Expanding Mechanisms of Governance: Uncertainty and Risk in Police Decision-Making Strategies in the Pursuit of Specialized Peace Bonds

Abstract

This study examines police officers’ decision-making practices through analyzing how they determine which offenders are candidates for an 810, or peace bond. This legal tool allows police officers to petition the courts for continued surveillance and conditions for offenders postrelease. Little, however, is offered in terms of assessment guidelines on how to make such determinations. As a result, police officers discretionary behaviors and additional legal factors play a key role in these determinations. Our findings advance the idea that “uncertainty” is the central object to be managed, and further complicate how risk is constructed and mobilized by suggesting that risk assessments result in over-precautionary practices.

Policing inherently requires that officers assess danger and risk when dealing with the public. The growth of proactive and precautionary policing initiatives, however, has made it is increasingly common for police officers to play a significant role in determining which offenders are at risk of future criminality. This need is facilitated by an increase in offender management programs that aim to identify if offenders are high-risk post-release. Part of this process requires police agents to make judgment calls and interpretations to evaluate risk, danger, and fear, which often come from experience, expectations, and racial bias (Lecoq et al., 2021; Trinkner et al., 2019; Woods, 2019). Despite the attention given to problematizing risk assessment and policing practices, studies on how these decisions are actually made, and the role of police officers’ discretionary power in these processes remains relatively opaque (Joh, 2017; Langevin & Curnoe, 2014). To date, empirical studies that examine how risk and uncertainty logics operate in preventive crime control initiatives are limited (Heilbrun et al., 1999; Mawby & Worrall, 2004; O’Malley, 2000; Petersilia & Turner, 1993; Lecoq et al., 2021). To narrow the gap on policing practices (Harmon, 2013), there is a need for empirical knowledge, particularly given that these decisions are often made with an absence of training on how to assess risk. Studies that examine these practices increase transparency of policing risk practices, which is imperative to improve accountability measures and better regulate police powers (Lecoq et al, 2021; Joh, 2016; Woods, 2019).

In particular, this study examines the decision-making process used to pursue a peace bond. This legal tool is a post-release sanction used in Canada to regulate offenders deemed at risk of future criminality and is defined in the Criminal Code of Canada under section 810 including various forms. Specifically, we examine 810.1 and 810.2 (referred to hereafter as “810s”), which are “specialized peace bonds” (Doerksen, 2019) used to protect both individuals and the public from “reasonable fear of harm or injury” as opposed to quelling minor social disorders.

Correctional agents, including judges and police officers, are involved in the process of deciding when to pursue peace bonds. This process does not involve a standardized risk assessment tool specifically for 810s, but rather police officers justifying that the offender poses a reasonable risk of harm and future criminality. Our micro-level analysis specifically investigates this process by focusing on and examining the question of how “fear,” uncertainty, risk, and perceptions of threat or danger influence and impact police decision-making pathways. To unpack this process, we asked police officers in multiple Canadian jurisdictions how they make the determination to pursue a specialized peace bond.

Our analysis demonstrates that the decision to pursue an 810 involves strategies that go beyond the legal requirements. Specifically, the initial steps are subjective, as they depend first on the officers’ general sense of the offender, followed by a search for indicators and evidence to confirm these hunches, which are heavily drawn from experience. In cases where not enough indictors are found to determine risk, police officers treat this as evidence of “uncertainty” and, as a result, justify a peace bond. We classify these practices as two distinct types of cases—slam-dunk and sit down—which provide a framework for analyzing decision-making processes and show the disconnect between distinctions of risk and uncertainty. What becomes clear in this analysis approach is the high probability that any offender will be deemed a risk, whether through evidence or uncertainty, leading to the recommendation for an 810. We also find that the perceived likelihood that a judge will approve the application plays a significant role in this decision process; suggesting that the expectations of judges regarding the quality of evidence and knowledge also factor into high-risk designations. Further, our analysis highlights that the administrative process for determining risk is central to understanding how an individual is deemed a risk. This finding, along with the absence of a risk tool to guide this process, illustrates that risk assessments practices are highly dependent on individual agents experiences and the likelihood of success. Thus, our study contributes to unpacking the extra-legal factors that impact how high risk is constructed and demonstrates the role of agents in constructing the risky individual. This study is empirical in nature and, therefore, allows for unique insights on the practices of 810s specifically, and decision-making practices in high risk designations generally that further illustrates the need to regulate discretion in policing practices (See Kerbs et al., 2009).

Determining Risk: Police Perceptions, Decision-Making and Conceptual Overlap

In an effort to regulate decision making, risk assessment tools and logics are commonplace practices used to predict recidivism, inform police response, and manage offenders overall. Despite the goal of risk assessment tools to standardize evaluations of risk, studies increasingly illustrate disparities in the assessment process (See Schaefer & Williamson, 2018). Often, risk tools are applied unsystematically, overlooked, and incorrectly used, resulting in adverse outcomes (Miller & Maloney, 2013; Viglione et al., 2015). Factors, such as limited resources, agents lack of training and interest, as well as data limitations, account for some of these disparities, which all have serious impacts on risk assessment, given that the process necessitates data and knowledgeable use to function (Schaefer & Williamson, 2018; Lecoq et al, 2021).

Even with growths in data collection practices, there are limitations to the quality, access, and scope of police data (See Hannah-Moffat, 2019; Joh, 2016; Ballucci, Under Review). For instance, compliance to data entry in policing is low (Schaefer & Williamson, 2018) and access to offenders’ criminal histories can be limited to local sources, meaning police agents must determine an offender’s risk level with minimal or incomplete information (AUTHOUT CITATION). This lack of data has a significant negative impact on risk designations, as the accuracy of risk assessment is controlled by data input (Cohen et al., 2020; Miller & Maloney, 2013; Viglione et al., 2015). Importantly, however, both an abundance and lack of information can result in a high-risk designation, as the former equates to long criminal histories and the latter to uncertainty in predictive capacities (Cohen et al., 2020). Uncertainty, unlike risk, is not calculable through assessments. Instead, it is mobilized when minimal information is available and suspicion exists; the goal here is to identify vulnerabilities and to “think the unthinkable” (Klima et al., 2011, p. 24). Reflecting the precautionary logic, uncertainty increasingly displaces risk regimes, and expands both the potential for criminalization and social control by governing the anticipation of criminal activity (Ericson, 2007).

The lack of distinct boundaries of risk often results in an oversimplified definition and conflation with similar concepts, such as fear and dangerousness, the former of which is implicitly linked to risk by definition (See Hollway & Jefferson, 1997; See Woods, 2019). Dangerousness, unlike risk, not only evaluates the capacity to commit future crime or harm, but also the seriousness of the crime, such as the vulnerability of the victim(s). These danger narratives are particularly impactful because those designated as dangerous offenders (i.e., sex offenders) are perceived as unlikely to be rehabilitated in addition to being likely to commit a more severe act in the future. Empirically examining how fear, risk, and uncertainty play out in offender management decisions is increasingly important given that these measures are prone to overemphasize risk (Farrall et al., 2009; Hannah-Moffat, 2004). Thus, scholars continue to document the perpetual development of risk frameworks that prioritize managing the fear of potential future criminality, and the use of risk assessment tools and risk reduction strategies and practices to quell uncertain fears (Dubber, 2001; Ericson & Haggerty, 1997; Neocleous, 2006).

Further, the conversations on fear, danger, uncertainty, and risk are prominent in the literature on police powers and surveillance studies (Dubber, 2001; Ericson, 2007; Ericson & Haggerty, 1997; Haggerty & Ericson, 2000; Humphrey & Gibbs Van Brunschot, 2014, 2015; Jochelson & Kramar, 2011; Neocleous, 2006; Woods 2019; Zedner, 2009). In the context of traffic stops, for example, Woods (2019) finds that the dominant danger narrative that all traffic stops warrant risk is “vastly oversimplified,” displaces contextual information, and greatly shapes police officers’ decisions when stopping citizens (p. 642). Woods’ unique and comprehensive analysis highlights the effects of fear and danger perceptions on both policing practices and those targeted. Inflated perceptions of fear and danger result in an overestimation of danger and hyper responses to policing situations, which brings attention to the importance of distinguishing and evaluating levels of risk to improve situational responses (Woods, 2019). Although there are limited studies on the 810 processes specifically (for exceptions see Doerksen, 2014, 2019), studies on the impact of police experiences and perceptions on decision-making provide valuable insight to contextualize our findings.

Despite best efforts to implement calculative, actuarial, and objective risk frameworks, police officers perceptions of risk, danger and fear are shaped by not only legal factors, such as an offender’s criminal history, but also extra-legal factors. These include things like their experiential knowledge and feelings (Ballucci, 2008; Hannah-Moffat et al., 2009; Haqanee et al., 2015; Kras et al., 2019; Pratt, 1999; Wortley, 2003), which are often unrelated to the individual offender (Kerbs et al., 2009), as well as demographics, demeanor, and victim-suspect relationships (see, e.g., Bittner, 1970; Black, 1980; Brown, 1988; Reiss, 1971). For decades, operationalizing officer’s experience (measured in years of experience) has been associated with years of service, but how officers attain experience is multifaceted and complex (Brown, 1988; Klahm & Tillyer, 2015; Muir, 1977; Rubinstein, 1973), and the impact of these experiences is necessary to understand structure decision-making. For example, police officers hold stereotypes and preconceived notions about risky characteristics as a result of continuous exposure to particular groups, ultimately influencing how they make decisions through observation and interpretation of actions and behavioral schematics (see, e.g., Klinger, 1997; Muir, 1977; Rubinstein, 1973; Smith & Alpert, 2007). The impact of experience is evident, for example, in the context of probation and parole decision-making, where professional characteristics and personal beliefs have been found to influence risk and need assessments more so even than the formal assessment tools themselves (Schaefer & Williamson, 2018).

These studies demonstrate that, in practice, determining risk is a fluid process, and uses a combination of factors to constitute definitions of high-risk that are continually debated and contentious (Maurutto & Hannah-Moffat, 2006; Ballucci, 2008). In the following section, we discuss our methodology and illustrate how officers identify and construct risk and uncertainty to support 810 applications.

Methodology and Data Analysis

For this project, we conducted semi-structured interviews and focus groups in a cross-sectional Canadian study across five different intensive supervision units targeting and monitoring small to medium sized groups of high-risk offenders. This process is done through mandatory contact with probation or police officers, as per their parole or through peace bonds such as the 810. Further, the offenders are managed using suitable, available resources as they are released into the community to decrease the risks of recidivism.

The respondents for this study were police agents involved in police services’ operations in one capacity or another (i.e., as an officer, senior administrator, Crown prosecutor, or risk assessor). Interviews were conducted at four sites, ranging from 50 to 140 minutes in length, while focus groups consisted of between two and four members and spanned two to four and a half hours long. Additionally, follow-up interviews were carried out at three out of five sites. The units in our study were located in two different provinces, across four different municipal police services, in urban settings with populations ranging from 200,000 to over 2 million. The interview questions focused on learning about high-risk offender management programs within each agency. Specifically, we asked how they determine which offenders are high-risk, as well as what factors and practices are used to determine who requires an 810. Our goal was to learn how high-risk offender designations were completed, and about the strategies and practices that were used to manage offenders while in the community. For background purposes, interviewees were also asked questions concerning their training and police history.

One member of the research team transcribed all interviews and focus groups; QSR Nvivo was utilized to aid in compilation, organization, and coding. After the data was reviewed and discussed, a thematic analysis was conducted. About 34 primary themes and 20 subthemes were identified, resulting in 54 codes within the data. Sources were selected based on a rough understanding of the limited available literature we intended to engage in conversation with, area of law, and trends in the data. Common themes, gained through their high reference count throughout the data, included “810 Orders.” Additionally, subthemes were created for variables to help in summarizing the data to pinpoint specific sections and included “Obtaining an 810,” “Active 810s,” and “Benefits of an 810.” Additionally, “Assessing Risk,” “Typical Workday Duties,” “Officer Evaluators of Risk,” “Risk,” “Managing Offenders,” and “Monitoring/Surveillance” were also common themes used in this paper’s analysis, through which subthemes narrowed down our focus. Moreover, a preliminary analysis of the data, in combination with understanding garnered through conducting and transcribing the interviews, structured the coding process. A first pass of coding looked at descriptive elements, whereas theoretical elements emerged on a second pass, offering insights into broader areas, such as risk and the respondents’ experiences of the benefits of managing offenders through peace bonds.

In the following section, we offer unique insights into how the concepts of fear, uncertainty, and risk are germane to police officers’ perceptions and responses through an analysis of documented practices and strategies used in the 810 determination process. What arises is two distinct types of cases that provide a framework for analyzing decision-making processes. Additionally, three typologies show that the impact of uncertainty and risk are the same; each result in suspicion and surveillance, albeit for different reasons, and suggest that high-risk designations are complex, moving beyond known factors to include uncertainty and the probability of success.

The Practices of Pursuing an 810 Peace Bond

Initiating and Deciding When to Pursue an 810

The decision to initiate an 810 customarily begins with the police service after receiving information from correctional authorities (see MacAulay, 2001). The process requires that both agencies do an in-depth review of offender files when considering an 810 to ascertain whether there are any risk indicators or evidence suggesting uncertainty in predicting future behavior. Some key indicators and rationales that lead correctional officers to flag and review a file for police follow up include situations where the potential victims are children, when conditions can help with re-integration into society, and when offenders that have reached their warrant expiry date ¹ are being released. The decision to move forward with a peace bond often involves consultation and the sharing of information between correctional practitioners, police officers, and the Crown to determine whether the evidence will meet the legal threshold of reasonable grounds for fear. As one respondent explains, at the core of the process is the question, “Does this person pose a risk to the community?” One police agent describes the initiation process as follows:

Corrections is the first [step] [. . .] The Parole Board of Canada has [. . .] review [. . .] each person as they come up for day parole, or full parole, and stat release [. . .] They’re the ones that have to look at everything and then [. . . forward us the documents to indicate] that they believe that this person poses a risk, and then it’s up to the jurisdiction that the person is coming to [. . .] to review all the material and decide, “Do I believe that this person poses a risk to the community?” Not to the police, but to the community, and that’s where you start.

Moreover, correctional agents make a recommendation on whether an 810 should be initiated using the information made available through Corrections Service Canada (CSC) parole review process. This is foundational to the process, as the collection of risk assessments and reports assists in evaluating risk and helps determine whether there are reasonable grounds to deem an offender a risk to the community. Although risk assessments provide a strong foundation in helping police agents determine potential peace bond candidates early in the process, they are only one piece of the puzzle. Documents that detail risk classifications act as important signals, but interpretation and meaning are needed not only for these assessments, but also for all the available information when deciding which cases to take forward for an 810.

Ultimately, in most cases, it is officers who decide whether or not a peace bond is worth pursuing, regardless of the fact that many have minimal risk assessment training. ² In other words, while they have risk tools at their disposal, and occasionally attend brief week-long training courses or conferences, these officers are not trained extensively in risk assessment and, in turn, must rely on their experience and background. Importantly, across all the sites studied, we found little to no training being given to officers who work on peace bonds. What we found was that respondents noted low levels of training prior to being placed in both their current and previous positions. When probed further about the training they underwent for their position, one respondent states that there was none. They explain:

[You just need] experience, [. . .] there are no courses on how to manage offenders. There are courses [on dealing with risk, like the one] I took it at [a local university] [. . . and other] courses on how to draft an 810. [But,] that’s something you do on your own. You have to be able to do the paperwork. But my experience comes from what I did before.

Another respondent echoes these sentiments: “There is no long answer. I was assigned, [and] the person that was in there sat down with me for half a day and said this is what you do. That’s it.” While training varies across the sites studied, it is kept to a minimum—ranging from none to roughly 1 week, used only to familiarize the agent with their duties. Respondents also mentioned that additional training is available, but that they are not always able to attend. This lack of training is concerning yet not surprising, particularly given the power of police officers in these decisions. This becomes increasingly problematic when you take into consideration that when risk assessment tools lead to inconclusive results, officers make decisions based on their experience. Although experience-related knowledge has value, comprehensive training provides structure, consistency, and clarity in decisions about which offenders require conditions and monitoring via peace bonds.

Nevertheless, despite the lack of guidelines and training, officers are able to articulate common factors and considerations. In the following statement, the agent tries to unpack predictions about reoffending. They highlight how the presence of some risk factors as opposed to others creates a level of uncertainty:

It’s hard to explain [. . .] You [just] know. Like, this person doesn’t have anything good in his life. You know he [has a history of violence and is likely] to [relapse]. But if he has a little bit of good, and a little bit of bad, or [maybe he only becomes violent] when he drinks and then other times he’s fine. [So things like] that can change. If he gets involved in alcohol, that’s the start of an offence cycle, so that could reduce or increase the risk. But if he’s not drinking [. . .] and that’s not what causes it, [then maybe] there’s mental health [struggles], or he’s not compliant with medication or whatever. You [have] to look at everything [. . .] as a whole. [. . .] You kind of have to look at “Okay, these are the risk factors” [so you can then say,] “Okay, well based on what I’m seeing, and his behaviours, this is what he could do.” If we address these issues, that might reduce the risk. It’s hard to explain how you come across these [signals] [. . .] I can’t put it into words [. . . You] just know by reading it [in reports]. Plus, you’re doing them so often. You can say, “Well, this guy’s not a risk” and you just know, right? [. . .] It’s more your experiences [. . .] you’re doing them all the time. [sic]

Although this agent mentions certain factors that risk assessments rely on, such as criminal history, dependence on alcohol, and mental health, they are unable to give clear or consistent rules for predicting an individual’s future actions within the community. In light of this level of uncertainty, agents rely on what they know of the offender, or their inclinations, to achieve a greater certainty of re-offending. The lack of guidelines and training concerning what constitutes high-risk empowers and holds police agents accountable; and, as a result, we show that uncertainty becomes a means and justification for suspicion and, therefore, substantiates governance.

The following respondent outlines the importance and power of “knowing” the offender by highlighting the types of indicators that suggest they are at risk of returning to criminality ³ :

[W]e know [a guy is] using because his hygiene [lessens]. So, we meet with them on a bi-weekly basis, go *sniff*, “Oh, you’re smelling a little funky,” we know he’s using again. [We also have a] strategy where we’ll start to follow him and [see if he’s] engaging prostitutes [. . . and seemingly] going into this realm again, [so] we’ll latch on to him and deal with that sort of stuff. So again, it’s understanding your people, having that ownership of them, for lack of a better word, when you can see these types of changes and be very preventative. . .

In this example, the agent illustrates the impact of extra-legal factors. When an officer feels they know an offender and their actions, their knowledge becomes a proxy to measure risk and work around or deal with uncertainty. Their individual knowledge and discretion impacts how they interpret and understand offenders’ actions, character, and behavioral patterns. This process of judgment acts as a means of providing additional support to the meet the legal requirements of an 810 application. For instance, when a police officer has knowledge of an offender, they can identify risk indicators and take necessary investigative steps. This is common practice and perceived as an effective method, as the it provides officers the power to interfere in and monitor offenders’ day-to-day lives.

Another agent expands on this idea and explains how external knowledge of the offender can be useful in dealing with uncertainty:

It would rely on what caused our fear. So, the hearings allow a little bit of hearsay, it doesn’t have to be everything proven. [Theoretically,] I can testify that, say on this day, he told me [. . .] that he saw a girl on the street that he was attracted to and he didn’t feel like he’d be able to stay away from her. [That could] be part of my grounds. [Say there’s] another occurrence three years [prior where] he said the same thing to another officer, [. . .] and right now [he’s being] investigated for watching a girl [where] the police were called because someone saw him [. . .] following her, [or] something like that [. . .]a judge [then] decides “Well, y’know, I can understand that’s concerning, but it’s not enough to get the order.” Luckily, we don’t have to do that too often. Usually it’s just these main cases, coming out of prisons.

This police agent demonstrates that, although hearsay and other non-verifiable factors are uncommon, they can still be used to secure a peace bond. Thus, this information is assessed and considered even though it does not provide the same assurance as formal risk assessments in demonstrating reasonable fear. Yet, at the same end, personal experience and knowledge are also often disqualified. To gain a clear picture of how such determinations are made, we examine police agents’ accounts of how and when they decide to pursue an 810. We look at how reasonable fear is established through high-risk indicators and high levels of uncertainty to justify state intervention.

Types of Cases Identified

Next, we outline the two different types of cases decision-making strategies identified in our data. To summarize, slam-dunk cases, which are built on risk indicators, involve high risk of criminality and low uncertainty. In these cases, certainty about an individual’s future criminality is substantiated by multiple risk factors present. Alternatively, sit-down cases lack identified risk and involves some uncertainty, which leads police agents to rely on the extra-legal factors described above. Unclear cases require additional information and deliberation beyond the limited professional evaluations put forward by CSC, such as police reports, transcripts, and agents’ personal experience and knowledge when determining what makes some dangerous. This category also includes cases where individuals have no prior criminal record, which also results in high levels of uncertainty; since little is known about these offenders, it is difficult to assess the possibility of re-offending. In “non-slam dunk” cases, speculation and conjecture is used to justify placing conditions on the individual. But, given that fear is the necessary legal component required to convince a judge to approve a peace bond, the unknown is enough to establish risk.

1. Slam-dunk cases

“That’s a no-brainer”: risk high and uncertainty low.

In slam-dunk cases, the offender profile presents little to no ambiguity for correctional agents. Key features here are multiple, widely accepted risk indicators from various expert and formalized sources, which allow police and correctional agents to establish with ease that an offender poses a high risk of recidivism. When asked to explain what makes a clear-cut case for obtaining a peace bond, one officer’s response encapsulates the typical slam-dunk case:

. . .A slam-dunk [case] is a guy that spent three years in prison for [sexual] offences against his children, [who, after getting out . . .] then does another three years for choking a woman and sexually assaulting her, and then gets held in during that sentence and is deemed a high risk to reoffend sexually by a psychologist, and doesn’t get out right until the end of their sentence. So, that one we don’t have a lot of choice. We’re gonna apply for that [. . . that’s] a no-brainer. [Another example is] a guy who [did] 15 years for aggravated sex assault and [. . .] didn’t learn anything while they’re inside—they still deny it or they behave poorly and breach a lot of their conditions. So, that’s a no-brainer, and a slam-dunk, and we have to apply. And I say probably half at least of ours are that simple for us to apply for.

For this agent, slam-dunk cases include released offenders who will quickly reoffend, have been deemed high-risk by experts, have a long criminal history, have victimized vulnerable people (such as children), and have demonstrated no effort to change while incarcerated. In these cases, pursuing a peace bond is an obvious next step for police agents. The factors considered are static and well-established risk indicators (see Bonta et al., 1998; Hannah-Moffat, 1999) and, therefore, their presence creates a strong case for an 810. ⁴ Another agent identifies similar indicators that impact this decision:

We look at if they participated in programming in prison. What was their behavior in prison? Were they involved in violence there? Did they continue their behavior? When they’re in custody, they also have to see a psychiatrist and a psychologist, so we’ll rely heavily on their reports. So they’re saying this guy is a psychopath, he’s still [going to] reoffend. Those are the types of people we would take.

In this account, expert reports on offenders, as well as their previous record of offences, may substantiate the potential for future criminality. Additionally, these examples show how an offender’s unwillingness to take responsibility for their own rehabilitation through programming is considered a strong indicator that they will offend again. Another respondent explains:

. . .[W]hen they’re incarcerated, they’re given assessments every so often. So, depending on what those assessments say, they may be deemed high-risk upon release. And these are the offenders that go right to the end of their sentence, [who] may have had a statutory release at some point [. . . before being] re-incarcerated for whatever reason [. . .] So, while they’re in prison, they’re encouraged to do programming, and some of them do, and through [this] an assessment is done — [sometimes] the program was unsuccessful, or because of their mental health issues the program was deemed unsuccessful [. . .] They may have went [sic] through the program, but [. . . sometimes it doesn’t] help them. Or, they do no programming whatsoever, so [now] they’re an untreated sex offender, or an untreated violent offender and they’ve said, “nope, we’re not doing anything.” So, now they’re deemed high-risk because they come back out into the community and we think because of what the assessments [say] [. . . and ] all the information that they’ve gathered from their previous offences — [. . .] maybe they’re prolific offenders, and most of them are, that’s usually not their one-time offence, they’ve usually done other things that [. . . mark] them [. . .] high-risk — so, from all of that, there might be interviews with family, [because] who else would be interviewed?

This respondent’s comments echo the importance of active participation and success in programs, as well as attitude. Offenders’ efforts to actively participate are evidence that they are taking responsibility for their own behavior, which, as shown elsewhere (See Ballucci, 2008; Hannah-Moffat, 2005), often reduces offenders’ risk scores. The converse of which, or a lack of such action, increases these scores.

Another significant factor in slam-dunk cases is whether the individual has served their sentence to their warrant expiry date. Whereas a vast majority of offenders are released prior to this due to a variety of reasons, such as prison overcrowding, good behavior, and rehabilitation, a small portion are incarcerated for their full sentence if there is a significant fear they will reoffend shortly after being released. When an offender is released at their warrant expiry date, barring an exceptionally rare long-term offender classification, there is no ability to place conditions of any form on the offender unless police pursue a peace bond. In discussing their unit’s case selection practices, one agent highlights that violent offenders who have fulfilled their sentences are often strong candidates for an 810. They explain:

We have some that are very violent, but they’re still in the provincial system, [rather than . . .] the federal system. When their warrant is expiring and they’re going to be released — that means they’re not on parole, they’ve done their complete time, but while they were in custody they would not admit that they were guilty. Even though they were convicted. They would not take any treatment. They’re very violent. It doesn’t have to just be sex, it could be weapons, it could be murder, it could be aggravated assaults. Whatever it is. [Then we] get the warrant expiry package from the jails. “This guy’s getting out. This guy’s a high risk.” [We] go through all the paperwork. If [we] believe, given the history and all the information that they gather on this person, that this guy or girl — mostly guys — have the propensity [to break the rules . . .] then what [we] do is reply through the Ministry of the Attorney General’s office to have these people put on what’s called an 810 order.

As the above respondent explains, peace bonds address a gap between an individual’s freedom not to be penalized by the state for an offence they have not committed, and, on the other hand, the interests of security and the regulation of uncertainty. This is a theme that becomes more prevalent in the following analysis. Another respondent from a different site confirms, “Of all the 810 guys, at least 90% of them are people that have just finished their jail sentences.” An offender is a prime peace bond candidate when, prior to their release, they still have not been rehabilitated, have shown poor behavior while incarcerated, and are still classified as high risk. Risk is also more easily evaluated and established when offenders have spent substantial time in correctional facilities and become the subject of many assessments and reports, or in other words, a “script for action” (Ericson & Haggerty, 1997), which can be used to justify an 810.

Central to a slam-dunk case is the presence of high-risk indicators that suggest a high degree of certainty that the individual still poses a risk to the community. This status is verified by extensive documentation of their past offences, the behavior they exhibited while in prison, as well as whether or not they have accumulated serious charges. These cases are also characterized as slam-dunk because there is greater potential that the judge will approve the application. In the absence of strong risk evidence, our analysis shows the impact and power of uncertainty as a rationale for pursuing an 810.

2. Sit-down cases

“You’d be able to pick it up too”: medium uncertainty and medium risk.

Whereas slam-dunk cases are often straightforward, sit-down cases require more scrutiny. Here, the evidence may suggest a strong possibility of future criminality, but that evidence may not be conclusive in a court of law. In other words, uncertainty is present, but additional evidence or vetting may be required in order to prove those fears in a court of law. Thus, a strong case must be carefully built, lest the application fail due to a technicality or a judge deciding that the evidence of risk is lacking. Our analysis shows that agents will often consult with others, interpret, and give meaning to information when deciding whether the offender’s actions indicate a reasonable risk to re-offend, in addition to using their experiential knowledge, to justify reasonable grounds for fear.

For this reason, intuition plays an early and important role in the decision to begin building a sit-down case. One officer explains:

So, the ones where we still fear they’re going to reoffend, but they’re not deemed quite as high-risk [because] they don’t have the same history of similar offences, [. . .] we actually sit down as a group and kind of talk about the offender and decide. The hard part is, we don’t know for sure which one is [sic] going to be the one that reoffends. There is a possibility that we won’t apply for one, and then a year later they reoffend. Or, we do apply, and they don’t re-offend ever again. It’s kind of a group decision at that point. And sometimes, if we’re really not sure, we’ll go to the Crown and involve them and ask them, “This is what this guy has done, do you think we can get an order for him?” and at that point, we decide whether we do or don’t. But a lot of that is based on experience, and if you read the files you’d be able to pick it up too. Just based on whatever training you have—“yeah, that guy is dangerous.”

The officer’s assertion that “you’d be able to pick it up too” suggests that an offender who poses a risk is easily recognizable to a civilian. Through this, risk indicators appear objective and universal. However, their comments also attribute their ability to identify risk as a product of their experience. Officers are “sitting down” as a group and consulting the Crown to discuss offenders, demonstrating that they use a combination of intuition, experience, and collaborative thinking in their decision-making processes. For this officer, both experience and intuition provide the justification for a designation of high-risk. The lack of strong evidence of risk factors however, does not preclude the officer from investigating and building a case for an 810. Instead, it creates uncertainty around the offenders risk level, which, in and of itself, is a risk factor. The minimal amount of information and knowledge of the offender serves as a reason for further investigate, rather than indicating that they are not a risk. Agents at alternative sites further this idea, they explain:

Within CSC, while they’re in there, they do different reports, programming reports, updates to their criminal profile. [. . .] So, they’ll send those documents through to us. We read through that stuff, the detectives themselves, and our staff sergeant kind of decide[s] if they meet the threshold for [us] to go through and write an assessment on them to [then] determine if they are fit for an s. 810 order. If we do decide that we’re going ahead with a full assessment to get an s. 810 order on them, we request transcripts, police reports, as much information as we can gather on them because the CSC documents are limited to whatever they were told within.

In this example, officers screen files to decide if the offender “meets the threshold.” Importantly, these files are often cases where uncertainty is present. In sit-down cases, where this issue abounds, police agents not only collect all available information, but also consider various extra-legal factors to help verify suspicions about an offender’s risk level. As we show, these can include subjective assessment, such as experiential knowledge, knowledge or intuition about the offender, resources, and likelihood of approval by the judge. This process is ultimately an exercise in achieving a similar level of certainty to that seen in slam-dunk cases.

As peace bonds are preventative in nature, they are intended to place probationary conditions on any individual regardless of whether or not they have committed a crime if reasonable fear of criminality exists. ⁵ In these cases, little information is available and alternative strategies to establish that a risk exists must be used. Instead of focusing on demonstrating the existence of multiple risk factors, they aim to establish a reasonable fear of risk by speculating on the potential meaning of what little is known about the prospective offender. Thus, uncertainty not only comes from limited information, but also from the impact of the crimes that might be committed. In cases where the victim is vulnerable, such as a child, or unexpected, such as a terrorism attack, relying on uncertainty is enough to justify action. One respondent explains:

We kind of need to justify why we feel they’re going to commit an offence. They don’t necessarily have to have a record to get the order, but it helps. We can have a case for somebody [who] hasn’t committed a criminal offence. [. . .] For an 810.1, it would be somebody who chronically has been hanging around children’s playground[s], or community centers, and we’ve had reports of them approaching children to speak to them and, say, giving out candy. It’s fairly rare, but we have had a couple where they just admit they have a problem or “I’m attracted to children and I need help.” Sometimes we’ll put the order on them almost on a voluntary basis to put conditions to say stay away from parks, stay away from somebody, don’t get into a relationship with somebody who’s got children. Some of them find it helpful, but it has to be a circumstance where we can justify it before a judge where we fear that they’re going to do something, otherwise we don’t get it.

In this example, an 810 is justified by speculating on the potential for criminality based on a single report. Although this is a form of evidence, it is questionable as to whether it alone offers enough to suggest a reasonable fear of crime. However, the fact that the risk of harm is to children plays a significant role in demonstrated reasonable fear, meaning there are less barriers necessitating the requirement to corroborate the potential of risk. Instead, the level of uncertainty, in combination with the police officers evaluation of the report, is enough to manifest a potential fear and pursue an 810. Further, factors such as an individual’s lifestyle choices, a report of potentially suspicious behavior, or confessions about sexual preferences are enough to establish fear. Thus, police officers use a combination of vulnerabilities of potential victims as well as speculation and conjuncture to make a strong case for an 810. According to the Criminal Code, only the potential for future criminality must be demonstrated to pursue an 810, and therefore, those without a criminal record are not excluded. However, the practice of governing on the basis of uncertainty, as seen in sit-down cases, raises questions concerning the extent to which grounds for reasonable fear can be expanded. If both the evidence and lack of risk factors can constitute a risk—who is not a potential risk?

Beyond Risk: Success and Resources

As our examples show, risk and uncertainty play a key role in these processes. However, there are additional factors that must be considered. One officer, while describing their process of evaluating offenders’, states that surveilling offenders is “[. . .] very effective, but at what cost? [. . .] [Y]ou’re using 6 to 8 dedicated police resources on 18, 25 people.” Moreover, while officers see the process as providing them the power to regulate and improve offenders’ behaviors, they seem hesitant to identify too many candidates for 810 because of the impact on police resources. This demonstrates that high risk designations go beyond the individual evaluation and extend to the ability of officers to manage substantial numbers with limited resources and uncertainty of success.

Consequently, applications are typically pursued when there exists realistic expectation of success. As a result, peace bond applications have high rates of success. For examples, one respondent boasts that their unit has a 100% success rate, explaining:

We’ll usually do our own risk assessment up front in our unit here, and then if we feel the risk is high, we will make an 810 application with [the Crown]. We’ll meet and discuss it. There’s been ones that are kind of teetering on the fence that we really think we [should] pursue one and the Crown says “Well, I don’t think you have enough” and we take their advice.

This agent’s comments show that the decision to pursue an 810 is not solely based on risk assessment outcomes, but also the likelihood of support from the Crown. This is another factor officers will consider beyond questions of risk and uncertainty. Although the Crown offers advice on the strength of a case, the judge ultimately decides if an 810 will be pursued. Thus, although police agents have a sense for what evidence will make a strong case, they also indicate that there is variability in judge assessments. One agent explains:

[T]he judge can decide well, I can’t rely on that assessment for whatever reason. If they feel that they don’t have a degree in psychology, then how can they make an assessment? It depends on each case, but a judge could technically say “well, I don’t feel that you’re an expert in that field so I can’t rely on your assessment [that they’re] a high-risk.

This comment illustrates how judges’ interpretation of the assessments plays into determining the overall strength of the case. For this reason, the decision to move forward with an 810 goes beyond risk to include success rates, as police officers avoid investing time in preparing applications if they feel the success rate is low. The decision to initiate an 810, therefore, is not solely about determining risk or uncertainty, but also about how judges decide what evidence is credible and trustworthy. This means that risk determinations involve consideration of both the legitimacy of the claim and whether it will satisfy the system.

Further, a judge’s decision is also shaped by their obligation to respect the legal rights of offenders. Although peace bond orders are often successful, they are not granted lightly. As one agent describes, judges must balance the protection of society with the rights of the offender:

It’s preventative in nature, but courts are very reluctant to use it for obvious reasons. You’re putting someone on conditions to report to probation and police and subject [them] to a lot of conditions for something they haven’t done. It’s what we think you may do. These are people who have already completed their entire sentence, and have done everything they’re supposed to do, but we’re saying that’s not good enough. So the courts, for obvious reasons, are very careful about people they put on [peace bonds].

Demonstrated here is an awareness of the impact and potential legal infringement on offenders’ rights, which highlights the importance of protecting both the community and individual rights and freedoms. Although there is an understanding of the necessity of risk metrics when attempting to convince a judge of the existence of reasonable fear of harm to the community, the requirements for high risk are not uniform or generalizable; different jurisdictions offer different opinions about peace bonds. Yet, officers and the Crown can challenge judges’ decisions to acquire a win, demonstrating the malleability of risk.

Discussion and Conclusion

Our study examines how police officers make decisions concerning risk. We show how high risk is constructed and the role of extra legal factors, such as experience and resources in this process. Officers mobilize archetypes to help determine and ensure that certain offenders will be placed under a peace bond. Although the impact of uncertainty discourses is evident in each of our case types, it is most obvious when individuals with little to no criminal history or charges are placed under parole-level sanctions. Here, very little is known about the individual and, therefore, uncertainty of future behavior is high. This prompts police officers to speculate and investigate to see if they can substantiate reasonable fear that an offender will commit an offence. As we show, however, in cases where multiple risk indictors are absent, the presence of uncertainty can potentially be enough to justify reasonable fear, particularly when the potential for criminality is determined to involve vulnerable victims. We also show that if a realistic expectation exists that the application will be successful, police will be more obligated to pursue a peace bond. Importantly, however, the likelihood of a successful application is very different from the level of risk the offender truly poses, or whether or not a peace bond is the best option for management. These non-uniform and informal practices of knowledge collection illustrate the power of uncertainty to support 810 applications.

Unlike studies that suggest that extra legal factors are diminishing (Engel et al., 2002, p. 252), we find that, despite the growth of risk assessment and analysis in policing, decision-making on risk is predicated on these factors. Thus, we build on the work of scholars and add to the literature that teases out extra legal factors that shape police decision-making (See Engles et al., 2002; Meehan & Ponder, 2002; Schaefer et al., 2006) by showing how legal requirements, resources limitations, and uncertainty discourses impact high-risk designations.

Our empirical analysis on how specialized 810.1 and 810.2 peace bonds are deliberated and decided, demonstrates the impact of precautionary logics on governance, and the extent to which this process is not only discretionary, but also expands policing powers. The decision to apply for an 810 is ungoverned, and officers have little to no accountability for their decisions, beyond the application being rejected. Given the extent to which extra legal factors play into this process, the need for an accountability measure and firm guidelines on who to impose an 810 upon is necessary. As it stands, offenders’ futures are left to the discretion of the officers’ interpretations of risk.

The implications of our study go beyond the specific literature on peace bonds to include concerns about police officers’ discretionary power and their level of accountability in crime prevention practices. The practice of applying for a specialized peace bond is highly impacted by perceptions, and are not only absent of distinctive guidelines and boundaries, but also accountability measures. Police officers’ decision-making capabilities that affords them a significant level of power, is extended in the precautionary framework, without consequence. Thus, decision-making guidelines for peace bonds specifically, and precautionary policing generally, require assessment and instituted guidance to better manage the potential of unfounded fear, which can, as seen in the context of traffic stops, result in policies and perceptions that further fuel violence through the over extension of risk (See Woods, 2019).

Our study raises concerns about the process of determining an 810 and calls for the implementation of further guidelines. The lack thereof, alongside police officer discretionary power, creates scenarios where various types of informal evaluation permit state intervention. In cases where the risk of future criminality is clearly defined, uncertainty diminishes, but risk remains high. On the other hand, cases that lack formal risk assessment or evidence have high uncertainty, which also results in high-risk designations. Each of these types of cases demonstrates how state power can be extended when various levels of risk and uncertainty are present or constituted. Yet, if all levels of uncertainty are equally powerful in justifying state intervention, the liberties of offenders, both proven and potential, are not the only ones at risk of the overreach and potential abuse of state power. For these reasons, like Stockdale, we call into question the “coherence of pre-emptive security” (Stockdale, 2013, p. 154), and call for further studies that document the impact of pre-emptive logics and strategies on not only those who are targeted, but also those who are not. If pre-emptive strategies continue to expand, this logic is sure to impact the perception of preventive governance as not only acceptable, but necessary.