DO POSITIVE PERCEPTIONS OF CORRECTIONAL STAFF MITIGATE INSTITUTIONAL VIOLENCE AMONG YOUTHFUL OFFENDERS?

Abstract

Juvenile justice facilities can be dangerous places for adolescents and may promote violent behavior among incarcerated youth. With high rates of violence among detained juveniles, youth who do not feel safe may resort to violent behavior to protect themselves. However, this “cycle of violence” may be interrupted if youth can turn to correctional staff for support. Using a 3-wave, longitudinal sample of 373 male incarcerated adolescents, the results indicated that as compared to violence exposure prior to incarceration, exposure to institutional violence more strongly predicts violent behavior. Further, the findings indicated that perceptions of staff serve as a buffer to violence exposure and may interrupt the cycle of violence. Policy implications are discussed.

Introduction

Adolescents who have been exposed to violence are at an increased risk of engaging in violence themselves (Fehon, Grilo, & Lipschitz, 2005). Indeed, it has long been understood that regardless of whether youths have been directly victimized or have merely witnessed violent behavior, the presence of violence in youths’ lives increases the likelihood that youth will engage in violent crime (Smith & Thornberry, 1995; Fagan, 2005), perpetuating the “cycle of violence” (Widom, 1989). There is such a strong association between exposure to violence (ETV) and aggressive behavior that it accounts for violence above and beyond other factors, such as psychotic disorders (Green, Browne, & Chou, 2017).

As compared to youth in the community, youth in the juvenile justice system have typically been exposed to more violence prior to their institutionalization (Wolff & Shi, 2012; Abram, Teplin, Charles, Longworth, McClelland, & Dulcan, 2004). Considering the high rates of violence among inmates in secure facilities (Wolff, Blitz, Shi, Siegel, & Bachman, 2007), incarcerated youth are at risk for further ETV while behind bars. Exposure to violence within the facility may set the tone for youth violence, potentially even outweighing their prior violence exposure. Understanding the etiology of violence within institutional settings is critically important, because youth who engage in institutional violence commonly face additional punitive sanctions, loss of privileges such as phone calls or recreation time, and solitary confinement. These harsh punishments can detrimentally impact their development, and often further increase their likelihood of using violence (de Valk, van der Helm, Beld, Schaftenaar, Kuiper, & Stams, 2015).

Researchers have begun identifying factors that impact the relation between ETV and violent behavior. For instance, adolescents’ perceptions of law enforcement and the justice system may influence the association between violence exposure and violence commission (Baron, 2017). In community samples, the effects of ETV are often linked to perceptions of law enforcement within a “code of the street” framework, a mindset that promotes violence in order to ward off victimization and maintain respect (Anderson, 1999). Adherence to this code is associated with distrust in law enforcement officers and negative perceptions of justice system legitimacy (Anderson, 1999). In other words, youth may be more likely to engage in violence if they do not trust law enforcement to protect them. Indeed, prior research has explored the relation between negative perceptions of law enforcement and adolescent offending in the community (Tyler & Fagan, 2008).

However, less is known about the effect of adolescents’ perceptions once they become incarcerated. Incarcerated youth often turn to institutional staff for emotional support in their otherwise chaotic environment (Barnert et al., 2015). When staff function as positive role models with the capacity to respectfully and diplomatically resolve conflicts, this may discourage youth from engaging in violence within the facility. Nonetheless, little is known about the relation between adolescents’ attitudes toward correctional staff and willingness to engage in violent behavior while incarcerated. Because detained youth are likely to experience ETV both in and out of the institution and may suffer developmentally detrimental consequences if they become involved in institutional violence, it is important to examine the role that adolescents’ perceptions of correctional officers play in adolescents’ likelihood of resorting to violence while incarcerated.

Exposure to Violence and Victimization

Exposure to violence and victimization are associated with numerous social and emotional problems, including increased risk of depression and anxiety, substance abuse, posttraumatic stress disorder (PTSD), poor academic performance and self-control, low levels of autonomy, and the use of violence as a coping mechanism (Lansford, Dodge, Pettit, Bates, Crozier, & Kaplow, 2002; Wolff & Shi, 2012; Fehon, Grilo, & Lipschitz, 2001). When youth are exposed to verbal and physical aggression, they are also at risk of becoming ensnared in the “cycle of violence” that leads to the perpetration of violent acts. The more violence that children witness or directly experience in their homes, schools, and communities, the more likely they are to employ violent tactics in their own lives (Widom, 2017). Youth who are verbally or physically abused by peers may also get stuck in a destructive cycle of victimization. Children who are bullied have been found to develop negative cognitive schemas and poor self-esteem, which contribute to low social expectations and subsequently altered social behavior. This ultimately results in a self-fulfilling prophecy of being continuously targeted for abuse (Calvete, Fernández-González, González-Cabrera, & Gámez-Guadix, 2018). The effects of early and continuous violence exposure such as this have been shown to have lasting effects, predicting violent behavior into adulthood (Bosqui, Shannon, Tiernan, Beattie, Ferguson, & Mulholland, 2014). ETV during adolescence appears to be particularly detrimental, as it is more strongly linked to violence perpetration than during other developmental periods (Benedini & Fagan, 2018).

Researchers and practitioners often ask whether direct or indirect victimization is more harmful. The research appears to suggest that direct victimization may be a more robust predictor of future violence (Shields et al., 2010), yet witnessing physical, sexual, or emotional abuse also significantly increases the likelihood of becoming involved in violent criminal activity (Currie & Tekin, 2012). As compared to community youth samples, juvenile delinquents not only report higher rates of victimization, but are also often exposed to multiple kinds of violence (Abram et al., 2004).

Consequently, provided ETV’s prevalence among adjudicated youth and its robust association with violent reoffending (Rivera & Widom, 1990), the juvenile justice system is increasingly recognizing the need to incorporate ETV as an individual factor for recidivism on risk assessment tools (Holloway, Cruise, Morin, Kaufman, & Steele, 2018). Extant research has examined the connection between ETV and the commission of violence by exploring various mediating factors and underlying mechanisms, including anger, callous-unemotional traits, PTSD (Kimonis, Ray, Branch, & Cauffman, 2011), social learning, information processing (Bandura, 1973; Buckley, Blanchard, & Neill, 2000; Shahinfar, Kupersmidt, & Matza, 2001), and empathy (McCloskey & Lichter, 2003). However, the majority of these studies focused on ETV prior to incarceration, or on community samples that have limited or no contact with the justice system. Considering the cumulative effect of ETV (Haynie, Petts, Maimon, & Piquero, 2009) and the likelihood of ETV during incarceration, there is a particular need to study its impact on youth in the justice system.

Once an adolescent comes into contact with the juvenile justice system, they are at an increased risk for further violence exposure (Wolff et al., 2007). While it is difficult to establish exact rates of violence within facilities (Wolff, Shi, & Bachman, 2008), it is clear that youthful offenders are more likely to experience verbal, physical, and sexual abuse during incarceration. ETV within the justice system certainly impacts offenders’ wellbeing while they are incarcerated, but institutional ETV can also have long-term effects after inmates are released into the community.

In adult samples, ETV while incarcerated is associated with higher levels of aggressive and antisocial behavior post-release, as well as increased emotional distress (Boxer, Middlemass, & Delorenzo, 2005). Detained adolescents exposed to traumatic violence are more likely to display symptoms of PTSD (Abram et al., 2004) and are more likely to recidivate (Chang, Chen, & Brownson, 2003). Further, as compared to adults, adolescents are significantly more likely to engage in institutional misconduct. Among newly admitted youthful offenders, prior ETV is linked with poor institutional adjustment and difficulty complying with authority (DeLisi et al., 2009) which may make them a target for violence. Research suggests that prior ETV is also associated with poorer perceptions of safety and fairness within the institution, which is in turn related to more institutional infractions (Lujan & Faniff, 2018). Taking prior ETV and institutional ETV into account, adolescents who are left feeling unprotected may be more likely to resort to violence to cope with the stress of being incarcerated. In sum, the violent context in which youth become embedded during incarceration may become a model for their behavior.

Adolescents’ Perceptions of Institutional Correctional Staff

Adolescents’ perceptions of the justice system have been linked to crime involvement, such that those who view law enforcement negatively tend to commit more crimes (Fagan & Tyler, 2005; Fine & Cauffman, 2015). Youth develop these perceptions based on their individual interactions with various types of law enforcement officers (Brunson, 2007; Schuck, 2013), as well as vicariously through the experiences of friends and family (Fine, Cavanagh, Donley, Steinberg, Frick, & Cauffman, 2016; Cavanagh & Cauffman, 2015). In other words, studies suggest that adolescents consider both their own personal experiences with police and the way that officers treat their communities, and then make decisions about whether or not to trust the justice system to protect them. The association between negative perceptions of law enforcement and crime is inherent in Anderson’s (1999) “code of the street,” a mentality that supports taking justice into your own hands, often by violent means, if you do not trust law enforcement to do so.

Anderson posits that adherence to this “code” becomes more stable over time, and at least one longitudinal analysis supports the assertion that those who subscribe to the violent behavioral code prior to incarceration continue to use this paradigm to guide their behavior in detention (Mears, Stewart, Siennick, & Simons, 2013). However, other research has found that, while this distrust in law enforcement may be relatively stable overall, perceptions may be somewhat malleable through adolescence (Moule, Burt, Stewart, & Simons, 2015). If youth with negative perceptions experience positive interactions with correctional officers in critical moments, it is possible that this may discourage reliance on violent street code.

In order to prevent exacerbating a violent cycle, law enforcement and justice system officials need to be cognizant of how they treat delinquent youth, even while holding them accountable for crimes. This is especially important when interacting with adolescents who are incarcerated. Incarcerated adolescents’ perception of their likelihood of success upon release is associated with the quality of a relationship with a staff member whom they turn to for support while detained (Marsh & Evans, 2009). Youth who have close relationships with staff report a more positive outlook on their chances of staying out of crime and avoiding substance abuse post-release, as well as their ability to resolve conflicts. Positive perceptions of juvenile justice institutions and perceived fairness of staff members are also linked to fewer antisocial outcomes one year after adolescents have been released (Schubert, Mulvey, Loughran, & Losoya, 2012). Altogether, these studies suggest that positive relationships with correctional staff may be strongly related to positive outcomes in youths’ lives, including a reduced risk of engaging in violence.

The Present Study

The present study extends the body of research on youth exposure to violence by investigating the cycle of violence within juvenile justice facilities. Given the immediate as well as long-term ramifications of perpetrating institutional violence, it is crucial to examine the salience of ETV for incarcerated adolescents and the mitigating role staff might play. With high rates of violence observed in institutional facilities, this study may have particular relevance for correctional agencies who are trying to reduce violence among inmates.

Using a sample of racially and ethnically diverse incarcerated male adolescents, the majority of whom were convicted of a violent crime, this study tests if youth who have been exposed to violence prior to entering the incarceration facility (pETV) are more likely to engage in violent behavior within the facility. In line with previous research, we expect to find that higher levels of pETV will be associated with more violent behavior while incarcerated. Then, we examine the impact of institutional violence exposure (iETV) on subsequent violent behavior, with the expectation that iETV will also be strongly associated with institutional violence, even among youth without pETV. Critically, we hypothesize that the culture of institutional violence in which youth become immersed will set the tone for their own actions. Finally, we investigate if adolescents’ perceptions of staff within the facility reduce the effects of ETV. Specifically, we hypothesize that youth with more positive perceptions of staff will be less likely to engage in institutional violence, likely because they view staff as a legitimate authority and have greater confidence in the staff’s ability to protect them. Uniquely, the longitudinal nature of the data enables us to examine iETV and perceptions of staff while accounting for pre-institutionalization exposure to violence and previous violent behavior.

Method

Participants

The sample consisted of 373 adjudicated adolescent male offenders between the ages of 14 and 17 (M = 16.42, SD = .80, range = 14–17) who were incarcerated in a secure juvenile facility in southern California. Descriptive statistics are provided in Table 1. The sample’s racial/ethnic characteristics was consistent with incarcerated youth in similar juvenile justice facilities in California at the time data was collected (Brown, 2006) and is consistent with the current juvenile detention population (Becerra, 2018).

Table 1.

Descriptive Statistics

Variable	Percent of Sample
Race/Ethnicity
Hispanic/Latino	0.51
Black/African American	0.32
Other/Multiracial	0.12
Caucasian/White	0.06
Committing Offense
Person	0.77
Property	0.25
Technical Violation	0.16
Status	0.05
Public Order	0.40
Drug	0.16

Procedures

All youths between the ages of 14 and 17 who were newly admitted to the facility or readmitted to the facility for a new offense were eligible to enroll in the study. Ninety-seven percent of the parents contacted consented to their child’s participation, and all youth provided their assent. To guarantee privacy of responses, a Confidentiality Certificate was obtained from the Department of Health and Human Services. After the consent/assent process, participants were administered a baseline interview by a trained research assistant within 48 hours of arrival to the facility (Time 1). Participants completed a second interview approximately two weeks after the baseline (Time 2) and a third interview one week later (Time 3; i.e., three weeks after baseline). The interviews took between one and two hours to complete, were conducted individually, and consisted of a series of behavioral, attitudinal, and contextual measures. To counter any reading challenges, research assistants were trained read each question aloud to the participant in a standardized manner, offer any clarification while remaining neutral to possible responses, and record his response. This study protocol was approved and overseen by the Institutional Review Board at the University of California, Irvine.

Measures

Prior Exposure to Violence.

At baseline (Time 1), a modified version of the self-report Exposure to Violence Inventory (ETV; Selner-O’Hagan, Kindlon, Buka, Raudenbush, & Earls, 1998) was administered. In this version, seven items assessed the variety of types of violent acts the youth had witnessed or experienced (e.g., mugging, beating, rape, attack). A variety score was calculated that indicated the number of different types of violent acts the youth had been exposed to by the time of the baseline interview (M = 3.71, SD = 1.62, range = 0–7). By utilizing a variety score, the severity of violence exposure is more easily captured. Exposure to multiple types of violence implies exposure to much more serious violence, and poly-victimization is linked to worse developmental outcomes (Finkelhor, Turner, Ormrod, Hamby, & Kracke, 2009).

Exposure to Violence in the Institution.

Four items derived from the self-report Exposure to Violence Inventory (ETV; Selner-O’Hagan, Kindlon, Buka, Raudenbush, & Earls, 1998) were administered at week 2 (Time 2) to assess exposure to violence with in the institution (e.g., seen someone physically attack staff; seen someone else attack an inmate; seen someone be verbally aggressive to staff; seen someone be verbally aggressive to an inmate). A variety score was calculated to indicate the number of different types of violent acts the youth had been exposed to in the facility (M = 1.54, SD = 1.21, range = 0–4). Considering the results were the same when the non-physical violence items were omitted, the full composite index was included in the final models with higher scores indicating more exposure to violence.

Perceptions of Staff.

A total of 8 self-report items were administered at week 2 (Time 2) to assess youths’ perceptions of staff. Five items were developed based on prior scales (Coates, Miller, & Ohlin, 1978; Moos, 1997) that assess youths’ perceptions of facility staff (e.g., “Staff deals fairly and squarely with all kids.” “Most of the staff here are pretty fair. “ “If a kid screws up, the staff tries to help the kid understand why it happened.” “If a kid has trouble getting along with someone in the program, staff will help him find ways to avoid fights.” “Most of the staff here don’t really care what happens to you; they’re just doing their job [reverse coded]. “). Youth self-reported their responses on a 5-point Likert scale ranging from strongly disagree (1) to strongly agree (5). Three additional items were designed for this study to assess youths’ perceptions of inappropriate staff behavior (e.g., Swear at kids; Use disrespectful language towards kids; Use racist comments towards kids). Youth self-reported their responses on a 5- point Likert scale ranging from Never (1) to Once a week or more (5). These three items were reverse coded. The eight items loaded onto a single factor (one eigenvalue > 1 = 2.50, variance = .86) and the scale was internally consistent (α = .762). Removing any item had a negligible impact on the alpha (range= −.76, .04), thus all eight items were retained for analysis. The final measure was created by mean-scoring the 8 items such that higher scores indicate more positive perceptions of staff (M = 3.40, SD = .65, range = 1.38, 5.00).

Violent Behavior.

A modified version of the Self-Report of Offending (SRO) (Huizinga, Esbensen, & Weihar, 1991) scale was administered at the week 2 (Time 2) and week 3 (Time 3) interviews. Participants were asked whether they had engaged in four violent acts since the prior interview (e.g., “beaten up, mugged, or seriously threatened another person,” “attacked someone with a weapon,” and “taken someone else’s things by force “). Due to the limited variability and low frequencies during a small time scale, dichotomous indicators were calculated to indicate whether the youth had committed a violent act during week 2 (15.14%) and week 3 (13.15%). Violence during time 2 was associated with violence during time 3, χ²(1) = 27.68, p < .001, Cramer’s V= .31, so we utilized the time 2 violence variable as a key independent variable and time 3 violence as the dependent variable.

Missing Data

Of the 373 youth who were sampled at baseline, 282 youth (75.6%) had complete data on all study variables. Compared to youth with incomplete data, those with complete data had no age differences (t(371) = −.35, p = .727, d = −.04), did not vary by race (χ²(1) = 1.89, p = .169), had similar perceptions of staff (t(315) = 1.23, p = .219, d = .22), and engaged in the same amount of violent offending within the institution at week 2 (χ²(1) = .02, p = .881) and week 3 (χ²(1) = 2.14, p = .143). However, as compared to youth with incomplete data, youth with complete data had been previously exposed to somewhat more violence (t(371) = −1.84, p = .066, d = −.22) and were exposed to more violence within the institution (t(371) = −9.37, p < .001, d = - 1.13). As compared with the youth with incomplete data, the sample of youth with complete data was exposed to somewhat more violence and thus represents a particularly policy-relevant group.

Analytic Plan

The key dependent variable was violence during week 3. All models were logistic regressions that accounted for a host of covariates: age, race (Hispanic/Latino coded 1; non-Hispanic coded 0), whether they had been adjudicated on an offense against a person (coded 1), time in the facility (number of days since the baseline interview), and violent behavior during week 2. The first model examined the association between exposure to violence prior to institutionalization (pETV) with violent behavior in the facility. Exposure to violence within the facility (iETV) was included as a key independent variable in the second model. In the third model, perceptions of staff were included to examine whether they reduced the association between exposure to violence within the facility and violent behavior.

Results

Bivariate associations between all key study variables are presented in Table 2. The first logistic regression model examined the association between exposure to violence prior to institutionalization and violent behavior, accounting for the covariates (age, race, adjudicating offense, time, and prior violent behavior). The results indicated that as expected, the more youth had been exposed to violence prior to entering the facility, the more likely they were to engage in violence in the facility (Table 3).

Table 2.

Bivariate Correlation Matrix

		1	2	3	4	5	6	7	8
1.	Age	-
2.	RaceAB	0.02	-
3.	Adjudicating OffenseAC	−0.15*	−0.11	-
4.	Time in Facility	−0.02	0.00	−0.06	−
5.	Exposure to Violence Prior to Incarceration	0.01	0.10	−0.03	0.02	-
6.	Exposure to Violence at Time 2	0.03	−0.15*	0.03	−0.10	0.29***	-
7.	Perceptions of Staff at Time 2	−0.09	0.08	0.02	0.03	−0.21***	−0.34***	-
8.	Violent Behavior at Time 2A	0.05	0.06	−0.03	−0.08	0.11	0.20***	−0.15*	-
9.	Violent Behavior at Time 3A	−0.01	−0.07	0.02	0.05	0.16**	0.23***	−0.26***	0.30***

^ABivariate correlations involving these variables are Spearman’s correlations

^BRace = Hispanic/Latino (1) vs non-Hispanic (0)

^CAdjudicating Offense= Adjudicated for offense against a person (1) vs not (0)

^*=p<.05,

^**=p<.01,

^***=p<.001

Table 3.

Results from Logistic Regressions Predicting Violent Behavior.

	Model 1				Model 2				Model 3
Variables	ORA	SE	95% CI		ORA	SE	95% CI		ORA	SE	95% CI
Age	0.87	0.2 1	0.5 4	1.40	0.85	0.2 2	0.5 2	1.40	0.82	0.2 2	0.4 9	1.38
RaceB	0.44*	0.1 8	0.2 0	0.99	0.57	0.2 4	0.2 5	1.31	0.67	0.2 9	0.2 8	1.57
Adjudicatin g OffenseC	1.22	0.6 0	0.4 6	3.21	1.13	0.5 6	0.4 2	3.00	1.09	0.5 6	0.4 0	2.97
Time in Facility	1.39*	0.2 1	1.0 4	1.87	1.58**	0.2 6	1.1 5	2.17	1.60**	0.2 6	1.1 6	2.21
Exposure to Violence Prior to Incarceration	1.33*	0.1 7	1.0 3	1.71	1.18	0.1 6	0.9 0	1.55	1.18	0.1 7	0.8 8	1.57
Violent Behavior at Time 2	7.97***	3.4 0	3.4 5	18.3 7	6.69***	2.8 9	2.8 6	15.6 2	5.72***	2.5 8	2.3 6	13.8 4
Exposure to Violence at Time 2					1.90**	0.4 7	1.1 8	3.08	1.67*	0.4 2	1.0 2	2.73
Perceptions of Staff at Time 2									0.40**	0.1 4	0.2 1	0.79
R2 pseudo	16.45				20.11				23.48

^AOR = Odds Ratio.

^BRace = Hispanic/Latino (1) vs non-Hispanic (0)

^CAdjudicating Offense= Adjudicated for offense against a person (1) vs not (0)

^*=p<.05,

^**=p<.01,

^***=p<.001

Exposure to violence within the facility was added to the second model. The results indicated that exposure to violence within the facility was strongly associated with subsequently engaging in violence. Critically, once exposure to violence within the facility was included in the model, the association between prior exposure to violence and violent behavior failed to reach significance. That is, the results indicated that as compared to their prior life experience, the institutionalized context in which youth become embedded appears to be more impactful for determining their violent behavior.

In the final model, perceptions of staff were included to examine whether they reduced the association between exposure to violence within the facility and violent behavior. The results indicated that perceptions of staff were strongly associated with the likelihood of engaging in violence. As expected, positive perceptions of staff were associated with a reduction in the association between exposure to violence within the facility. Specifically, the odds of engaging in violence within the facility reduced by approximately 14%. In other words, a one-unit improvement in youths’ perceptions of staff was associated with a 14% decrease in the odds of engaging in violence. This suggests that above and beyond the effects of witnessing violence in the facility, the more adolescents trust staff and view them as protective, the less likely they are to engage in violence while institutionalized.

Discussion

It has been suggested for decades that exposure to violence is linked with violence perpetration. The findings of the present study are consistent with previous literature, but make a critical contribution by expanding the scope to incarcerated youth. The results demonstrate that adolescents who had been exposed to violence prior to incarceration are more likely to engage in violence within juvenile justice facilities. However, critically, once we take into account violence exposure that occurs while youth are detained, this association disappears. Experiencing violence while incarcerated emerges as a comparatively stronger indicator of subsequent violent behavior, above and beyond violence exposure that youth bring with them into juvenile justice facilities. Institutional violence, then, appears to have an especially significant and detrimental impact on adolescent violence.

On the other hand, and perhaps most importantly, we found that juveniles who viewed staff as legitimate and protective were less likely to engage in violence. In other words, even when youth were being exposed to violence within the institution, the presence of a capable and reliable adult appeared to help youth avoid engaging in violent behavior. This echoes similar findings that staff behavior can impact how youth respond to conflict (Marsh & Evans, 2009). Consequently, institutional staff emerge as an important buffer for the effects of violence exposure within an incarceration setting.

There is a large body of evidence that contends exposure to violence leads to violent behavior, and that engaging in violence increases the chances that youth will become incarcerated (Fehon, Grilo, & Lipschitz, 2005; Fagan, 2005). This study expands upon previous research by investigating how the cycle of violence continues when youth do become involved with the juvenile justice system. Moreover, it calls attention to the seriousness of violence behind bars. Whether or not youth have been immersed in a violent environment before entering a facility, the violence that they witness while institutionalized appears to be related to their likelihood of engaging in violent behavior. This is in line with other research on juvenile incarceration, which suggests that institutionalizing youth in general increases the chances of various undesirable outcomes, including higher levels of antisocial behavior (Gatti, Tremblay, & Vitaro, 2009).

Additionally, our findings regarding adolescents’ perceptions of staff align with Anderson’s (1999) code of the street framework, in that positive perceptions of law enforcement appear to “flip” the values that the code espouses. In the traditional literature, youth who perceive law enforcement negatively generally engage in violence because they do not trust the system to protect them. This study suggests, correspondingly, that incarcerated youth who perceive staff positively may actually turn to them for help when they feel threatened. If youth perceive staff as procedurally just and protective, it could be that they, in turn, feel safer overall. Because of this, they may no longer feel compelled to use violent behavior to earn respect or ward off victimization, as expected by the code of the street. If adherence to the code of the street is indeed plastic during adolescence as Moule and colleagues (2015) suggest, positive perceptions of correctional staff may be paramount in preventing or desisting from violent behavior while incarcerated. Nonetheless, because we were unable to directly test these mechanisms, this explanation remains tentative until future research is able to parse them more thoroughly.

The study had several noteworthy strengths. Primarily, as compared to cross-sectional studies, the longitudinal design provides a robust platform for evaluating the impact of institutional violence exposure on violent behavior. Because youth were sampled immediately following their initial introduction into the facility, we were able to consider their previous violence exposure and prior violent criminal activity, which clearly account for a significant amount of their subsequent institutional violence engagement. Further, considering the youth were sampled multiple times throughout their first few weeks of incarceration, we were able to examine the effects of institutional violence on subsequent violent behavior, while accounting for prior violence exposure and perpetration. As compared with cross-sectional studies, these strengths provide a more nuanced and robust analysis of what affects youths’ violent behavior while institutionalized as time unfolds on a weekly basis. Additionally, our study shows the significant impact of institutional violence exposure and positive perceptions of staff after only a few weeks behind bars. A relatively understudied period of time within juvenile justice populations, our data uniquely examine adolescents’ first few weeks of incarceration, a critical period in which youth are forced to adapt to a highly stressful environment.

This study provides significant insight into adolescents’ experiences with violence while incarcerated, but it is important to note that the data were collected from a single juvenile justice facility in California. Consequently, the results may not represent the experiences of institutionalized youth elsewhere in the state or country. Similarly, though the racial and ethnic breakdown of participants is representative of juvenile detainees in Southern California, it may not be generalizable to populations of institutionalized youth in other parts of the country. For instance, youth of Hispanic/Latino descent comprised the majority of our sample. Despite the fact that they represent the fastest growing demographic group in the United States, Hispanic/Latino youth may represent a smaller proportion of juvenile delinquents in other facilities and therefore behave differently as a function of their minority status within the institution. Indeed, this could also affect their perceived feelings of safety or reliance on staff. Further, the sample was entirely male. The results may not generalize to female juvenile inmates. Future research could address these limitations by sampling multiple juvenile justice facilities with demographic characteristics that represent local juvenile detention rates. Finally, these results may not generalize to adult justice populations, as adults may be less likely to seek out and rely on authority figures to guide their behavior. This may be particularly true of adults who have had a history of negative interactions with justice system personnel.

It is likely that there are youth whose offending behaviors are not detected by the system and therefore do not result in official records. Studies suggest that self-reported offending and official records are associated even among serious youthful offenders (Piquero, Schubert, & Brame, 2014). However, it is possible that self-report may be a better indicator of offending than official records among this type of sample considering some evidence that abused and neglected individuals self-report proportionately more offenses not known to officials as compared to those who have not been victimized (Maxfield, Weiler, & Widom, 2000). Nonetheless, considering measures based on multiple data sources are likely to be more valid (Farrington et al., 1996; Sullivan & McGloin, 2014), future research should utilize a triangulated measurement strategy.

While it is without question that youthful offenders should be held accountable for their crimes, these findings have important implications for the treatment of incarcerated adolescents. The results first suggest the need for policies aimed at preventing damaging violence exposure within juvenile justice facilities. While prior exposure to violence may be a risk factor for youth who enter institutions marked by little to no violence, it appears as though experiencing violence upon institutionalization can be even more impactful. The results indicated quite clearly that regardless of whether or not youth had been exposed to violence prior to incarceration, youth who were exposed to more violence upon institutionalization were significantly more likely to engage in violence themselves.

Furthermore, our findings elucidate the crucial role that correctional staff play in youth’s perceived feelings of protection and procedural justice and their likelihood of engaging in violence. When youth felt as though staff dealt fairly with all kids, they were less likely to commit violence in subsequent weeks. It could be that when juveniles are able to expect consistent fair treatment toward themselves as well as their peers, they do not feel compelled to solve their problems through violent means. Predictable behavior from staff may impart feelings of protection and procedural fairness, and in turn, youth may not feel it is necessary to take justice “into their own hands.” Correctional officers should thus aim to treat all delinquents with the same level of respect and use appropriate, consistent disciplinary action. Relatedly, it is possible that youth were able to avoid fights with fellow inmates when they felt like staff would help them navigate interpersonal issues. If staff receive regular training in nonviolent conflict resolution strategies that reinforce interpersonal communication skills and de-escalation tactics, they will certainly be better equipped to mentor delinquent youth who inevitably look to them for support. It is impractical to assume that all institutional violence will be eliminated, yet this study’s findings suggest that it is imperative that staff understand the power they wield in helping youthful offenders avoid violent conflict while institutionalized. When staff act as protective and procedurally just role models who treat adolescents fairly, their behavior can promote greater feelings of security and reduce violence among the youth in their care.