Abstract

Purpose: Investigate the effect of body-worn cameras (BWCs) on jail resident injuries during response-to-resistance (RTR) events. Methods: One-year randomized controlled trial among 12 housing units in the Loudoun County Adult Detention Center in Virginia. Negative binomial regression investigated the effects of BWCs on jail resident injuries during RTRs, and logistic regression predicted whether RTRs result in an injury. Results: A 58% reduction in predicted injuries in unit-months where jail deputies were assigned BWCs, corresponding to an average of 0.17 resident injuries in unit-months without cameras and 0.07 injuries in unit-months with cameras. BWCs also reduced the likelihood of injury occurring by a factor of 0.12 (injuries occurred in 28.4% of RTRs without BWCs, versus 8.8% of RTRs with BWCs). Conclusions: Little is known about use-of-force incidents in prisons and jails, nor the rate at which these incidents result in injuries to incarcerated residents. Yet, in response to calls for greater transparency and accountability among criminal justice agencies, correctional administrators have increasingly turned to BWCs to address these issues and enhance the safety and security of their staff and resident populations. This study addresses critical knowledge gaps and offers one of the first tests of a correctional BWC program.

Highlights

• The first randomized controlled trial of body-worn cameras in a jail setting

• Adds to limited literature on correctional uses of force and related injuries

• Found a 58% reduction in the number of resident injuries from body cameras

• Estimates injuries in 28.4% of responses to resistance without body cameras

• Estimates injuries in 8.8% of responses to resistance with body cameras

Keywords

Body-worn cameras, jails, corrections, randomized controlled trial, injuries

Introduction

One of the primary goals of prison and jail administrators is to maintain safe and secure institutions for both staff and incarcerated residents (Cressey, 1959; Denhof, Spinaris, & Morton, 2014). This is achieved through the implementation of various management strategies, including the careful and consistent enforcement of rules (Clemmer, 1940; Cloward, 1960; Freeman, 2003; Haggerty & Bucerius, 2021; Liebling, 2000). Through the course of these management efforts, correctional officers (COs) sometimes use force to ensure compliance with rules and to maintain social order (Crawley, 2004; King & Elliot, 1977; Marquart, 1986; Schultz, 2023; Symkovych, 2019). While a certain amount of force is expected and would be aligned with the policies of a correctional agency, there have been several recent and highly publicized examples of excessive force that resulted in serious bodily harm to incarcerated residents (U.S. Department of Justice, 2022; Finton, 2023; Fuller, 2023; Koval, 2023; Maxwell, 2023; O'Connor, 2023; Rembert et al., 2023; Santo & Neff, 2023). Even necessary and lawful applications of force can result in injuries, with police research showing that injuries to community members may occur in 17 to 64% of officer-initiated use-of-force incidents (Smith et al., 2010). Though there is a long history of abusive staff conduct directed at incarcerated residents in jails and prisons (Butterfield, 2004; Specter, 2006), no comparable estimates of use of force (both unlawful and lawful) and resulting resident injuries exist for correctional agencies.

One promising approach for enhancing institutional safety and mitigating uses of force in correctional settings is body-worn cameras (BWCs). At least ten states and several jail jurisdictions have begun providing, or made plans to provide, BWCs to their COs (Brodie, Kirkland, & Kelly, 2020; Welsh-Huggins, 2021; Winton, 2021). Correctional administrators originally used BWCs with their special operations and emergency-response teams to improve accountability within the agency and provide a record of CO-resident interactions that had the potential to result in serious physical altercations (Brodie et al., 2020; Bui, 2016). More recently, officials have deployed cameras throughout their facilities with the specific goal of reducing staff misconduct and enhancing the overall safety of the institution (Verdon, 2022; Winton, 2021).

The uptake of BWCs in correctional agencies has been spurred by previous policing research. Early studies found significant and sizeable reductions in use-of-force incidents among police officers equipped with BWCs (Ariel, Farrar, & Sutherland, 2015; Jennings, Lynch, & Fridell, 2015). Findings from other studies further indicated that BWCs could reduce the likelihood of injuries to both community members (Henstock, 2015) and officers (Stolzenberg, D'Alessio, & Flexon, 2019). More recent research has produced mixed results on these outcomes in law enforcement settings (Lum et al., 2020; Wilson et al., 2022), but there remains a strong sense of optimism among officials and policymakers that BWCs are well-suited for the correctional environment (Verdon, 2022; Winton, 2021). Despite this potential, there have been no studies to date on the impact of BWCs on use-of-force incidents or sustained injuries. This study seeks to fill this gap through a year-long randomized controlled trial (RCT) testing the impact correctional BWCs have on resident injuries during responses to resistance (RTRs) in the Loudoun County Adult Detention Center (LCADC) in Virginia.

Prior Research

Uses of force and injuries in correctional agencies

The use of force in prisons and jails is a necessary component of ensuring compliance with institutional rules, maintaining social control, and preventing violence and other infractions (Crawley, 2004; King & Elliott, 1977; Liebling & Price, 2003; Marquart, 1986; Symkovych, 2019). In fact, Schultz (2023) argues that use of force is an expected byproduct of modern correctional organizations—a strategy that COs carefully deploy and justify under their agency’s bureaucratic structures. However, a significant gap in extant literature is the lack of empirical research on the prevalence and scope of correctional use-of-force incidents and resulting injuries. A survey by the American Correctional Association found a per-facility average of 70 use-of-force incidents in 1993 across 325 prisons from 49 state departments of corrections and the Federal Bureau of Prisons (Henry, 1994). More recent data of the California prison system noted 7592 use-of-force incidents in 2022 (California OIG, 2023).

There is also no research on how often CO uses of force result in injurious outcomes (see Smith et al., 2010 for a discussion of this in law enforcement settings). Rembert et al. (2023) analyzed 239 federal court cases involving excessive uses of force, many of which resulted in injuries to incarcerated residents. The authors identified four themes that could explain these incidents: (1) perceptions of disrespectful behavior, (2) professional boundary violations, (3) a desire to protect female staff, and (4) angry and retaliatory actions. Hepburn, Griffin, and Petrocelli (2000) examined how to prevent use-of-force related injuries through nonlethal weapons (pepper spray, stun gun). The authors reported that while nonlethal weapons did not change overall use-of-force levels in the Maricopa County, Arizona, jail system, the use of such weapons reduced injuries to residents and COs (Hepburn et al., 2000).

In addition to this scholarly research, there has been broader public interest on these topics, as evidenced by several recent news stories on CO abuse in both jails (Finton, 2023; Fuller, 2023; Koval, 2023; Maxwell, 2023) and prisons (Neff, Santo, & Meagher, 2023; O'Connor, 2023; Santo & Neff, 2023). Reporters from the Marshal Project recently completed an in-depth investigation into such abuses by COs in the New York state prison system and found 290 cases of serious abuse over a 12-year period, which was likely only a fraction of these incidents (Santo & Neff, 2023). It is important to note that these stories were likely selected by editors specifically because of their “newsworthiness” (Chermak, 1995) and because they served to frame a narrative about CO abuse. Still, such stories complement the data collected through more rigorous research, providing context and illuminating issues that have lacked scholarly attention.

Whenever they do occur, uses of force, abuse, and injuries can be detrimental to the institutional and public safety goals of an agency. For example, in his foundational work on social order in prisons, Sykes (1958) contended that COs must maintain order by balancing the deprivations inherent in carceral settings with reciprocity. Misbehavior occurs when the control becomes unbalanced and the deprivations become too severe, such as when staff are excessive or capricious in their enforcement of rules (Sykes, 1958). In more recent research, COs also described their “thuggish coworkers” as “stupid and overemotional” whose actions are a “hazard” to the institution (Schultz, 2023, 670). Research suggests that the exposure to violence within correctional facilities, which could include abuse by staff, can lead to post-traumatic stress symptoms among residents that last years after their release from custody (Novisky & Peralta, 2020). Violence and other negative conditions of confinement can also beget more violence and have deleterious effects on recidivism and other post-release outcomes (Boxer, Middlemass, & Delorenzo, 2009; Hopwood, 2021; Listwan, Sullivan, Agnew, Cullen, & Colvin, 2013).

The potential of BWCs

Body-worn cameras hold potential for reducing injuries stemming from uses of force in prisons and jails. Not only do they provide a direct account of what transpired between COs and incarcerated residents (Neff et al., 2023), they also bring with them a sense of oversight. Most correctional institutions are already saturated with existing stationary cameras that were specifically implemented to deter resident misbehavior and aid CO investigations of misconduct incidents (Allard, Wortley, & Stewart, 2006, 2008; Debus-Sherrill, La Vigne, & Downey, 2014; Lawrence, Peterson, Robin, & Shukla, 2022). While incarcerated residents, family members, or advocacy groups can file public records requests for stationary camera footage on uses of force or other incidents, there are roadblocks to fulfilling these requests due to the security risks and privacy considerations associated with correctional settings (Prison Policy Initiative, n.d.; Chamberlain, 2020). For these reasons, we suggest that COs have embraced stationary cameras as a means of controlling the incarcerated population, not to hold themselves accountable for their behaviors.

On the other hand, BWCs have, from their inception, been viewed primarily as a strategy for improving transparency and accountability in criminal justice agencies (PERF, 2018; Peterson, 2023; President's Task Force on 21st Century Policing, 2015). As a result, most of the research on BWCs has focused on their impact on critical outcomes related to officer and community member behavior, such as use of force, complaints, injuries, and officer-initiated activities (Lum et al., 2020; White, Gaub, Malm, & Padilla, 2023a, 2023b; Wilson et al., 2022). Though the results of these studies have been mixed, they have served to spotlight transparency and accountability deficiencies in law enforcement.

Thus, just as BWCs have become ubiquitous in policing and helped stimulate a national conversation on police reform, there is promise that these devices may similarly draw attention to these issues in prisons and jails and spur the system into responding (Cunningham et al., 2023). Peterson and Lawrence (2021) documented a similar phenomenon in the Milwaukee Police Department, whereby use-of-force incidents dropped immediately after deployment of BWCs, but gradually increased in the following months. It is because of the additional scrutiny BWCs are likely to bring to correctional agencies that policymakers have explicitly focused on the ability of these devices to mitigate excessive uses of force and enhance the safety of staff and incarcerated people (Verdon, 2022; Winton, 2021).

Prior BWC research

Despite their potential and their expanding use in the field, there have been no studies to date on the impact of BWCs on mitigating use-of-force incidents and subsequent injuries in correctional settings. Currently, the only academic research on BWCs in prison or jail settings has focused on the perceptions of and support for these devices among COs (Beales & Marsh, 2016; Dodd, Antrobus, & Sydes, 2020, 2023; Peterson et al., 2023; Sydes, Dodd, & Antrobus, 2022). Yet we can ascertain the potential impact of cameras on critical correctional outcomes by examining research from the field of law enforcement.

Dozens of studies, for example, have examined the impact of BWCs on use-of-force incidents among police officers (White et al., 2023a). Some earlier studies examining these outcomes produced promising results. BWCs resulted in a nearly 50% reduction in use-of-force incidents in Rialto, California (Ariel et al., 2015) and a 53% reduction in RTR events in Orlando, Florida (Jennings et al., 2015). Results from more recent studies are mixed, with some still finding significant reductions in uses of force (Braga, Sousa, Coldren, & Rodriguez, 2018; Groff, Haberman, & Wood, 2020) and others producing null effects (Gaub, Todak, & White, 2021; Peterson, Yu, La Vigne, & Lawrence, 2018; Yokum, Ravishankar, & Coppock, 2019). In a directory of 30 studies examining the impact of BWCs on use of force, only 14 found significant reductions in this outcome, while another 14 found no impact and two found significant increases in use of force (White et al., 2023a; see also Lum et al., 2020 and Wilson et al., 2022).

There have been fewer studies examining the degree to which BWCs affect injurious outcomes during use-of-force incidents. What studies do exist primarily focus on injuries to police officers (also measured as assaults against officers). While BWCs seem to have reduced assaults against officers and injuries in some agencies (e.g., Ariel et al., 2018; Stolzenberg et al., 2019), most studies show no impact of BWCs or an increase in these outcomes (Ariel et al., 2016a, 2018; Lum et al., 2020; Wilson et al., 2022). There is limited research on how BWCs may affect injuries to community members during police use-of-force incidents. One study of the Birmingham South Local Policing Unit, UK, found a 188% increase in injuries to community members during police use-of-force incidents without BWCs (i.e., 16 injuries from BWC-wearing officers compared to 46 injuries from their control counterparts over the six-month study period; Henstock, 2015).

While studies on police outcomes have produced mixed results, there is a critical need to extend this line of research into corrections. Predating the rapid expansion of BWCs among police officers, there had already been a concerted effort by police executives and lawmakers to curb their use of force. For example, Peterson et al. (2018) argue that part of the reason BWCs did not affect use of force in the Milwaukee Police Department was the department’s earlier implementation of policies that drastically reduced use-of-force incidents. While there have been some efforts on this front in corrections, such as crisis intervention training (Tucker, Van Hasselt, Mendez, Palmer, & Browning, 2012), these do not appear to be as widespread, comprehensive, or narrowly focused on use of force. It could thus be argued that BWCs are even better suited today for prisons and jails than for police departments.

Moreover, there are fundamental differences between correctional versus law enforcement settings that makes the implementation of BWCs in prisons and jails unique. For example, compared to police-community encounters, COs interact with incarcerated residents on a more consistent and long-term basis. By design, prisons and jails also include a high concentration of individuals charged with or convicted of serious and violent crimes. Correctional facilities also have a higher concentration of individuals who may be more vulnerable to injuries during uses of force, such as those under serious psychological distress and those with mental health problems (Maruschak, Bronson, & Alper, 2021). Finally, police-community interactions primarily occur in public settings, where other community members and bystanders can act as capable guardians to police misconduct and even record the interactions with their cellphones for broader release to the public (Walker, 2022). This is not possible in correctional settings, where most interactions between COs and residents occur away from the public eye. This study aims to address many of these research gaps by presenting results from one of the first studies on BWCs in a correctional setting. It will also add to the limited literature on the relationship between cameras and injuries sustained during use-of-force incidents.

Methods

The present study

The present study, which was supported by funding from the National Institute of Justice (Award No. 2018-75-CX-0019), evaluated the impact of correctional deputy BWCs on resident injuries stemming from RTR incidents (i.e., instances when residents resisted deputies in any way). Relying on a RCT of the BWC program at the LCADC, we have two main objectives for our study. First, we aim to estimate the degree to which BWCs impacted overall levels of RTR-related injuries among incarcerated residents in the LCADC. As we argued above, we believe BWCs are particularly well suited to affect these outcomes in correctional settings. Our primary hypothesis, therefore, is that the implementation of BWCs in the LCADC will lead to a significant reduction in injuries to incarcerated residents.

A secondary aim of our study is to estimate the degree to which other characteristics of the RTR event relate to injurious outcomes. There is a substantial lack of empirical data and scholarly research on uses of force in prisons and jails, generally, and injuries resulting from these incidents, specifically. Given this, we do not have a specific hypothesis for this part of our analysis; rather, we will explore whether any RTR characteristics can predict when an event may lead to a resident injury. We believe this has significant implications for corrections practice and research.

Although this study cannot be generalized to jails located in other counties, the LCADC is representative of local corrections facilities across the U.S. The facility is run by the Loudoun County Sheriff’s Office and provides jail services to Loudoun County, Virginia, which is the third most populous county in the state. At the year this study was implemented (2020), the county had a population of almost 421,000 people, including 52% white residents, 21% Asian residents, 14% Hispanic residents, and 7% Black residents (U.S. Census Bureau, 2023). LCADC administrative data revealed that the facility had an average daily population of 222 residents during the study period from November 2020 to October 2021 (ranging from 193 to 233). Just over 80% of these residents were male. In terms of their racial makeup, 51% of residents were white, 24% were Black, 21% were Hispanic, and 3% were Asian. During this period, roughly four-fifths of residents had a length of stay of less than two weeks, while only 4% had a length of stay of over six months. Only 20% of the individuals in the LCADC were serving sentences for misdemeanor or felony convictions, with the remainder awaiting trial.

The LCADC houses residents in eight maximum, medium, and minimum-security units, each of which have one to four pods (for a total of 20 housing pods in the facility). The facility also has four general units that focus on non-housing areas; these include medical, hallway, intake, and transportation. The LCADC provides several reentry and rehabilitation services, including work release, workforce, drug treatment, and mental health programs. Table 1 provides more information on the characteristics of these units.

The LCADC is operated by 124 staff members, consisting of 102 front-line deputies and 22 supervisors. Most of these staff were white and male. In 2019, before the LCADC implemented their BWC program, they conducted a pilot of the cameras with six staff members in their Special Operations Team (SOT). The study team conducted a focus group with the SOT deputies to learn about their experiences using BWCs and learned that the cameras were mostly well-received by the team. The SOT deputies felt that BWCs worked well within their mission and operations. For example, at the start of any event, SOT deputies are required to introduce themselves, announce the ammunition and weapons they have, and what they are being deployed to do. Previously, these statements would be recorded with cell phones, but this process became more efficient with BWCs. The deputies also recognized the potential for BWCs to enhance accountability by creating a record of their actions through video evidence, which could help address any instances of misbehavior.

On the other hand, the deputies also felt that the BWCs could be distracting during RTR events, resulting in team members losing focus of their surroundings during potentially dangerous situations. SOT deputies also conveyed that non-SOT deputies exhibited increased hesitancy in their activities due to the presence of BWCs. This was particularly evident when these deputies observed SOT deputies wearing BWCs, often prompting them to defer to the SOT deputy to lead an interaction with a resident, even if they had originally been responsible for the event. The LCADC did not obtain perspectives about facility-wide deployment of BWCs from deputies outside of the SOT before they launched this program (however, see Peterson et al. [2023] for our examination of deputy perspective during the study period). Nonetheless, the LCADC felt the pilot BWC deployment was a success and aimed to release BWCs facility-wide in 2020. This provided us with a unique opportunity to examine the impact of BWCs on relevant correctional outcomes through a rigorous RCT design.

The study team acted as an independent, third-party research organization that identified the LCADC as a partner at the time the proposal was submitted to the National Institute of Justice. Post award, the study team worked closely with the facility to ensure a transparent and accurate evaluation could be completed. The grant covered the costs associated with the BWC equipment for the study period. In return, the LCADC ensured the study team had access to staff and data to conduct the in-depth evaluation. As part of this partnership, we worked closely with the LCADC to ensure they followed our randomization protocols, which are described in the following section. All human subject related research, including interviews and surveys with supervisors and deputies, and administrative data collection and analyses, were approved by an external institutional review board, the WCG IRB (Study number: 1262879).

There is not a policy specific to BWCs for the LCADC. Instead, the facility relies on the Loudoun County Sheriff’s Office’s general orders policy 411.17 “Body Worn Video Cameras,” which requires deputies to “have BWCs recording during every law enforcement–public encounter related to a call for service or law enforcement action, subject stop, traffic stop, and/or deputy services provided that such activation does not interfere with officer/deputy safety or the safety of others…” (LCSO, 2023b). In the context of the LCADC, deputies were expected to activate their BWCs every time an interaction occurred with a resident. This often resulted in deputies activating their camera for long periods of time, for example, when conducting a unit walk through or work activities in the more public sections of the facility. That said, it is worth noting that we were unable to account for how often deputies turned on their BWCs and/or complied with departmental policies about wearing or activating BWCs during the study period. Instead, we followed an intent-to-treat approach and assumed all deputies in the treatment group used them during resident interactions and RTR events. We know from the police literature that BWC activation rates among officers vary widely (Katz, Choate, Ready, & Nuño, 2014; Katz, Kurtenbach, Choate, & White, 2015; Lawrence, McClure, Malm, Lynch, & La Vigne, 2019). Had we been able to measure activation rates, it is possible that this would have moderated the impact of BWCs on our outcome of interest (see Ariel et al., 2016b; Hedberg, Katz, & Choate, 2017). That said, our review of BWCs from 13 randomly selected RTR incidents found that all deputies involved activated their cameras as expected by policy (Cunningham et al., 2023).

Randomized controlled trial procedures

The evaluation of BWC programs in law enforcement agencies has sparked a scholarly discussion on the challenges associated with contamination in experimental studies (Ariel, Sutherland, & Sherman, 2019; Braga, Coldren, Sousa, Rodriguez, & Alper, 2017; Lawrence & Peterson, 2019). When multiple police officers respond to high priority calls, it can result in contamination when a treatment group officer equipped with a BWC responds to the scene with a control group officer who does not have a camera. Such contamination can compromise the internal validity of the study when deputies who are not equipped with BWCs are involved in the same event as those who have the cameras (Ariel et al., 2019; Braga et al., 2017). Consequently, some researchers suggest that individual-level randomization can lead to misleading experimental results (Ariel et al., 2019).

This issue is particularly challenging in correctional settings where the nature of the environment and the job responsibilities of correctional deputies may make it difficult to avoid contamination. Deputies are often required to work in pairs or teams and may respond to RTRs, fights, or other incidents together, increasing the risk of contamination. As a result, it may be difficult to isolate the impact of BWCs on individual deputies, as the presence of a camera on one deputy may influence the behavior of other deputies involved in the same incident.

One method used to address the issue of contamination in policing studies is to randomize officers based on their business-related groupings, such as by shift, unit, or district (Ariel et al., 2015, 2016a, 2016b; Farrar & Ariel, 2013). Ensuring equivalence between experimental groups using this approach can be quite challenging, especially in settings where the number of available groups for randomization and permanent BWC assignment is limited. In correctional settings, for example, only a few housing units may exist, each with different architectural layouts and populations of residents at varying security levels, making it very difficult to achieve equivalence between randomly selected units.

Instead, randomizing by unit-time has been found to be an effective method for reducing contamination in BWC RCTs (Ariel et al., 2015). This method randomizes all work shifts into BWC and non-BWC conditions at the beginning of a week or month and repeats the process at the beginning of the next work week or month. This ensures that all officers or deputies are either equipped with BWCs or do not have them, reducing concerns of contamination. Additionally, this method ensures equivalency between the experimental groups as the same groups are used repeatedly over time. Though this approach still has limitations (see Lawrence & Peterson, 2019 for a more detailed discussion), it was best suited for this study.

To that end, we executed a clustered RCT in the LCADC between November 2, 2020 and October 31, 2021 (approximately one year). Before the start of each month, the study team randomly assigned the 12 units, comprising 8 housing units and 4 general units, to treatment and control conditions. In other words, six of these units were allocated BWCs for the entirety of the month, during both day and night shifts, while the other six units continued business-as-usual without cameras. We then replicated this at the beginning of each subsequent month, resulting in 12 monthly randomization assignments throughout the study period. These assignments across the study period are summarized in Table 2.

We closely monitored the LCADC’s adherence to this random assignment to ensure fidelity. We found that only 3.0% of all unit-days (131 out of 4,368 [12 units × 364 days]) in the study contained errors, indicating that BWCs were assigned incorrectly or not assigned at all in very few instances. This low error rate was reassuring and indicated that the BWC intervention was aligned with the study’s theoretical design.

Data and Analytic Strategies

We collected data on all RTRs conducted by LCADC deputies during the 364-day study period (i.e., November 2, 2020 to October 31, 2021). The LCADC provided us data on both deputies and residents involved in each RTR event, with a total of 633 deputy responses to 97 RTRs. The number of unique residents involved in the 97 RTR events is unknown (i.e., a single resident could have been involved in more than one RTR). The data fields we analyzed did not vary among deputies within a single RTR event, except for the type of control method used by each deputy (i.e., one deputy may have used a resistance control while another used a physical control against the same resident). Therefore, we aggregated the deputy data to the resident level, resulting in 97 unique RTR events that occurred with one resident and one or more deputies. All analyses were conducted in STATA v. 15.1.

The outcome of interest in the following models was injuries to residents stemming from RTR events. There were a total of 18 resident injuries from the 97 RTR events over the one-year study period. Of note, the LCADC recorded all these injuries as “apparent minor injuries,” while events without injuries were reported as “no injuries noted or visible.” After consultation with LCADC officials, we understood these injuries to include anything less than serious bodily harm. While none of the reported injuries required residents to be transported to an outside medical facility, many of them likely required medical care available within the LCADC. We also explored the possibility of assessing deputy injuries occurring during RTRs, but there were no such injuries reported during our study period.

Count model to examine the change in injury prevalence

Our first analysis examined differences in the number of resident injuries in treatment versus control housing units over the study period. Our unit of analysis is the unit-month since units were randomized at the beginning of each month, resulting in a total of 144 cases (12 units × 12 months). Given that the outcome is a count measure and showed evidence of skewness and overdispersion, we utilized the nbvargr command in STATA to determine whether the Poisson or negative binomial distribution was more appropriate for our data (Hilbe, 2011; Long & Freese, 2006; Macdonald & Lattimore, 2010). The results showed that the negative binomial distribution was a better fit for our resident injury outcome than the Poisson distribution.

We then employed a negative binomial regression model to conduct our analysis. We used the nbreg command in STATA to fit the model. To account for differences in the number of injuries that could occur during an RTR event in each unit, we included the total number of RTR events within each unit-month as an exposure variable in the count models. Not accounting for these differences could lead to biased results (Hutchinson & Holtman, 2005). Furthermore, we employed the vce(cluster clustvar) option to account for the potential correlation of the outcome within the same unit-month. This is because the observations are not independent within the same unit-month, but they are independent across different unit-months. By clustering the standard errors at the unit level, we could obtain robust standard errors that are adjusted for the correlation within the same unit of analysis.

We report the incidence rate ratio (IRR) of the difference in the count of injuries in months with BWCs compared to months without the cameras. IRRs are easier to interpret than the unstandardized coefficients produced by count models. For example, an IRR of 1.10 in our models would indicate that the presence of BWCs increased the number of resident injuries across unit-months by 10%, while an IRR of 0.90 would indicate that BWCs reduced the number injuries by 10%. We also report the predicted margins, which offer a more straightforward interpretation of the results by estimating the average monthly count of injuries in the BWC versus non-BWC units.

The primary variable of interest in the count model was whether deputies in a unit-month had been assigned to wear a BWC. To increase precision of our model, we include additional covariates that were judged in advance to have the potential to influence the outcome. Adding such covariates to an RCT analysis is widely accepted and, in most cases, the preferred approach. An analysis from Lee (2016) examined this issue in detail, and determined that unadjusted, crude analyses in RCTs underestimate the effect sizes of the treatment, and that an unbiased estimate of effect size can be obtained only by adjusting for all predictors of the outcome. Similarly, Kahan, Jairath, Dore, and Morris (2014) conducted simulations to examine the impact of covariate adjustment on 12 outcomes from 8 studies and found that adjustment for predictive covariates can lead to substantial increases in statistical power and should therefore be routinely incorporated into the analyses of RCTs.

As such, we also included four additional covariates in our count model. First, we measured the cumulative number of months each unit was equipped with BWCs, which helped account for the additional experience with BWCs that the deputies may have gained the longer they had been equipped with BWCs (see Peterson & Lawrence's, 2021 “program maturity hypothesis”). The second covariate was the number of deputies per RTR, which was the average number of deputies responding to RTR events during each unit-month. For example, if there were 20 deputies responding to 5 RTRs, the value of this variable would be 4; if there were no RTRs in a unit-month, the value would be 0. Third, we measured the security level of the unit, coded as either a maximum-security unit = 1 or a minimum or medium security unit = 0. Finally, we included the sex of the residents in the unit in our analysis, coded either as male = 1 or female and mixed-sex units = 0. Four of the six units categorized into the mixed-sex category were general units (i.e., medical, hallway, intake, and transportation), while the other two were housing units. There was only one unit that housed solely female residents, and this unit was combined with the mixed-sex category to improve model fit.

Table 3 displays descriptive statistics of the control variables included in the count model on injuries. The table separates this information for the entire study period and for the experimental groups, which are the unit-months with and without BWCs. To ensure equivalence between the groups, we assessed the Cohen’s d effect size and t statistics with values exceeding +/− 0.20 and +/− 1.96, respectively, as an indication of a lack of equivalence. The results indicate that the average number of responding deputies, proportion of maximum-security units, and proportion of male units were equivalent across both groups. These variables were included in the regression models to ensure the integrity of the study’s design.

Logistic model to predict injury

Our second analysis aimed to identify event characteristics that predict resident injury in RTR events. To achieve this, we used the event-level dataset comprising the 97 RTRs and conducted a logistic regression analysis to predict whether the event resulted in an injury (0 = no injury, 1 = injury). This allowed us to consider a range of situational and behavioral factors that could impact resident injuries beyond the aggregated data used in the count model, as detailed in Table 4.

We present results from this analysis using four models, each of which builds off the previous. The first, model A, was an unconditional regression that only included whether BWCs were present or not. This model allowed us to assess the change in resident injuries from the presence of BWCs with no other controls. Model B includes situational information, such as the number of months the unit was assigned BWCs, the work shift of the RTR (day = 0 or night = 1), the number of responding deputies, security level of the unit (max security = 1), sex of the residents in the unit (female/mixed = 0, male = 1), and the purpose of the contact (cell extraction =1 or other [escorting, intake, etc.] = 0).

Model C adds behaviors associated with the resident, including the residents’ assaultive behavior (0 = no, 1 = yes) and resistance level (passive [reference], active, or aggressive). The LCADC defines passive resistance as “a subject’s verbal and/or physical refusal to comply with a deputy’s lawful direction, causing the deputy to use physical techniques to establish control,” active resistance as “a subject using physical evasive movements directed toward the deputy, such as bracing, tensing, pushing, or pulling to prevent the deputy from establishing control over the subject,” and aggressive resistance as “a subject attacking movements toward a deputy that may cause injury but are not likely to cause death or great bodily harm to the deputy or others” (LCSO, 2023a).

Finally, model D adds behaviors associated with the responding deputies, specifically on the control methods used by the deputies during the RTR. The LCADC data included 12 types of control methods, which were recoded into three dichotomous categories (not present = 0, present = 1): (1) Physical controls (arm-bar takedowns, close-quarter strikes, empty-hand controls, and pressure point controls); (2) Restraint controls (use of an emergency restraint chair, handcuffs, a hobble restraint, or a spit-hood); and (3) Weapon Controls (OC spray, personal weapons, or tasers). These categories were not mutually exclusive and some RTRs involved more than one type of control method. Results from these logistic regression models include the odds ratio (OR); we also report the predicted margins to estimate the percentage of RTRs with and without BWCs that resulted in an injury.

Results

Amount of Injuries

Table 5 presents the results of the negative binomial regression models examining the impact that BWCs had on the number of resident injuries. The results indicate a 58% reduction in predicted injuries in unit-months where deputies were assigned BWCs (IRR = 0.42, p < .05). This corresponded to an average estimated count of 0.17 resident injuries in unit-months without cameras, compared to 0.07 injuries in unit-months with cameras. BWCs are clearly associated with a significant reduction in resident injuries.

Neither the number of months the unit was previously assigned BWCs, average number of deputies responding to the RTR, nor whether the event occurred in a maximum-security unit significantly affected the amount of injuries. On the other hand, units that housed male residents experienced 4.88 times more injuries than units that housed female residents or mixed-sex units (p < .001). However, there were only 6 RTRs in all-male units, four of which resulted in injuries. Comparatively, all-female and mixed-sex units experienced 5 and 86 RTRs resulting in 0 and 14 injuries, respectively. It is notable that the majority of RTRs (73 of 97) and injuries (12 of 18) occurred in the facility’s non-housing units (medical, hallways, intake, transportation), most of which occurred in the intake unit (43 RTRs with 8 injuries).

Predictors of Injury

We next examined the event characteristics that predict resident injuries based on the 97 RTRs observed during the year-long study period. Table 6 details the results from the four logistic regression models. Results from model A—the unconditional regression model—indicate that BWCs reduce the likelihood of a resident injury occurring, although the relationship was not statistically significant.

Results are similar across model B (adding situational characteristics), model C (adding resident behaviors), and model D (adding deputy behaviors), and will be discussed in tandem. The main takeaway from these models is that BWCs appear to significantly reduce the likelihood of injuries occurring during RTRs, once we include control variables. As detailed in model D, which controls for situational characteristics, resident behaviors, and deputy behaviors, the likelihood of an injury occurring was reduced by a factor of 0.12 when involving deputies assigned BWCs (OR = 0.12, p < .01). Based on the predicted margins, we would predict injuries to occur in 28.4% of RTRs without BWCs compared to 8.8% of RTRs with BWCs.

Many of the situational characteristics did not significantly relate to whether an injury occurred across all models, including the number of months that BWCs were assigned to the unit, the shift (night versus day), the number of responding deputies, the security level of the unit (maximum security versus other), nor whether the RTR was the result of a cell extraction. Similar to the negative binomial models examining the number of injuries across unit-months, the leading situational factor in these models was if the RTR occurred in an all-male unit (models B, C, & D, p < .01). Again, however, this is most likely due to the low amount of RTRs that occurred in all-male units with high levels of injuries.

Models C and D included aspects of the residents’ behaviors and deputies’ control methods (i.e., force types used). Interestingly, both fell short of statistical significance. Residents’ resistance levels were only marginally related to the likelihood of a resident injury (p < .10), while restraint and weapon controls were just shy of marginal significance (p = .10 and p = .11, respectively). These findings may be explained, at least in part, by the limited statistical power associated with a low base rate of both RTRs and injuries.

Discussion

This study sought to examine the relationship between BWCs and RTR-related injuries to incarcerated residents. Based on a limited body of police research (Henstock, 2015), and the potential suitability of BWCs for correctional environments (Cunningham et al., 2023; Verdon, 2022; Winton, 2021), we hypothesized that BWCs would reduce the number of injuries residents sustained during RTRs. Results from our negative binomial regression model supported this hypothesis, indicating that the unit-months in which deputies were assigned to wear BWCs had 58% fewer injuries than those in the control condition.

In practical terms, we estimate that BWCs prevented approximately 8 resident injuries in the LCADC over the one-year study period (i.e., 5 resident injuries occurred in treatment units versus 13 in control units). While this number is seemingly small, it is not inconsequential. Although there is limited research specific to RTR-related resident injuries, there are several studies that underscore the importance of fair and just treatment in correction facilities. For example, Steiner and Meade (2014) argue that residents’ “perceptions of correctional officer legitimacy will no doubt be influenced by instrumental concerns such as the outcomes of their encounter with staff and their level of satisfaction with those outcomes” (p. 141). Injuries to residents, even when justified, can undermine perceptions of legitimacy. In turn, these perceptions directly affect COs’ ability to maintain institutional social order, effectively enforce rules, and gain compliance from the resident population (Bottoms, 1999; Dilulio, 1987).

Maintaining the safety and wellness of incarcerated residents is also one of the primary goals of correctional administrators (Cressey, 1959; Denhof et al., 2014). Likewise, preventing resident injuries of any kind can enhance an agency’s mission to uphold public safety. For example, while not specific to injuries committed by COs, several studies have linked victimization of incarcerated residents in correctional settings to their subsequent criminal behavior post-release (Boxer et al., 2009; Listwan et al., 2013).

Further, even though the injuries that occurred in the LCADC during our study period were minor, all injuries have financial and legal impacts. As we noted above, even a “minor” injury in this case did not mean it was insignificant. Some excessive use-of-force incidents involve only minor injuries but can result in serious and lengthy legal battles (see Rembert et al., 2023). Further, many of these minor injuries required internal medical attention, which translate to real costs for the LCADC. We also know that, despite the occurrence of only minor injuries during our study, more serious RTR-related injuries do occur in jails (Finton, 2023; Fuller, 2023; Koval, 2023; Maxwell, 2023). It is possible that, if the preventative impact of BWCs in the LCADC holds over time, the Loudoun County Sheriff’s Office may eventually experience fewer serious RTR-related injuries, which have more significant implications for both social and actual costs.

The presence of a documented injury, even a minor one, also has implications for the legal rights of incarcerated people. For instance, uses of force with minor injuries can still be considered “excessive” and lead to serious and lengthy legal battles (see Rembert et al., 2023). Additionally, the Prison Litigation Reform Act created the “physical injury requirement”, which prevents people incarcerated in jails and prisons from filing civil lawsuits against correctional agencies for claims relating to emotional or psychological injuries, without a documented physical injury (Prison Litigation Reform Act, 1996). Although there is legal debate about the application of this requirement and its potential to impede otherwise meritorious lawsuits (Detmold, 2013; Winslow, 2001), the reduction in injuries realized through the LCADC’s BWC program could limit the number of lawsuits filed against the Loudoun County Sheriff’s Office. Lawsuits can be very costly in terms of both litigating the claims and paying out civil judgments to plaintiffs who prevail.

Given the sparse research on RTR-related injuries, our second goal of the study was to identify which characteristics predict the likelihood of an RTR resulting in an injurious outcome. Despite including numerous event-level covariates in our analyses, only the presence of a BWC significantly reduced injuries. The finding bodes well for the many correctional administrators across the US who are planning to or already have deployed BWCs in their facilities.

We also found that the vast majority of both RTRs (73 of 97) and injuries (12 of 18) occurred in the general-purpose areas of the facility, including medical, intake, transportation, and hallways. Of these areas, the intake unit had the highest number of RTRs (43) and injuries (8). These events were much less prevalent in the LCADC’s housing units. Given the dearth of research on use of force in corrections, we cannot comment on the extent to which this finding is consistent with other jails, but the finding certainly has practical value for the LCADC in terms of understanding safety and security risks in their jail. We encourage the LCADC to explore the increased RTR and injury rates in these areas, particularly the intake unit, with an eye toward understanding the factors that increase risk and implementing procedures to reduce that risk. Potential responses could include increasing staff and BWCs in these areas, providing enhanced training (e.g., de-escalation), and enhancing other physical security measures.

Though injuries were less common in the housing units generally, our multivariate analyses determined that RTRs occurring in all-male units had an increased likelihood of an injury. There were only 6 RTRs in the male housing units, but 4 resulted in injury. Again, we do not have data from other jails to serve as a comparison, but the finding is worthy of investigation by the LCADC. Jail administrators could consider providing specialized training for deputies assigned to all-male housing units, such as crisis intervention training, de-escalation, or other tactics that reduce RTR-related injuries (Comartin, Wells, Zacharias, & Kubiak, 2020; Hepburn et al., 2000; Rembert et al., 2023; Tucker et al., 2012). Such training would be especially valuable when deputies engage in pre-planned events in male housing units that are likely to lead to RTRs, such as cell extractions or shakedowns. In addition, given the differential injury rates in male housing units, jail administrators should consider the gender-responsive needs of their residents. There is a large corpus of research on how such correctional management approaches can positively impact residents’ institutional adjustment, safety, and wellness (Bloom, Owen, Covington, & Raeder, 2003; McCampbell, 2005; Wright, Salisbury, & Van Voorhis, 2007; Wright, Van Voorhis, Salisbury, & Bauman, 2012).

In addition to these implications for policy and practice, the study adds to the body of corrections research. There has been extremely limited research to date on RTRs or uses of force in correctional facilities, and even less on injuries sustained during these events. In another manuscript associated with this project, we explore the impact of BWCs on the number of RTRs and other elements of the RTR, such as control methods used by deputies during RTRs and resident resistance levels. We found that RTR events were approximately 40% lower in unit-months with BWCs and associated with significantly lower uses of physical controls by deputies and active resistance by residents (Lawrence, Peterson, White, Cunningham, & Coldren, 2023). This study adds to this scholarship by focusing on RTR-related injuries and the event-level characteristics that predict such outcomes. To our knowledge, there has only been one other study to examine the impact of an intervention (in this case, the deployment of nonlethal weapons) on the likelihood of injuries during correctional uses of force (Hepburn et al., 2000).

Moreover, while there is a limited body of research estimating the percentage of police-officer use-of-force incidents resulting in injuries (Smith et al., 2010), no such research has been conducted in correctional environments. Instead, the public only hears about injuries to incarcerated individuals through court cases (Rembert et al., 2023) or when the news media uncover especially egregious examples of CO abuse (Neff et al., 2023; Santo & Neff, 2023). Our study helps fill this gap, with our results yielding an injury rate of 18.6% (18 resident injuries from 97 RTR events over the one-year study period). These rates seem low compared to the rate of injuries in police officer use-of-force incidents, which have been estimated to be between 17 and 64% (Smith et al., 2010). However, we caution against such comparisons because of the fundamental differences between police and corrections, such as the custodial nature of interactions, the relative lack of guns in correctional facilities, and the more consistent and frequent contacts between deputies and residents.

Finally, though it is beyond the scope of the current paper to explain the specific mechanisms through which BWCs reduce use-of-force related injuries, we can situate our findings within the broader literature on CO culture. Higgins and colleagues argue that CO identities are constructed through “us-them” ideologies rooted in assumptions about the dangers posed by incarcerated people (Higgins, Smith, & Swartz, 2022; Higgins, Swartz, Navarro, & Hughes, 2023). Danger-based othering is ingrained in CO culture and a leading cause of official mistreatment (Higgins et al., 2022). From this, one could potentially argue that by making institutions safer, BWCs can disrupt or mitigate COs perspectives of these dangers. However, our measure of “safety” was based solely on resident outcomes. In an accompanying study, we found that LCADC deputies’ perceptions of safety did not improve after being equipped with BWCs (Peterson et al., 2023; see also Sydes et al., 2022). Thus, while BWCs may lead to COs using less force overall, it is unclear if they are improving or otherwise impacting CO culture.

Rather than alleviating negative beliefs about incarcerated residents, BWCs may simply act as a deterrent to CO misbehavior. We found that not only did the number of RTRs with injuries decrease when LCADC deputies had BWCs (count model), but the likelihood of an RTR resulting in an injury also decreased when BWCs were present (logistic models). In other words, BWC-equipped deputies were more careful about when and how they used force. Schultz (2023) similarly describes use of force as an “organizational behavior—an expected part of the job that officers skillfully negotiate” (2023, 669). COs intentionally use force, even excessive force, to maintain order, but conduct a sort of risk calculus to ensure these actions are justifiable under departmental rules. It stands to reason, then, that BWCs become part of this calculus.

Limitations

There were a few limitations to the study. First, our study was set in a single, mid-sized county jail in northern Virginia. While the LCADC is similar to many other jails across the country in terms of its size, policies, and operational challenges, it is unclear whether our results are generalizable to other correctional agencies. Second, both of our analyses had relatively small samples (i.e., n = 144 unit-months in the negative binomial regression model, and n = 97 RTR events in the logistic regression models) and were based on only 18 RTR-related injuries. It would therefore be valuable for future research to replicate this study in a larger facility or one with a higher base rate of injuries.

Third, the LCADC did not collect housing unit-level data on the resident population, nor were they able to provide data on the characteristics of the deputies and residents involved in RTR events. There were thus several covariates we would have liked to include in our models that may have predicted injuries, but which were not available to us. These include the tenure and sex of the deputies involved in the RTR event; the race or other demographics of the resident; the average size of the population and length of stay of residents per housing unit; and the offense types that resulted in the residents’' incarceration. We believe future research should examine the importance of these variables in understanding jail-related injuries.

Fourth, the full study occurred during the COVID-19 pandemic, which impacted the operations of prisons and jails across the country (Carson, Nadel, & Gaes, 2022). In Loudoun County, officials implemented several policies to mitigate the spread of COVID-19 in the LCADC and minimize in-person contact between staff and residents (Cline, 2020). It is possible that these policy changes resulted in fewer interactions, and thus fewer RTRs and injuries, during the study period than what would normally be expected. However, we could not test this assumption because of a policy change within the LCADC that made it impractical to compare the number of RTRs that occurred during our study to pre-pandemic periods. Specifically, in October 2020 (right before our study), the LCADC transitioned from documenting instances where the deputies directly used force on the residents (i.e., a use of force) to documenting instances when residents resisted deputies in any way (i.e., an RTR), resulting in an immediate measurement-driven increase in the number of incidents recorded after this policy change.

Conclusion

In response to the increasing demands for transparency and accountability in the U.S. justice system, prisons and jail administrators have increasingly turned to BWCs as a tool for enhancing the safety and security of their staff and resident populations. However, very little is known about the extent to which force is employed in these institutions, the rate at which use of force results in injuries to residents, and the efficacy of BWCs in mitigating these harms. Findings from this study help address these critical knowledge gaps. We found that, while RTR-related injuries are relatively rare events, BWCs significantly reduced these outcomes within the LCADC over our one-year study period. These results offer support for correctional policymakers and officials across the country who have already deployed, or are planning to deploy, BWCs in their facilities. Future research should replicate and extend our study in other jail settings and in state and federal prisons.

References

Allard, T. J., Wortley, R. L., & Stewart, A. L. (2006). The purposes of CCTV in prison. Security Journal, 19(1), 58–70.

Allard, T. J., Wortley, R. L., & Stewart, A. L. (2008). The effect of CCTV on prisoner misbehavior. The Prison Journal, 88(3), 404–422.

Ariel, B., Farrar, W. A., & Sutherland, A. (2015). The effect of police body-worn cameras on use of force and citizens’ complaints against the police: A randomized controlled trial. Journal of Quantitative Criminology, 31(3), 509–535.

Ariel, B., Sutherland, A., Henstock, D., Young, J., Drover, P., Sykes, J., Megicks, S., & Henderson, R. (2016a). Wearing body cameras increases assaults against officers and does not reduce police use of force: Results from a global multi-site experiment. European Journal of Criminology, 13(6), 744–755.

Ariel, B., Sutherland, A., Henstock, D., Young, J., Drover, P., Sykes, J., Megicks, S., & Henderson, R. (2016b). Report: increases in police use of force in the presence of body-worn cameras are driven by officer discretion: a protocol-based subgroup analysis of ten randomized experiments. Journal of Experimental Criminology, 12(3), 453–463.

Ariel, B., Sutherland, A., Henstock, D., Young, J., Drover, P., Sykes, J., Megicks, S., & Henderson, R. (2018). Paradoxical effects of self-awareness of being observed: Testing the effect of police body-worn cameras on assaults and aggression against officers. Journal of Experimental Criminology, 14, 19-47.

Ariel, B., Sutherland, A., & Sherman, L. W. (2019). Preventing treatment spillover contamination in criminological field experiments: The case of body-worn police cameras. Journal of Experimental Criminology, 15, 569-591.

Beales, N. & Marsh, L. (2016). On body cameras in prison. Practice: The New Zealand Corrections Journal, 4(1), 39-41.

Bloom, B., Owen, B. A., Covington, S., & Raeder, M. (2003). Gender-responsive strategies: Research, practice, and guiding principles for women offenders. Washington, DC: National Institute of Corrections.

Bottoms, A. E. (1999). Interpersonal violence and social order in prisons. Crime and Justice, 26, 205-281.

Boxer, P., Middlemass, K., & Delorenzo, T. (2009). Exposure to violent crime during incarceration: Effects on psychological adjustment following release. Criminal Justice and Behavior, 36(8), 793-807.

Braga, A., Coldren, J. R., Sousa, W., Rodriguez, D., & Alper, O. (2017). The benefits of body-worn cameras: new findings from a randomized controlled trial at the Las Vegas Metropolitan Police Department. Final report to the National Institute of Justice, 2013-IJCX-0016. Arlington, VA: CNA Corporation.

Braga, A. A., Sousa, W. H., Coldren, J. R., & Rodriguez, D. (2018). The effects of body‐worn cameras on police activity and police‐citizen encounters: A randomized controlled trial. Journal of Criminal Law and Criminology, 108(3), 511–538.

Brodie, D., Kirkland, W., & Kelly, S. (2020). Body-worn cameras in Correctional Settings [Webinar]. BWC Training and Technical Assistance. https://www.bwctta.com/tta/webinars/body-worn-cameras-correctional-settings

Bui, L. (2016, February 23). Maryland county equips some detention officers with body cameras. Washington Post.

Butterfield, F. (2004, May 8). Mistreatment of prisoners is called routine in U.S. New York Times.

California Office of the Inspector General (OIG). (2023). Monitoring the use-of-force review process of the California Department of Corrections and Rehabilitation. Sacramento, CA: California OIG.

Carson, A. E., Nadel, M., & Gaes, G. (2022). Impact of COVID-19 on state and federal prisons, March 2020–February 2021. Washington, DC: Bureau of Justice Statistics.

Census Bureau, U.S. (2023). Table P2: Hispanic or Latino, And Not Hispanic or Latino by Race, 2020 Decennial Census. U.S. Census Bureau. Accessed on February 1, 2023 through the Explore Census Data, available at https://data.census.gov/

Chamberlain, M. (2020). Managing risk associated with video in today’s jails. Lexipol. https://www.lexipol.com/resources/blog/managing-risk-associated-with-video-in-todays-jails/

Chermak, S. M. (1995). Victims in the news: Crime and the American news media. Boulder, CO: Westview Press.

Cline, N. (2020, April 17). Number of inmates at Loudoun’s detention centers dropping amid COVID-19 pandemic. Loudoun Times-Mirror. https://www.loudountimes.com/news/number-of-inmates-at-loudouns-detention-centers-dropping-amid-covid-19-pandemic/article_b20087e2-80c6-11ea-bfb2-efca6b52dca2.html

Clemmer, D. (1940). The prison community. New Braunfels, TX: Christopher Publishing House.

Cloward, R. A. (1960). Social control in the prison. In R. Cloward (Ed.) Theoretical studies in social organization of the prison (pp. 20-48). New York: Social Science Research Council.

Comartin, E. B., Wells, K., Zacharias, A., & Kubiak, S. (2020). The use of the Crisis Intervention Team (CIT) model for corrections officers: Reducing incidents within a county jail. The Prison Journal, 100(5), 581–602.

Crawley, E. (2004). Doing prison work: The public and private lives of prison officers. Cullompton, UK: Willan Publishing.

Cressey, D. R. (1959). Contradictory directives in complex organizations: The case of the prison. Administrative Science Quarterly, 4(1), 1–19. https://doi.org/10.2307/2390646

Cunningham, B. C., Peterson, B. E., Lawrence, D. S., White, M. D., Coldren, J. R., Lafferty, J., Richardson, K. (2023). Comparing the uses and benefits of stationary cameras versus body-worn cameras in a local jail setting. Arlington, VA: CNA Corporation.

Debus-Sherrill, S. A., La Vigne, N. G., & Downey, P. M. (2014). CCTV in jail housing: An evaluation of technology-enhanced supervision. Security Journal, 30(1), 367–384.

Denhof, M. D., Spinaris, C. G., & Morton, G. R. (2014). Occupational stressors in corrections organizations: Types, effects and solutions. Washington, DC: National Institute of Corrections.

Detmold, H. (2013). ‘Tis enough,’ twill serve: defining physical injury under the Prison Litigation Reform Act. Suffolk UL Rev., 46, 1111.

Dilulio, J. (1987) Governing prisons: A comparative study of correctional management. New York: Free Press.

Dodd, S., Antrobus, E., & Sydes, M. (2020). Cameras in corrections: Exploring the views of correctional officers on the introduction of body-worn cameras in prisons. Criminal Justice and Behavior, 47(9), 1190-1208.

Dodd, S., Antrobus, E., & Sydes, M. (2023). Body-worn camera activation in prisons: Understanding correctional officers’ decision-making and use of discretion. Security Journal, 1-21. https://doi.org/10.1057/s41284-023-00380-7

Farrar, T., & Ariel, B. (2013). Self-awareness to being watched and socially desirable behavior: A field experiment on the effect of body-worn cameras and police use of force. Washington, DC: Police Foundation.

Finton, L. (2023. March 2) Surveillance footage from jail shows officers kneeling on inmate’s back for almost 6 minutes. Memphis Commercial Appeal. https://www.commercialappeal.com/story/news/2023/03/02/video-released-of-shelby-county-jail-officers-beating-inmate/69964005007/

Freeman, R. M. (2003). Social distance and discretionary rule enforcement in a women’s prison. The Prison Journal, 83(2), 191-205.

Fuller, L. (2023, March 17). 3 correctional officers accused of beating restrained inmate indicted on federal charges. WSD Local. https://www.wpsdlocal6.com/news/3-correctional-officers-accused-of-beating-restrained-inmate-indicted-on-federal-charges/article_c9562434-f50c-11ed-ba30-0b6ff99f3f33.html

Gaub, J. E., Todak, N., & White, M. D. (2021). The distribution of police use of force across patrol and specialty units: A case study in BWC impact. Journal of Experimental Criminology, 17(4), 545-561.

Groff, E. R., Haberman, C., & Wood, J. D. (2020). The effects of body-worn cameras on police-citizen encounters and police activity: Evaluation of a pilot implementation in Philadelphia, PA. Journal of Experimental Criminology, 16, 463-480.

Haggerty, K. D., & Bucerius, S. M. (2021). Picking battles: Correctional officers, rules, and discretion in prison. Criminology, 59(1), 137-157.

Hedberg, E. C., Katz, C. M., & Choate, D. E. (2017). Body-worn cameras and citizen interactions with police officers: Estimating plausible effects given varying compliance levels. Justice Quarterly, 34(4), 627–651.

Henry, P. (1994). The 1993 national survey on use of force incidents and procedures in correctional institutions. An unpublished report prepared for the American Correctional Association under the Prison Setting Field Evaluation Project. Alexandria, VA.

Henstock, D. (2015). Testing the effects of body worn video on police use of force during an arrest: A randomized controlled trial. Unpublished master’s thesis. Wolfson College, Cambridge: UK.

Hepburn, J. R., Griffin, M. L., & Petrocelli, M. (2000). Safety and control in a county jail: Nonlethal weapons and the use of force. Final Report submitted to the National Institute of Justice.

Higgins, E. M., Smith, J., & Swartz, K. (2022). “We keep the nightmares in their cages”: Correctional culture, identity, and the warped badge of honor. Criminology, 60(3), 429-454.

Higgins, E. M., Swartz, K., Navarro, J. C., & Hughes, K. (2023). ‘They might not kill you today but they’re going to get you in the end’: The correctional subculture and the schematization of danger. The British Journal of Criminology, https://doi.org/10.1093/bjc/azac099

Hilbe, J. M. (2011). Negative binomial regression. Cambridge: Cambridge University Press.

Hopwood, S. (2021). How atrocious prisons conditions make us all less safe. New York, NY: Brennan Center for Justice.

Hutchinson, M. K., & Holtman, M. C. (2005). Analysis of count data using poisson regression. Research in Nursing & Health, 28(5), 408-418.

Jennings, W. G., Lynch, M. D., & Fridell, L. A. (2015). Evaluating the impact of police officer body-worn cameras (BWCs) on response-to-resistance and serious external complaints: Evidence from the Orlando police department (OPD) experience utilizing a randomized controlled experiment. Journal of Criminal Justice, 43(6), 480-486.

Kahan, B. C., Jairath, V., Dore, C. J., & Morris, T. P. (2014). The risks and rewards of covariate adjustment in randomized trials: An assessment of 12 outcomes from 8 studies. Trials, 15, 139.

Katz, C.M., Choate, D., Ready, J., & Nuño, L. (2014). Evaluating the impact of officer worn body cameras in the Phoenix Police Department. Phoenix, AZ: Arizona State University, Center for Violence Prevention and Community Safety.

Katz, C. M., Kurtenbach, M., Choate, D. E., & White, M. D. (2015). Phoenix, Arizona, smart policing initiative: Evaluating the impact of police officer body-worn cameras. Washington, DC: Bureau of Justice Assistance.

King, R. D., & Elliott, K. W. (1977). Albany: Birth of a prison–end of an era. London, UK: Routledge & Kegan Paul.

Koval, K. (2023, April 19). Video shows Cook County correctional officer beating inmate ‘more than 30 times’. WGN TV. https://wgntv.com/news/chicago-news/video-shows-cook-county-correctional-officer-beating-inmate-more-than-30-times/

Lawrence, D.S., & Peterson, B. E. (2019). How do body-worn cameras affect the amount and makeup of police-initiated activities? A randomized controlled trial in Milwaukee, Wisconsin. Journal of Experimental Criminology, 16, 481-503.

Lawrence, D. S., McClure, D., Malm, A., Lynch, M., & La Vigne, N. G. (2019). Activation of body-worn cameras: Variation by officer, over time, and by policing activity. Criminal Justice Review, 44(3), 339–355.

Lawrence, D. S., Peterson, B. E., Robin, L., & Shukla, R. (2022). The impact of correctional CCTV cameras on infractions and investigations: A synthetic control approach to evaluating surveillance system upgrades in a Minnesota prison. Criminal Justice Policy Review, 33(8), 843–869.

Lawrence, D.S., Peterson, B.E., White, M.D., Cunningham, B.C., Coldren, J.R. (2023). Effects of correctional body-worn cameras on responses to resistance: A randomized controlled trial in a jail setting. Manuscript submitted for publication.

Lee, P. H. (2016). Covariate adjustments in randomized controlled trials increased study power and reduced biasedness of effect size estimation. Journal of Clinical Epidemiology, 76, 137-146.

Liebling, A. (2000). Prison officers, policing and the use of discretion. Theoretical Criminology, 4(3), 333-357.

Liebling, A., & Price, D. (2003). Prison officers and the use of discretion. In L. Gelsthorpe & N. Padfield (Eds.), Exercising discretion (pp. 77– 96). Cullompton, UK: Willan Publishing.

Listwan, S. J., Sullivan, C. J., Agnew, R., Cullen, F. T., & Colvin, M. (2013). The pains of imprisonment revisited: The impact of strain on inmate recidivism. Justice Quarterly, 30(1), 144-168.

Long, J. S., & Freese, J. (2006). Regression models for categorical and limited dependent variables using Stata. College Station, TX: Stata Press.

Loudoun County Sheriff’s Office (LCSO). (2023a). LCSO General Orders, Policy 403.1 Use of Force. Loudoun County, VA: LCSO. Retrieved from: https://public.powerdms.com/LOUDOUNVA/tree/documents/115725

Loudoun County Sheriff’s Office (LCSO). (2023b). LCSO General Orders, Policy 411.17 Body Worn Video Cameras. Loudoun County, VA: LCSO. Retrieved from: https://public.powerdms.com/LOUDOUNVA/tree/documents/115824

Lum, C., Koper, C. S., Wilson, D. B., Stoltz, M., Goodier, M., Eggins, E., Higginson, A., & Mazerolle, L. (2020). Body-worn cameras’ effects on police officers and citizen behavior: A systematic review. Campbell Systematic Reviews, 16(3).

MacDonald, J. M., & Lattimore, P. K. (2010). Count models in criminology. In A. R. Piquero & D. Weisburd (Eds.), Handbook of quantitative criminology (pp. 683–698). New York: Springer.

Maruschak, L. M., Bronson, J., & Alper, M. (2021). Indicators of mental health problems reported by prisoners. US Department of Justice, Office of Justice Programs, Bureau of Justice Statistics.

Marquart, J. W. (1986). Prison guards and the use of physical coercion as a mechanism of prisoner control. Criminology, 24(2), 347-366.

Maxwell, A. (2023, May 18). Three former Camden County Sheriff’s Office employees indicted after inmate beating. News 4 Jax. https://www.news4jax.com/i-team/2023/05/18/3-camden-county-sheriffs-officer-employees-indicted-after-inmate-beating/

McCampbell, S. W. (2005). The gender-responsive strategies project: Jail applications. Gender-responsive strategies for women offenders. Washington, DC: National Institute of Corrections.

Neff, J., Santo, S., & Meagher, T. (2023, May 22). How a “Blue Wall” inside N.Y. state prisons protects abusive guards. New York Times. https://www.nytimes.com/2023/05/22/nyregion/ny-state-prison-guards-abuse.html

Novisky, M. A., & Peralta, R. L. (2020). Gladiator school: Returning citizens’ experiences with secondary violence exposure in prison. Victims & Offenders, 15(5), 594-618.

O’Connor, J. (2023, March 16). Illinois prison guard gets 20 years for inmate beating death. AP News. https://apnews.com/article/prison-beating-guard-sentencing-illinois-5303b0aa40bf3b883d07ef62c688426d

Peterson, B. (2023, March 31). How can body worn cameras maximize police accountability? New analytic tools might be the answer. In-Depth (CNA Corporation). https://www.cna.org/our-media/indepth/2023/03/how-can-body-worn-cameras-maximize-police-accountability

Peterson, B. E., & Lawrence, D. S. (2021). Do the effects of police body-worn cameras on use of force and complaints change over time? Results from a panel analysis in the Milwaukee Police Department. Criminal Justice and Behavior, 48(6), 734-754.

Peterson, B. E., White, M. D., Cunningham, B. C., Lawrence, D. S., Lafferty, J. & Coldren, J. R. (2023). Body-worn cameras in a correctional setting: Assessing jail deputy attitudes before, during, and after implementation. Justice Evaluation Journal. https://doi.org/10.1080/24751979.2023.2244039

Peterson, B. E., Yu, L., La Vigne, N., & Lawrence, D. S. (2018). The Milwaukee Police Department’s body-worn camera program. Washington, DC: Urban Institute.

Police Executive Research Forum (PERF). (2018). Cost and benefits of body-worn camera deployment. Washington, DC: Police Executive Research Forum.

President’s Task Force on 21st Century Policing. (2015). Final report of the President’s Task Force on 21st Century Policing. Washington, DC: Office of Community Oriented Policing Services.

Prison Litigation Reform Act, 42 U.S.C. § 1997e. (1996). https://www.govinfo.gov/content/pkg/USCODE-2021-title42/pdf/USCODE-2021-title42-chap21-subchapI-A-sec1997e.pdf

Prison Policy Initiative. (n.d.). A guide to public records requests for advocates seeking reform of the criminal legal system. https://www.prisonpolicy.org/trainings/records.html

Rembert , D. A., Joseph, J. J., Threadcraft-Walker, W., Threadcraft, M., Brown, D., Soyele, O. E., & Henderson, H. (2023). Criminal liability for correctional officer excessive use of force. Crime, Law and Social Change, 79, 105-128.

Santo, A., & Neff, J. (2023, May 22). We spent two years investigating abuse by prison guards in New York. here are five takeaways. The Marshal Project. https://www.themarshallproject.org/2023/05/22/new-york-prison-corrections-officer-discipline-findings

Schultz , W. J. (2023). Correctional officers and the use of force as an organizational behavior. Criminology, 61(3), 654-675.

Smith, M. R., Kaminski, R. J., Alpert, G. P., Fridell, L. A., MacDonald, J., & Kubu, B. (2010). A multi-method evaluation of police use of force outcomes: Final report to the National Institute of Justice. Washington, DC: National Institute of Justice.

Specter, D. (2006). Making prisons safe: Strategies for reducing violence. Washington University Journal of Law & Policy, 22, 125-134.

Steiner, B., & Meade, B. (2014). The safe prison. In F. T. Cullen, C. L. Jonson, & M. K. Stohr (Eds.). The American prison: Imagining a different future (pp. 129-150). Sage.

Stolzenberg, L., D’Alessio, S. J., & Flexon, J. L. (2019). Eyes on the street: Police use of body-worn cameras in Miami-Dade County. Davie, FL: Weston Publishing, LLC.

Sydes, M., Dodd, S., & Antrobus, E. (2022). Body cameras behind bars: Exploring correctional officers’ feelings of safety with body-worn cameras. Criminology & Criminal Justice, 22(2), 323-342.

Sykes, G. M. (1958) The Society of Captives: A Study of a maximum security prison. Princeton, NJ: Princeton University Press.

Symkovych, A. (2019). The legal and illegal use of force by prison officers in Ukraine. The Prison Journal, 99(1), 89-111.

Tucker, A. S., Van Hasselt, V. B., Mendez, J., Palmer, L., & Browning, S. L. (2012). Crisis Intervention Team (CIT) Training in the jail/detention setting: A case illustration. International Journal of Emergency Mental Health, 14(3), 209–216.

U.S. Department of Justice. (2022, September 21). Four Former Georgia correctional officers sentenced for assaults on inmate and cover-up [Press release]. https://www.justice.gov/opa/pr/four-former-georgia-correctional-officers-sentenced-assaults-inmate-and-cover

Verdon, K. (2022, March 4) San Diego County jails will use body cameras following damning state audit. Reason. https://reason.com/2022/03/04/san-diego-county-jails-will-use-body-cameras-following-damning-state-audit/

Walker, D. (2022). “There’s a camera everywhere”: How citizen journalists, cellphones, and technology shape coverage of police shootings. Journalism Practice, 16(10), 2022-2039.

Welsh-Huggins, A. (2021, October 29). Prison systems adding body-worn cameras to security plans. Courthouse News Service. https://www.courthousenews.com/prison-systems-adding-body-worn-cameras-to-security-plans/

White, M. D., Gaub, J. E., Malm, A., Padilla, K. (2023a). Impacts of BWCs on use of force: Directory of outcomes. Body-Worn Camera Training and Technical Assistance Program. https://bwctta.com/impacts-bwcs-use-force-directory-outcomes

White, M. D., Gaub, J. E., Malm, A., Padilla, K. (2023b). Impacts of BWCs on citizen complaints: Directory of outcomes. Body-Worn Camera Training and Technical Assistance Program. https://bwctta.com/impact-bwcs-citizen-complaints-directory-outcomes

Wilson, K., Eaton, J., Foye, U., Ellis, M., Thomas, E., & Simpson, A. (2022). What evidence supports the use of body worn cameras in mental health inpatient wards? A systematic review and narrative synthesis of the effects of body worn cameras in public sector services. International Journal of Mental Health Nursing, 31(2), 260-277.

Winslow, J. (2001). The prison litigation reform act’s physical injury requirement bars meritorious lawsuits: Was it meant to. UCLA L. Rev., 49, 1655.

Winton, R. (2021, March 12). Federal judge orders body cams in five state prisons citing abuse of disabled inmates. Los Angeles Times. https://www.latimes.com/california/story/2021-03-12/cameras-in-five-prisons-including-lancaster-after-abuse-of-disabled-inmates-judge-rules

Wright, E. M., Salisbury, E. J., & Van Voorhis, P. (2007). Predicting the prison misconducts of women offenders: The importance of gender-responsive needs. Journal of Contemporary Criminal Justice, 23(4), 310-340.

Wright, E. M., Van Voorhis, P., Salisbury, E. J., & Bauman, A. (2012). Gender-responsive lessons learned and policy implications for women in prison: A review. Criminal Justice and Behavior, 39(12), 1612-1632.

Yokum, D., Ravishankar, A., & Coppock, A. (2019). A randomized control trial evaluating the effects of police body-worn cameras. Proceedings of the National Academy of Sciences, 116(21), 10329-10332.

Corresponding author

Daniel S. Lawrence, [email protected], CNA Corporation, 3003 Washington Boulevard, Arlington, VA 22201-2194, USA.

Acknowledgments

This project was supported by Award No. 2018-75-CX-0019, awarded by the National Institute of Justice, Office of Justice Programs, U.S. Department of Justice. The opinions, findings, and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect those of the Department of Justice.

Declarations of interest statement

The authors have no conflicts of interest to report associated with this project.