiences and the temporal and spatial context within which interactions with the police happened were also significant in shaping participant perceptions of police action and its legitimacy (2021, pp. 684-685).Further analysis of the same data set by Kyprianides et al. (2022) this time with a sharp focus on whether police procedural justice, effectiveness, legitimacy and perceived risk of sanction were predictors of compliance with the law among the homelessreaffirmed many of these assumptions.
The implications of the studies discussed above are twofold.First, they provide continued support for the proposition that what the police do matters less to people than how they do it, in respect of legitimacy judgements and future compliance.Whilst this is an assumption largely confirmed among criminally involved populations, in some dynamic social contexts the causal relationship between perceived procedural justice, legitimacy and compliance may be less than straightforward.This was in evidence in the settings of police detention and the policing of an extreme marginalised population (the homeless), where relations between powerholders the (police) and audiences (suspects) are an embedded process, not simply a one-off, ahistorical interaction.The second inference from these studies is that context is important.It can facilitate reciprocal relationships between procedural justice and legitimacy and/or result in participants emphasising notions of distributive fairness, police effectiveness and deterrence as significant predictors of law-abiding behaviour (more so than procedural justice).Effectiveness and distributive fairness are, of course, embraced within the PJT concept of quality of decision-making, which stresses normative expectations of policing within democratic societies.These expectancies would include police effectiveness in tackling localised crime and disorder and the equitable distribution of police resources and the outcome of their actions (or inactions) throughout the community (Tankebe 2013).

Legitimacy as a dialogic and dynamic concept
Despite broad empirical support for procedural justice, elements of the theory remain contested.For example, Tyler's (1990) original view of legitimacy (an obligation to obey) has since been variously disputed, with recent meta-analysis, by Walters and Bolger (2019), confirming empirically that the two concepts are distinct.Tankebe (2013, p. 124) has also drawn attention to the 'danger of assuming that feelings of obligation to obey legal authorities are always normatively justified'.People might report feeling an obligation to obey the police because they recognise that the police have a normatively valid claim to power/authority, but equally they might say they do so out of fear of the consequences of non-obedience or 'dull compulsion' and instrumental calculations.Obligation, according to Tankebe (2013, p. 105), should therefore be considered as a 'dependent variable', sometimes a consequence (rather than a constituent) of perceived legitimacy, sometimes not.He further suggests that police legitimacy is best understood not as a single transaction between the police as powerholders and the public as their audience, but as a perpetual 'discussion' between the two (also, Bottoms and Tankebe 2012, p. 129), with police officers making claims to (i.e.attempting to establish and cultivate) legitimacy and the public responding to this claimsupporting or rejecting it.Police officers may adjust their claim in light of the audience response, and the process continues (2012, p. 129).Within this approach the meaning of legitimacy is taken from a definition of legitimacy offered by David Beetham (2013, p. 19, emphasis in original): Legitimacy is power that is acknowledged as rightful by relevant agents, who include power holders and their staff, those subject to the power and third parties whose support or recognition may help confirm it.

Dialogic legitimacy and IOM
Dialogic legitimacy has yet to be applied to a multi-agency policing setting.But it makes sense to study IOM police legitimacy as a dynamic interaction among officers and scheme participants because IOM police-participant relationships are typically lasting.Over time, it is possible that the dynamics of these relationships change for both parties, partly based on the police officer's changing sense of authority and power and partly based on the participants' reactions to IOM police claims to power and authority (also, Tyler and Jackson 2013).Tankebe (2013) and also Bottoms andTankebe (2017, 2021, p. 90) offer some suggestions as to the possible substance of the claims and responses.To do so, the authors transfer Bernard Williams' (2005) concept of 'Basic Legitimation Demands' into the criminological literature in the form of 'Basic Legitimation Expectations' (Bottoms and Tankebe 2021, p. 87).These 'expectations' are conditions that citizens consider will allow them to acknowledge police power as 'rightful', and thus will reasonably anticipate police officers to address within the dialogue if their claims are to be viewed as legitimate.Bottoms and Tankebe (2021, p. 75) point to several requirements operational policing should usually satisfy so as to provide a justifiable response to citizens' expectations.Requirements might include police officers acting fairly (procedurally or distributively) and/or performing their duties effectively and/or behaving lawfully.A further possibility is that different citizens might have varied basic legitimation expectations of police behaviour, partly because of a particular social context in which they exist, their social status and/or their historical relations with police.This line of theorisation may help us better understand perceptions of police legitimacy in IOM.

Data and method
Data for this study came from a larger study 1 investigating the behaviours and operational practices of police officers working within one IOM unit ('Sunnyvale') in a single English police force.The research also examined how IOM was experienced by those subject to it: IOM participants.The present investigation is based on interviews with IOM participants, conducted during ethnographic fieldwork (n = 400 h), carried out between September 2012 and October 2013.
IOM participants interviewed (n = 20) had the following characteristics: aged from 18 to 39 years (m = 29.5,S.D. = 6.143032), female (n = 3), male (n = 17), self-identified as Black British (n = 1), selfidentified as White British (n = 19), unemployed (n = 20), in custody (n = 10), in the community (n = 10).Each participant had been selected for the Sunnyvale IOM scheme through an 'Offender Group Reconviction Scale' Version 3 (OGRS3) and an 'Offender Assessment System' (OASys)actuarial software, which estimates the probability that individuals will be reconvicted within two years of sentence or released, if sentenced to custody (Howard 2009).Respondents were recruited with the assistance of Sunnyvale workers, but not Sunnyvale police officers.Staff were asked to identify 20 participants willing to discuss their experiences of the scheme with an independent, academic researcher.My aim was to interview Sunnyvale participants who were currently engaging with the scheme.In this sense, the sampling was not intended to be representative of all people currently targeted by Sunnyvale IOM. 2 Half the selected participants (n = 10) resided in the local community, with the remainder (n = 10) in custody.The risk profiles of the participants I interviewed were not made available to me during the study.Participation in the study was voluntary and, at the start of each interview, respondents were provided with detailed information about the research, assured confidentiality and anonymity and required to sign a consent form.Participants were also told that they could stop the interview at any point and withdraw their consent to further use of the data.Respondents in custody were interviewed on the prison estate, individually and alone in a closed room, with no other persons present.Community-based participants were interviewed in coffee shops.Police interview rooms or probation office consultation suites were seen as unlikely to create the environment required to capture free and frank perspectives on IOM policing.
Each participant was interviewed once and with agreement their interviews recorded.Interviews were on average 1-hour in duration.After each interview was completed, the data were transcribed verbatim using only the audio recordings, with participants anonymised and given pseudonyms.The data was then securely stored, as per university regulations, for later analysis.
Interviews were semi-structured wherein I prompted detailed core information from each participant, while also probing more deeply into their experiences of IOM policing.A series of open-ended questions explored their routine interactions with Sunnyvale police (both support arrangements and enforcement practices), the treatment they had experienced, the role of the police within society, whether participants would help (or seek help from) the police and how fairly they felt they were treated during routine interactions with IOM police officers.
Where participants reported 'fairness' and 'unfairness' in the treatment they had received, I asked a series of follow up, open-ended questions to try and understand aspects of these experiences e.g.whether Sunnyvale police explained reasons for (intensive) police contact, whether participants felt listened to by officers, whether they believed Sunnyvale police wanted to (and could) help them, whether police contact seemed proportionate, whether officers were polite, or physically and/or verbally abusive and so on.To explore participants' concepts of police legitimacy, I asked what (if anything) respondents thought gave Sunnyvale police 'the right' to intervene in core aspects of their lives in order to conduct the intensive monitoring and supervision to which they reported being subject.Specifically, I wanted to find out whether or not respondents felt the scheme required a compulsory element as a means of encouraging participant engagement.These markers of perceptions of fairness and legitimacy differ slightly from those typically emphasised by PJT, and this is in part the result of the enduring and dynamic nature of police-participant interactions IOM policing generates.
Data were coded along the dimensions noted above and analysed by a single coder.Whilst such an approach has the advantage of avoiding the potential for the imposition of multiple frames over the data, it is important to pause here and reflect on the steps taken to establish interrater reliability.This was done through a process of 'peer debriefing'a consensus methodwhere an external recheck of the data is carried out by a debriefer so as to strengthen the validity and reliability of the data (Creswell and Miller 2000).In this case, it was the research supervisor who reviewed the data and analysis, challenging preconceived assumptions/biases, considering alternative interpretations and managing subjectivities.
Finally, analytic abduction is used as a methodological approach to develop the empirical analysis and contribute to theory refinement.This method of inquiry exists alongside deduction and induction but is also particularly sensitive to novel and/or surprising findings and their ability to extend, advance, or revise established theories (Halpin and Richard 2021).In the present case, the analysis is informed by the extensive research on PJT.
Excerpts of participants' responses were extracted in order to illustrate the themes that emerged from the data.As noted above, participants were asked to report on their experiences with IOM police officers.Perhaps inevitably, some recounted more general experiences of policing likely related to historical encounters with uniformed police not involved in IOM.Nonetheless, care has been taken to ensure that the thematic findings, unfolded below, exemplify the most common patterns in their accounts of policing specific to the context of IOM. 3

Fairness as trustworthy police motivations, underpinning the distribution of resources
Reflecting on interactions with Sunnyvale police, IOM participants described a range of personal experiences drawn from a variety of settings and contexts.From the enforcement practices of uniformed IOM police officers, tasked specifically with disrupting the criminal activities of IOM participants (the 'stick'), to home visits conducted by IOM police offender managers, tasked with intelligence gathering and the provision of rehabilitative support (the 'carrot'), their experiences covered the breadth of Sunnyvale's police functions.
Eighteen out of twenty IOM participants reported extensive personal experience with uniformed police in stop, account, search and arrest type situations.These involuntary contacts, which likely included encounters with both uniformed IOM police and uniformed police not operationally linked to IOM, were felt by participants to be of excessive frequency.As Mike complained, 'I got stopped three times in a week by three coppers.They watch me, follow me and everything … '.Another man, Darren, provided the following example: I got stopped four or five times on my road.It might be a long road, but I've been stopped by one set of police officers and then walked down the road a bit and like another set of police officers pulled me over.I've said, "Look I've just been pulled over".They're like "Oh well" and started laughing at me … like they think they're funny.
Participants also reported instances of police offender managers 'turning up' at their homes uninvited: I got out from the courthouse and went to my girlfriend's house.They were waiting for me.I wasn't expecting them.They just turned up and said, "We're police, working with probation and we're here to help you.But if you step out of line, we'll send you back to prison before the courts will.Make sure you don't step out of line.We're keeping an eye on you".
IOM police interactions were thus not just unwelcome, because of their inordinate frequency, they were also experienced as intrusive.Yet, whilst it was common for participants to express confusion, frustration, and distrust as to why they were targeted, both patrol officers and offender managers have significant capacity to explain their actions by reference to the mechanisms used by the scheme for the risk-management of participants.Risk profiles of Sunnyvale participants are measured according to their current risk of re-offending and colour coded according to a dynamic Red Amber Green (RAG) system.Participants are deemed 'Red' where current intelligence signals they pose the greatest risk of re-offending, whereas 'Amber' denotes a lack of police knowledge about the degree of offending activity a person might be concerned with, but 'professional judgement' suggests that the participant is involved, or at risk of being involved, in crime.'Green' participants are those not believed by Sunnyvale police to be committing crime and are engaging in rehabilitative efforts.Interviews with patrol officers confirmed the importance of participant risk profiles in generating Sunnyvale police action: Their status puts a target on them.We go through the briefing [files] for the whole of [Sunnyvale].People that are highlighted, suspected of committing offences, we will specifically target areas where we're likely to see those people.
Accordingly, high-risk participants, who routinely spent time in 'hot spots' known for criminality, were likely to encounter Sunnyvale patrol officers deliberately targeting places they believed risky people might be found.This type of police activity is in line with the 'Stick' aspect of IOM policythat is to provide 'an enhanced level of surveillance and control in relation to those participants identified as at risk of reoffending' (Criminal Justice Joint Inspection 2014, p. 35).However, Red Sunnyvale participants were also in something of a dilemma.They had 'previous form' and were considered by the police 'more likely to be at it' and therefore needed to have an 'eye kept on them', as one patrol officer put it.Thus, they received high levels of police attention, both from patrol officers (on the streets) and from offender managers (home visits and probation meetings).But increased attention also raised their profile during police 'briefings' (where offender managers funnelled intelligence to patrol officers), resulting in more police contact.
Once they know who you are, they're definitely more suspicious of you.Since I've been on this scheme, I've been told that my picture is on a wall in a police station.Each day or whatever the police talk with probation to see if there is any intelligence on whether these people are actively thieving or using [drugs] and I worry about that.I could be walking around and [uniformed] police know I'm on this scheme.So, that puts me in a bracket with them.
The extent to which a person was known to police and/or risk-management strategies adopted by IOM might explain variances in regularity of contact between uniformed Sunnyvale police and participants.Moreover, the Sunnyvale participants I encountered were nearly all unemployed, impoverished and/or of no fixed abode.As a result, they often felt compelled to occupy public spaces, increasing their 'availability' to the police (Waddington et al. 2004).Nonetheless, those on the receiving end of the police practices did not make these connections.Rather, interviews captured a perception among most participants that they were victims of authoritarian policing, untethered to organisational rules and more closely tied to personal discretion.In short, participants distrusted police motivations when subject to what they characterised as unwanted street stops.'They don't like me.Every time they see me, they just pull me over', one man, James recalled, whilst Neil complained that uniformed police regularly stopped and questioned him for 'no good reason'.
The procedural rules justifying (or restricting) the pre-emptive targeting of Sunnyvale participants are based on wide ranging information on participants' conduct, and whilst the knowledge may not include clear evidence of offending, it should be rooted in intelligence (Police Operations Guide 2010).For Sunnyvale police, therefore, the action that follows is not seen as arbitrary.'We go wherever the analysis, and the intelligence, tells uswe're not after easy hits', a uniformed patrol manager explained.This perspective is a central aspect of the dialogic construction of legitimacy, wherein the police are continuously making claims to legitimacy and, as Bradford (2017, p. 60) observes, 'evidence-based policing is nothing if not a claim to legitimate authority over the realm of crime control'.However, Neil's complaint implies he experienced some police encounters as arbitrary because, from his perspective, the intervention was not a response to evidence or suspicion of offending.
This disjuncture between police and participant views on what is driving IOM police interventions is again captured in the following interview extract.Here a Sunnyvale participant criticises the police for carrying out a series of street stops, initially understood to result from a form of risk-based targeting.Later it becomes clear that the stops were influenced by evidence-based suspicion and the man is more accepting of the interventions: Peter: Whenever I went out, I got stopped.Happened six times in one day and kept happening all through a period of time.All I can think is they've red-flagged me.Went on for about a week.Interviewer: What were the average times between stops?Peter: Within hours of each other.When the next police car went past or one would go past and the next one would stop me, or they would go around the block and stop me.One time they said they'd been told like by a member of the public that I was walking about with a charity box … another time [they said] 'people having been ringing in say cars are being broken into'.I was like OK these are reasons to stop me.
As these responses reveal, perhaps most strikingly in Peter's comments above, risk-based targeting by patrol officers (i.e. the RAG system, noted above as the basis for stops and visits) was seen as less acceptable than interventions based on firm evidence of past offending.Procedural justice denotes a perception of objectivity and openness, as well as opportunities for representation and correction of mistakes made in respect of police decision-making (Tankebe 2013, p. 111).However, in the first instance Peter describes, the police decision-making process underpinning the intervention is effectively concealed and not prone to challenge or accountability mechanisms.The activity was thus experienced by Peter (and other Sunnyvale participants) as a form of harassment rooted in subjective (police) discretion.In the second example, however, Peter seems to accept the intervention, with his comments implying a measure of cooperation.The difference was that the police gave a reason for the stop.This addresses several dimensions of procedural fairnessopenness (Peter is informed of the reason for police activity), motive-based trust (the reason givensuspicion/evidenceis seen as a good one), and respect (Peter's right to know the reason is respected by Sunnyvale officers)seemingly supporting Peter's acceptance of the encounter as a legitimate use of police authority.This finding indicates a clear difference between police activity Sunnyvale participants were prepared to accept and that which they were not.At the core of the distinction seems to be the trustworthiness of police motives.This concept was again emphasised in participant concerns raised about the quality of interpersonal treatment from Sunnyvale police.

Fairness as trust and respect during encounters with IOM police
It was not simply that Sunnyvale participants objected to being stopped and questioned frequently; most felt demeaned and disrespected by patrol officers during these unwanted encounters.Many reported that patrol officers were discourteous and used provocative language.Spencer for instance said, 'Some call me 'sunshine' and poke me in the chest for no reason and be very antagonistic.One refuse [d] to let me in his car when I was homeless.He radioed another [police] car, saying 'my mate deals with the smelly ones''.Matt, on the other hand, complained that police officers had referred to him as 'scum' and a 'fucking divvy'.Observations confirmed this tendency among both sets of Sunnyvale police officers to describe participants in disparaging terms rooted in stereotypes.People were frequently characterised as 'vile', 'dirty scroats', 'shit-bags', 'crackheads' and (on one occasion) subject to racist behaviours (Cram 2020).These false designations were unproblematically and indiscriminately applied to Sunnyvale participants.However, participants also pointed to incidents of disrespectful police talk coupled with disproportionate and/or inappropriate police action.One man for instance said, 'I've had a couple of hidings off them', whereas a female participant recalled, 'I haven't literally been punched in the face, but once they thought I was a bloke and chucked me over the fence and on the floor and … you know, roughed me up when the handcuffs [went] on'.It was Jason, however, who perhaps provided the most revealing example: When I've been arrested sometimes, I've had police sat on top of me, slap me across the face put their fingers in my face, shout in my face.Been really, really, aggressive towards me and at that specific moment in time, I haven't done anything wrong; I haven't been charged with anything, whether I have or haven't done anything wrong.
Similar to the reactions of Neil and Peter, (discussed above, in respect of street stops), Jason appears to distrust police motiveseven the legality of the conduct.For him, what appears to be of primary importance is not so much the (seemingly accepted) violent conduct of the police (although few would suggest that this is unimportant), but that the conduct lacked any justification in terms of formal process.The actions of the police appear not to be shaped by formal laws and rules, but rather by independently defined objectives that have little to do with law enforcement: imposing discipline, control, and authority on participants (also, Choongh 1997).This is problematic as far as perceptions of police legitimacy (that might otherwise be anchored in quality of decisionmaking dimension of procedural justice theory) are concerned, as again there is a lack of open, trustworthy, and respectful behaviour on the part of the police.

Evidence of instrumental and/or prudential concerns
There was also some evidence that procedurally unjust encounters (i.e.involving disrespectful and physically abusive police behaviour) resulted in deference and cooperation based on a sense of prudential and/or instrumental obligation.Pósch et al. (2021, p. 7) suggest that such a reaction may result from the asymmetrical power relations that police officers have over citizens.Consider Aaron's remarks: I was walking down the road and the police was pulling someone across the road and the copper shouted over: "You been found dead yet?" to me, "You been found in a field?"And I said, "Fuck off" and he ran straight over and grabbed me and said, "Say sorry", bending my arm back.I'll nick you in a minute".So, ended up saying sorry, rather than getting [arrested].
Negotiation of power balances within typically adversarial relationships between participants and the police moved beyond the streets and into the homes of Sunnyvale participants during visits from offender managers.Home visits provided an opportunity to 'glean information and intelligence' from participants, but also 'inform them about the scheme and see if they're willing to engage', Brian, a police offender manager explained.Nonetheless, the visits also reaffirmed unequal police-participant power dynamics: police officers were to be treated with deference, whether in uniform or not.As Martin's comments illustrate: If [an offender manager] is asking questions, then you're expected to answer them.They put that on the table from the off.It's like if you've got nothing to hide then you've got nothing to worry about.So, if he's asking questions, I'm expected to answer them.If I don't then I appear as if I'm being awkward.So, either way I'm going to do it.If I was awkward and difficult and said "Oh I ain't talking to you", that would have consequences.I can't do that because he definitely governs whether I stay out of prison or not.If I'm difficult, he's going to think, "Well he's obviously not trying to help himself".
The imbalance problematised by Martin mirrors one found in more formal circumstances, where questioning might take place, such as the police station (also, Choongh 1997, p. 83).Yet what he describes took place within the comfort of his own home.Even in this private space, he reports a sense of powerlessness in the process and a resigned need to acquiesce (similar to Aaron, documented above).This reaction is consistent with what Tankebe (2013) would describe as 'dull compulsion' (also, Carrabine 2004, in the prisons context) and appears to underpin Martin's inclination to cooperate with police offender managers.
Largely, however, physical, and verbal abuse was not viewed by participants as a legitimate use of police authority, as evidenced by the resulting negative downstream effect on the willingness (of those on the receiving end) to work with the policeparticularly offender managers.Spencer (also extracted above) for instance said: These schemes are supposed to help people, but how can a police officer help me?The only thing they've done for me over the last couple of years is kick my door off its hinges.I won't speak to them … I don't trust any of themacross the board.
The unwillingness of Spencer to work with Sunnyvale police offender managers appears to stem from a lack of trust in the motives of Sunnyvale police, but his comments also illustrate a concern over the transactional benefits (or lack thereof) of future encounters (i.e.what can police offender managers do for me?).Echoing these sentiments, another man, Lee, also questioned the effectiveness of this type of police activity: Who the fuck wants to chat to a police officer?Come on let's have it right; do you honestly think?You know what they said to me? "Anything you say to me is confidential".You're a fucking police officer.Do you honestly think you're going to chat to me mate?Get fucking real.Stop wasting your time.
Disrespect (in the form of physical and verbal abuse) was thus firmly tied to the perceived illegitimacy of Sunnyvale police action.The result of such conduct, on the part of the police, was that some participants were uncooperative with and/or were hostile to uniformed police, whilst others seem to cooperate out of a sense of pragmatism in the face of powerlessness.Past experiences of hostile and disrespectful uniformed policing also seemed to harm relations (or potential relations) between Sunnyvale participants and police offender managers.Simply put, the gloss of the police offender manager approach appeared to have done little to allay the posture of distrust participants typically felt towards the police.These findings provide further support for previous work done on the consequences of procedural [un]fairness during traditional police encounters for perceptions of police legitimacy (e.g.Tyler 2006, Reisig et al. 2007, Tyler and Fagan 2008, Tyler et al. 2010).But just as Sunnyvale participant notions of legitimacy were jeopardised by aggressive and disrespectful encounters with police (that from their perspective were unjustified), 'fair' or indeed legitimate interventions were again premised on the perceived trustworthiness of police motives and respectful police behaviour.Consider, for instance, the following exchange: Interviewer: Are you treated fairly, by [Sunnyvale] police?Emma: Yeah, they're good, sometimes.Interviewer: Can you give me an example?Emma: We'd nicked something.Got home, sold it.Police pulled up and said, "What you been doing today?"I'm like, "Ah shit".But they bailed me on the street [and] said, "Look, come on up [to the police station], we'll sort it out".
The 'bail decision' was viewed as fair because it meant postponing the formal exercise of police powers (detention) until a later date.For Emma, a heroin user, street bail also nullified the immediate threat of experiencing withdrawal symptoms whilst in custody.As Emma continued to explain: Whilst Tyler (2006) has argued that fair treatment is more important to people than the actual outcome of interactions with the police, in this instance the outcome (not being detained) was of paramount importance to Emma and thus, in part at least, determined her perception of fairness.But relational elements were also present, i.e. flexibility (within clear boundaries), honesty, understanding and apparent concern for her wellbeing on the part of the officer.For the patrol officer the decision, may respond to the implicit 'carrot' of the IOM approach: participants known to be engaging with support offered by Sunnyvale will be treated less robustly by patrol officers they encounter (also, Annison et al. 2015).Informal bargaining, in this instance, achieved objectives for both police officer and participant.
The overlap between instrumental concerns/benefits and aspects of procedural fairness was also to be found in participant accounts of relations with police offender managers.Marc, for example, said: When I came out of prison, a plain clothed police officer, come up to me at the gate and said, what's your plans?I told him I was going for a probation order.He said if you need help with anything give me a call.Said he'd be in contact with probation and keeping an eye on me.I found it helpful.Even though he's a copper.If there's help there for me, which there is with [Sunnyvale], then it's all good.
Similar to Emma's account, it appears Marc's perception of the fairness of the intervention also stems from the 'soft' way in which the authority was exercised.For him, what was valued was the straightforward, honest approach of the police officer, as well as instrumental returns (i.e.support).The latter seemingly wrapped up in a sense of trust and confidence in the motives of the police officer: a belief that the offender manager cared genuinely about Marc's welfare and improvement.But this sense of motive-based trust, so pervasive throughout participant accounts, played an important part in Sunnyvale participant legitimacy judgements.In the final empirical section, I turn to this issue of trust, attempting to tease out what it represents and its relationship to participant perceptions of Sunnyvale policing.

Further emphasis on the trustworthiness of IOM police motives
Something about the trustworthiness of police motives was especially central to participant accounts of Sunnyvale policing.This was perhaps most evident in one of the most striking and consistent findings in the research: participants understood Sunnyvale police interventions through the lens of their criminal behaviour.Thus, many, like Steven, accepted the inevitability of intensive police action: My home and private life could get invaded at any time of the day.That's how I just accepted how things were.Certain places I was going to, they were known houses anyway.At any time, the police could turn up and we'd just scatter or just sit there and see what would happenprobably get searched.
Other participants more directly related their offending history to the amount and type of police attention they were likely to receive.It was this link between their offending record and police interventions that in their view justified some Sunnyvale police operational practice.As Matt put it, when asked what, if anything, gave Sunnyvale the right to intervene in his life: I understand it.I've got 40 previous for burglary dwellings and I've had 6-year sentences, 5-year sentences, I've had this 4-year sentence just gone, 3 and a half years before.I've had some big sentences; so, I can understand it.He's just a flat-out burglar.A lot of police are one track mindedonce a criminal always a criminal.
Another man complained of 'loads of stops, [Police National Computer] checks, searches, being pushed against the wall, slammed to the groundstuff like that'.However, when pressed on what, if anything, justified the coercive and invasive Sunnyvale police action, said: 'My behaviour.The only way I can put it.I am a madman when I'm like that.I don't know myself.You know what I mean?I can understand it fully'.Another participant Paul, echoed this view, during the following exchange: Interviewer: You said you've had a lot of experience of Sunnyvale policesurveillance, police offender managers turning up to probation appointments, being stopped, searched in the street and so on.What do you think justifies, or say, gives them the right to intervene in your life in that way?Paul: Because I'm an offender, prolific.Police are worried when I get out, what am I going to be up to.If I wasn't up to something, up to no good, then it would bother me.
Whilst, for one or two participants, the link between previous offending and Sunnyvale police practices was non-existent or tenuous, most imposed this frame upon their treatment as a way of making sense of their experiences.The roots of this phenomenon can be traced back to literature that suggests the police are seen as legitimate when people believe they represent a proper moral purpose (Beetham 1991, Sunshine and Tyler 2003, Tyler and Wakslak 2004, Bradford et al. 2014).Thus, whilst the distribution of IOM police resources, and quality of interactions with Sunnyvale police (particularly uniformed cops) was perceived by participants to be 'unfair' when aspects of police practice strayed beyond formal (legal) and informal (police discretion) 'rules' that defined the boundaries of ongoing relations, the trustworthy motives driving police action (responding to offending) was accepted to a degree as justified.

Discussion
PJT emphasises fairness as a central factor in shaping evaluations of police legitimacy.These judgements are built on perceptions of two aspects of police authority: quality of decision-making and quality of treatment.The data discussed above provides a clear-sighted view of how participants experienced encounters with Sunnyvale police officers.Whilst a small minority reported positive interactions with both sets of IOM police officers, the majority complained of systematic and repeated targeting by Sunnyvale uniformed police, who regularly stopped them on the street, but also of unsolicited visits to their homes by police offender managers.Overall, this 'routine hassle', a form of 'communicative surveillance' (Lister et al. 2008), precipitated a sense among participants (regardless of age, gender, and race) that they were continuously monitored.As one man observed, 'I was thinking, fucking hell, they're watching me'.Disciplinary practices (e.g. the use of formal powers to extract deference from participants (Choongh 1997)) were also adopted by uniformed officers who could be aggressive and verbally abusive.
PJT highlights a linkage between fair and respectful policing and police legitimacy (e.g.Tyler 2006), but in an apparent departure from this assumption most participants placed less emphasis on the way officers dispensed authority and more on their own criminal behaviour when considering the legitimacy of Sunnyvale policing.As one man explained, 'You give them the right by doing what you done … I'm an offender, I've re-offended, re-offended, re-offended'.Physically and verbally abusive treatment at the hands of the police was seen as an unwelcome 'occupational hazard' which, alongside intense police attention, required management and negotiation as part of ongoing police-participant relational work.It was also understood and accommodated as a predictable aspect of a prolific criminal lifestyle.But this did not mean that participants believed their offending history legitimised all forms of Sunnyvale police action; there were limits.Indeed, there was evidence that participants cared a great deal about fair and respectful treatment during interactions with police.This reveals a paradox in Sunnyvale policing and its relationship with the past and exposes a tension between police and participant views on the legitimacy of Sunnyvale policing.Sunnyvale police believed risk assessments were a legitimate basis for focused intervention, whereas some participants (e.g.Peter) suggested that only interventions responding to evidence of recent offending were justified.Nonetheless, when asked, 'what gave IOM police the right to intervene in their lives to this extent', the same participants suggested that the 'right' (a degree of legitimacy) arose from their prior offending and the risk of reoffending it denotes.
So, how was this tension managed during police-participant interactions?What behaviour did prior offending justify in the eyes of participants and what did it not justify?In answer, we might turn to participant perceptions of the fairness of Sunnyvale police action.These views were bound up with both the quality of decision-making and the quality of treatment dimensions of procedural fairness.Aspects of police conduct, such as straightforwardness, honesty, the relaxation/suspension of formal powers and process and decision-making driven by evidence, were strongly related to a sense of fairness, whilst repeated street stops, unannounced home visits, disrespectful language and disciplinary practices were closely tied to perceptions of unfairness.
Nonetheless, it was motive-based trust (i.e.trust in the perceived motivations of powerholders (Tyler andHuo 2002, Tyler 2006)) that was the primary distinguishing dimension of procedural fairness, which shaped participant perceptions of the legitimacy of Sunnyvale police actionor lack thereof.Its presence was a requirement of the quality and character of legitimation claims made by IOM police if such claims were to be accepted by IOM participants.The street stops and home visits, recalled by IOM participants, can be considered claims about the right of IOM police to make the stop/visit, to question IOM participants, and to expect them to behave in certain ways.These claims were evaluated by participants, whose resulting actions were a result of his or her belief that the officer's claims are legitimate (or not).
We saw that participants questioned and distrusted police motivations when street stops, often infused with hostility and disrespect, were carried out without a perceived evidence-base.Aspects of these initial claims to legitimacy were rejected because such encounters went further than what was accepted by the majority of participants and were viewed as unwarranted personal attacks and outside of the scope of the legitimate use of police authority.These reactions mirror previous research that found aggressive order maintenance policingparticularly where it leaves 'citizens feeling humiliated, violated, or even victimized'can erode perceptions of procedural justice and legitimacy (Gau andBrunson 2010, p. 256, Novich andHunt 2018, p. 64).The result tended to be compliance/cooperation predicated on instrumental concerns.But a history of these encounters, coupled with prior (also largely negative) experiences of uniformed policing outside of the context of IOM that were revealed in some interview responses, also placed stress on future relations between IOM participants and police offender managers, with the latter viewed as untrustworthy.This finding supports the general argument advanced by Skogan (2006) that previous negative encounters with the police reduce perceptions of police trustworthiness.
Alternatively, when stopped or visited at home, for what were recognised as justifiable (and often communicated) reasons (e.g.suspicion of criminal activity, and/or the offer of support from offender managers) most participants evaluated these basic legitimation claims, accepting them as they understood better the context of the action and trusted the motivations of the police officers involved.Also, perhaps, accepting an implicit claim to legitimacy on the basis of legality.For some participants, this perception appeared to contribute to willing cooperation with Sunnyvale police, although, again, prudential and/or instrumental cost/benefit calculations were also operative.
This positioning of motive-based trust, as central to the basic legitimation expectations of participants, may be a result of the context in which these insights have developed.There is a difference between routine (but often sporadic) police-citizen encounters (the main focus of PJT) and enduring Sunnyvale police-participant relations.In the latter context, participants have histories (prior offending), which frame the meaning and scope of police activity (enforcement and ongoing rehabilitative work) for both parties.Here, police legitimacy judgements are not easily anchored in the quality of decision-making dimension of PJT.This is because the concept depends on neutrality and objectivity, which involve IOM police officers treating those they encounter as simply another citizen.Neither Sunnyvale police nor Sunnyvale participants expect that to be true.Rather, both parties accept that participants will be targeted on the basis of their criminal record and intelligence on their risk of re-offending that falls short of reasonable cause to suspect them of criminality.In short, the basic legitimation expectations of IOM participants may be lower than most citizens.Certain key conditions must, however, be fulfilled, and thus the focus of participant basic legitimation expectations of Sunnyvale policing shifts towards the quality of treatment dimension of PJT.Relational aspects of procedural fairness (i.e.motive-based trust, honesty, respect) are likely to have a greater effect on legitimacy judgements, within the context of Sunnyvale, because evaluations of procedural fairness can move beyond perceptions of whether police are following (or breaking) formal laws and rules into the locus of long-term Sunnyvale police-participant relationships.
Research done on legitimacy in the context of prisons and community sanctions (e.g.Sparks et al. 1996, Bottoms and Rose 1998, Bottoms 2001, Franke et al. 2010, Crewe 2011, Rowe et al. 2018) lends empirical credence to this last point.In these settings, judgments on the fairness and legitimacy of (probation/prison) staff action correlated highly with a positive perception of offender-staff relationships.Fairness centred largely on 'good relations', i.e.where relationships possessed aspects of trust, respect, straight-talk, support, honesty, informal relations and the deployment of what Liebling (2011, p. 491) refers to as the 'quiet flow of power'.Where good staff-participant relationships were present the formal authority of staff was likely to be rendered legitimate in the minds of those they supervised.
The work of some prison and probation staff and that of police offender managers is similar: an attempt to bend cynical minds towards a path away from crime.This positions relationships as a key site in which to build legitimacy (McNeill and Robinson 2013).Fairness is an important part of this development, but relational work also requires trust.This again returns us to PJT, which recognises the importance of public perceptions of police motives in the construction of police legitimacy (Tyler andWakslak 2004, Tyler 2006).The argument is that, if people believe the police are sincerely trying to do what is good for the people with whom they deal, this helps shape the belief the police are legitimate.When we extrapolate this from PJT, into the context of IOM, one can postulate that Sunnyvale policingenforcement and/or relational rehabilitative workwas perceived by participants as legitimate because of trust in the motives underlying the action: a response to prolific offending.

Conclusions
Given that PJT suggests that perceived legitimacy promotes law-abiding behaviour, the findings from the present study should be met with optimism.The men and women subject to increased surveillance under IOM did not, it seems, view the action as intrinsically incompatible with legitimacy.Because of their offending records, IOM participants did not expect to be treated like other citizens, but they did expect their relationships with the police to be conducted with a measure of respect.Additionally, while previous offending (a key component of risk assessment) was regarded as legitimatising further police intervention, there was little acceptance of risk-based intervention (e.g.Peter's comments on 'Red flagging').Perhaps this was because there was not enough explanation of how decisions were taken to legitimatise them, or such decision-making was not seen by IOM participants as treating them as individuals.
Nonetheless, the degree of perceived legitimacy afforded to Sunnyvale policing by those subject to the scheme has not resulted in widespread compliance.Fieldwork revealed a scheme perpetually dealing with the same individuals.This raises the possibility that whilst Sunnyvale policing retained a measure of legitimacy in the eyes of participants, it may be disrupting the overarching objectives of IOM: steering prolific participants away from crime and towards social reintegration.Close monitoring and surveillance of Sunnyvale participants may further instrumental law enforcement objectives (e.g.deterring participants from committing offences and/or expediting the process of catching and convicting those that do offend), but may impinge on their sense of autonomy and freedom (core building blocks in the reconstruction of a positive identity required by participants as part of the desistance process (Weaver and McNeill 2010)).In the same way, the hostile and disciplinary policing complained of by participants may reduce opportunities for rapport building between police offender managers and participants, which is important for intelligence gathering, securing compliance with authority and influencing change.
The present study offers preliminary evidence from one IOM area and, although there is some indication that IOM policing differs across the UK (e.g.Annison et al. 2015, Hadfield et al. 2020), other research (e.g.Williams and Ariel 2013) has also cast doubt on the success of IOM in promoting desistance among recidivists.Future studies should, therefore, examine whether IOM participant perceptions of unfair police treatment, even if accepted as an inevitable aspect of a criminal lifestyle, reduces the chances of long-term desistance among this group.For now, however, the empirical enquiry unfolded above provides us with a better understanding of police legitimacy in the IOM context and the role of procedural justice in this regard, as well as insight into which procedural justice components are most important here.It also adds weight to the argument that antecedents of legitimacy and antecedents of procedural justice can vary across contexts.

Limitations
The findings offer several insights into IOM participant perceptions of fairness and legitimacy of intensive monitoring by IOM police officers.But the study does have certain limitations to consider.First, the study relies on self-reporting, leaving open the possibility that participants did not recall or fully articulate the complete depth of their experiences accurately.Second, IOM staff identified respondents for the study.This was done to overcome the logistical challenge of one PhD researcher attempting to approach all IOM participants (n 688) to ask if they would be willing to participate (and then pick 20).This leaves open the possibility that IOM staff (perhaps subconsciously) selected participants who they believed were more likely to express positive experiences of IOM policing.Third, although the data were considered to be of sufficient quality for this examination, the size of the sample (n = 20) reflects the highly localised nature of the study.This raises questions as to whether the findings can be generalised beyond the immediate experiences of study participants.Future studies that address these limitations, among others, will contribute towards advancing our understanding of how those subject to IOM perceive the frequent contact they have with the police officers responsible for their supervision.

Notes
1.For the larger study, I interviewed Sunnyvale police officers (n = 15), probation staff (n = 8), criminal justice intervention workers (n = 3) and prison officers (n = 2).Due to ethical concerns, supporting data cannot be made openly available.Further information about the data and conditions for access are available at Cardiff University Research Data Archive at: http://doi.org/10.17035/d.2023.0289644896.2. At the outset of the fieldwork the Sunnyvale scheme had a total of 688 participants; although, it should be noted that the notion of participation is not synonymous with engagement.3.For a more detailed personal/autobiographical account of the fieldwork and the various challenges I encountered, see Cram (2016).
When the copper pulled up, I'd just scored.I had it in my hand, and I thought, "Oh shit, what am I going to do about it?"And he seen my behaviour change and he knew.He knew but didn't search me.He bailed me on the street.I had to go back at 3 o'clock.[When I went to the police station] he said, "I knew you'd scored, but I wasn't going to search you.I don't condone it, but I know you needed it."