Institutional Infractions Among Young Prisoners

Abstract

Using a longitudinal design, this study describes changes in institutional infractions among a sample of 75 young males in a Portuguese prison. The inmates were assessed at 1, 3, 6, and 12 months after entry. The total number of infractions peaked at the third month and then declined. Although the pattern of severe infractions was irregular, minor infractions increased until the sixth month and decreased thereafter. Major predictors of inmates’ infractions during the first year in the institution were fewer visits, being single and non-White, having higher hostility levels, younger age at first imprisonment, and being a property offender.

Keywords

young prisoners institutional infractions longitudinal study

Introduction

The research literature identifies young age as one of the strongest predictors of prison infractions (Gendreau, Goggin, & Law, 1997; Gonçalves, Gonçalves, Martins, & Dirkzwager, 2014; Schenk & Fremouw, 2012). Prison infractions have in turn been associated with a higher chance of recidivism on release (Cochran, Mears, Bales, & Stewart, 2014; Trulson, DeLisi, & Marquart, 2011), making young prisoners a group with increased risk and needs. Despite this evidence, predictors of young prisoners’ infractions have been largely understudied compared with those in adults (Tasca, Griffin, & Rodriguez, 2010), and even less is known about changes in their behavior during incarceration (Boessen & Cauffman, 2013).

Trying to fill this gap in knowledge, this study examines (a) changes in young prisoners’ infractions over time, and (b) to what extent social support, mental problems, institutional risk, and other covariates are associated with this process. More knowledge on the development of and risk factors for inmate misconduct is important for prison management because it may help to optimize the classification of young prisoners, thus improving prison safety and inmate rehabilitation (Fernandez & Neiman, 1998; Lovell & Jemelka, 1996).

Inmate Adjustment to Prison

Adjustment to prison has been defined in several ways, but prisoners’ infractions have received the most attention as disciplinary problems have a huge impact on the order, safety, and management of correctional facilities (Trulson, 2007; Wright, 1985).

Penologists have developed a range of theoretical explanations for inmates’ adjustment to prison. Early theories explained inmates’ adjustment as a response to the features of the prison situation (deprivation theory; Sykes, 1958)—such as prison population size, security level, gang activity, availability of programs, and type of supervision (Gadon, Johnstone, & Cooke, 2006; Gendreau et al., 1997; Gonçalves et al., 2014). In response to the deprivation model, the importation model was developed (Irwin & Cressey, 1962). It proposes that inmates’ adjustment to prison life is related to pre-existing or imported personal characteristics such as their age, educational level, racial identification, criminal history, substance abuse, and mental problems (Gendreau et al., 1997; Gonçalves et al., 2014; Schenk & Fremouw, 2012).

More recently, scholars have emphasized the interaction between inmates and their environment (Wright, 1985). For instance, research confirmed that young inmates are more likely to misbehave in more crowded prison institutions (interactionist theories; Wooldredge, Griffin, & Pratt, 2001). Moreover, the well-known study of Zamble and Porporino (1988) showed that young inmates may have more reduced coping resources to deal with prison life, which can result in more adjustment problems (coping theories).

Empirical research has supported other criminological theories as well, showing that prisoners, including the youngster, with lower self-control (general theory of crime), fewer bonds to conventional society (social-control theory), and a deviant family background and early delinquent career (lifestyle-course theories) are more likely to misbehave (DeLisi, Beaver, et al., 2010; DeLisi, Trulson, Marquart, Drury, & Kosloski, 2011; Steiner & Wooldredge, 2009).

Although rarely used in the correctional context, social support theories may also be a relevant framework to explain inmates’ adjustment to prison. For instance, it has been suggested that people who benefit from more social support, especially from law-abiding others, are less likely to be involved in criminal behavior (Cullen, 1994). Applied to the prison context, this implies that inmates who receive more social support during their time in prison—for instance, in the form of visits from their family and friends—may be less likely to misbehave (Cochran, 2012).

Patterns of Institutional Infractions

In general, the early period of incarceration seems most stressful, which is reflected in higher levels of disruptive behaviors such as violence, self-injuries, and suicide attempts (Harvey, 2007; Kuanliang, Sorensen, & Cunningham, 2008; Liebling, 1999). Therefore, the initial phase of incarceration is a crucial period to study young inmates’ adjustment to life in prison (Cesaroni & Peterson-Badali, 2010; Monahan, Goldweber, & Cauffman, 2011).

Unfortunately, few studies have actually examined adjustment patterns of prisoners’ infractions over time, particularly among young offenders. Among adults, Zamble and Porporino (1988) observed that disciplinary infractions are higher in the beginning of the prison term and decline in the following months and years (see also Zamble, 1992). Toch and Adams (2002) showed that infractions were highest between the sixth and ninth months and decrease thereafter. More recently, using a group-based trajectory analysis, Cochran (2012) observed that infraction rates were higher between the sixth and seventh months, except for a large group of inmates (69%) that did not infract over the first year in prison.

Even less is known about the longitudinal course of adjustment to prison life among young prisoners. The scant research among this population has used either long (i.e., years) or short (i.e., weeks) follow-ups that fail to cover inmates’ behavior during the first year in prison, when major changes are expected to happen. Kuanliang et al. (2008) showed that the peak for violent misconduct among young prisoners was typically a year after entrance into the institution. Similarly, McShane and Williams (1989) reported that the number of major disciplinary infractions of juveniles was higher during the second year in prison. In addition, Boessen and Cauffman (2013) observed that whereas violent offending remains relatively stable during the first 8 weeks in prison, non-violent offenses continually increase during the first 6 weeks. Besides being scarce, some of the previous studies are retrospective, and all were conducted with North American samples.

Social Support, Mental Problems, and Institutional Risk

We are not aware of any study exploring the relationship between social support and misconduct among young prisoners to date. However, among adults, the results are inconclusive (Day, Brauer, & Butler, 2015). Several authors reported that social support mechanisms reduced prison misconduct (Cochran, 2012; Graeve, DeLisi, & Hochstetler, 2007; Jiang & Winfree, 2006), but others found no significant effects (Clark, 2001; Coid et al., 2003; Lahm, 2008).

Mental health may be another factor associated with inmates’ infractions. The literature seems to demonstrate that prisoners with mental problems, both young and adult, are more likely to misbehave (DeLisi, Caudill, et al., 2010; Lovell & Jemelka, 1996; Toch & Adams, 2002; Trulson, 2007). Such evidence is however inconclusive among the youngster. Some authors reported that young prisoners with more mental problems incur in fewer infractions (McReynolds & Wasserman, 2008), whereas others found no association between these variables (Van der Laan & Eichelsheim, 2013). In the present study, the Brief Symptom Inventory (BSI; Derogatis, 1993) is used to assess young inmates’ mental health problems. To our knowledge, no study has used the BSI to predict inmate misconduct so far.

Finally, an inmate’s institutional risk has been related to his or her prison infractions (Gonçalves et al., 2014; Lee & Edens, 2005; Shermer, Bierie, & Stock, 2013; Trulson, 2007). The Level of Service Inventory–Revised (LSI-R; Andrews & Bonta, 1995) is among the most widely used institutional risk tools in the correctional setting, and is used in the present study as a predictor of young inmates’ infraction rates. Although the LSI-R has consistently been related to prison infractions in adults (Campbell, French, & Gendreau, 2009; Gendreau et al., 1997), studies examining the LSI-R among young adults (i.e., those aged 16 until early adulthood) are lacking.

Other Covariates

Several other covariates have been associated to prison adjustment. At the personal level, predictors can be grouped into socio-demographic, criminological, and clinical variables. Although limited in number, some studies examined predictors of young prisoners’ infractions. Nevertheless, predictors found in adult inmates may apply to the juveniles as well.

Among socio-demographics, younger age, lower education level, and racial identification have been linked to young prisoners’ infractions (DeLisi, Caudill, et al., 2010; DeLisi et al., 2011; McReynolds & Wasserman, 2008; Trulson, 2007; Trulson, DeLisi, Caudill, Belshaw, & Marquart, 2010). Additional risk factors found among adult prisoners include being a foreigner and being single (Jiang & Fisher-Giorlando, 2002; Shermer et al., 2013). Clinical variables have been less explored, but substance abuse and prior mental treatment were related to prison misconduct, both in young and adult prisoners (DeLisi et al., 2011; Steiner & Wooldredge, 2009; Trulson et al., 2010). Criminological predictors such as younger age when first imprisoned, longer time served in prison, and the type of crime also appear to predict misconduct in young prisoners (Boessen & Cauffman, 2013; Trulson, 2007; Van der Laan & Eichelsheim, 2013). Additional risk factors in adults include being sentenced and prior prison infractions (Coid et al., 2003; Drury & DeLisi, 2010; Lee & Edens, 2005).

This Study

In sum, juveniles’ adjustment to prison life is a relatively new issue (Kuanliang et al., 2008) with several important questions to be answered. The present study will contribute to current knowledge in several ways. First, it uses a longitudinal design, which will result in knowledge on the temporal patterns of young prisoners’ adjustment. Second, it will enhance theory building on predictors of young prisoners’ adjustment. For instance, the social-support theory will be tested by examining the effect of visits on inmate infractions. Third, we were able to test the validity of the LSI-R and BSI in predicting young prisoners’ infractions. Those tools are easy and quick to administer and thus may be valuable for correctional practice. Finally, this research will provide a transcultural perspective on adjustment to prison, examining inmates of an unexplored cultural context (Portugal).

Imprisonment in Portugal

In Portugal, criminal responsibility starts at the age of 16.¹ Based on a rehabilitative philosophy, the law comprises protective measures for young offenders aged 16 to 21 both in the criminal code (e.g., reduced prison sentences) and in the execution of criminal sanctions (e.g., mandatory rehabilitation plan), but conditions of imprisonment are basically the same as for adults. Separation between young and adult prisoners is required, but contact frequently occurs because most prisons hold both adult and young inmates. There is only one prison in Portugal that only holds young prisoners. This is the research site of this study.

This correctional facility is located in a costal middle country city, and includes five operational units that house remand and sentenced offenders (separated). The correctional facility also has a drug-free unit and a well-equipped health care unit. Its architectural design is a campus design (see Morris & Worrall, 2014), and its actual capacity is 214 places. The mean occupancy rate in 2011 was 210 inmates (98%). The prison is classified as high security,² and most inmates come from urban regions (General Directorate of Reintegration and Prisons [DGRSP], n.d.).

Most inmates in this correctional facility are held in individual cells. Newcomers are assessed by educational and clinical staff within 72 hr and sent to the “Observation” block for an initial evaluation that guides the development of their rehabilitation plan. During this period, which lasts around 60 days, few prisoners are engaged in activities and most spend around 20 hr a day inside their cell. Progressively, they are enrolled in work, school, and other activities, and moved to other units. Those in more advanced stages of the sentence and with good institutional behavior may qualify for the open regime and may be moved to a block allowing more freedom and autonomy, as a preparation for their release.

Prisoners in the common regime can have visits two times per week in sessions of one hr with a maximum of three visitors. Visits are prevented or limited when the inmates are under isolation sanctions. In 2011, there were a total of 353 disciplinary infractions, of which 121 were considered severe and 232 minor.³ The most applied sanctions were written reprimands (18%), prohibition of using objects (e.g., videogames; 50%), isolation in the own cell (18%), and isolation in disciplinary cell (10%; DGRSP, n.d.).⁴

Method

Participants

The sample is composed of 75 males aged 17 to 22 years (M = 19.15, SD = 1.40) at the time of admission to the institution. Their racial identification was mixed—44% Whites, 45% Blacks, and 11% Gypsies. Most participants were Portuguese citizens (59%), although a considerable portion comes from ex Portuguese colonies (29%)⁵ or other countries (12%, mostly Brazil). Their educational level ranges from one to 12 years of schooling, and the mean was under the nine years of mandatory school (M = 6.85, SD = 2.35). The majority of the inmates were single (84%) and had a drug use history (80%), but their consumption patterns include almost exclusively Hashish and Cannabis. None was assigned to drug treatment programs in prison. Also, 37% had a mental treatment history prior to their arrest. Only 40% (n = 30) were already sentenced at the time of admission (75% one year after). The crimes they were accused of were disproportionately property crimes (71%), whereas violent (16%) and drug-related ones (13%) represented the remaining offenses.⁶ Inmates’ age at first imprisonment ranged from 16 to 21 years (M = 18.43, SD = 1.48).

Measures

Dependent variables

Institutional infractions are operationalized as the number of rule violations for which the inmates were found guilty. Infractions were organized into (a) total infractions, (b) severe infractions, and (c) minor infractions. Severe infractions were considered those that resulted in segregation, either by placement in a disciplinary cell or inside the inmates’ own cell. Minor infractions are all others resulting in less severe sanctions.

Independent variables

The first independent variable of this study is time in prison. To evaluate the main effect of time on inmates’ infractions and differences across waves, we created a variable corresponding to the number of months since the inmate entered the current facility. Time was centered at the first month; thus, the intercept of the estimates represent the mean of the outcome at the first month in prison. The other independent variables can be classified into socio-demographic, clinical, and criminological.

Socio-demographic

Socio-demographic variables include age (continuous in years, mean-centered), education (continuous in years, mean-centered), marital status (0 = having a lasting relationship, that is, through marriage or living together before prison; 1 = being single), nationality (0 = foreigner; 1 = Portuguese), racial identification (0 = non-White, that is, Blacks and Gypsies; 1 = White). As a measure for social support, the number of visits that an inmate received during each observation period was used.⁷

Clinical

Clinical variables include drug abuse history (0 = no; 1 = yes) and mental treatment history (0 = no; 1 = yes). Additional variables were measured through the BSI.

The BSI is a self-reported inventory of mental problems for persons aged 13 and older that take about 10 min to complete and may support clinical decision making at intake (Derogatis, 1993). It includes 53 items related to nine subscales: Somatization (seven), Obsessive/Compulsive (six), Interpersonal Sensitivity (four), Depression (six), Anxiety (six), Hostility (five), Phobic Anxiety (five), Paranoid Ideation (five), and Psychoticism (five). The instrument also provides information on the overall level of psychological distress (i.e., the Global Severity Index—GSI). The BSI has been validated and showed good psychometric qualities (Derogatis, 1993). In our study, the Cronbach’s alpha was high (α = .94), and all but one subscale proved to be reliable, with alphas ranging from .60 (for Paranoid Ideation) to .88 (for Somatization).⁸ The BSI is validated for the Portuguese population (Canavarro, 1999).⁹

Criminological

Criminological predictors include age at first imprisonment (continuous in years, mean-centered), type of crime (dummy coded: drug, property, violent), penal status (0 = remand; 1 = sentenced), time previously served in correctional facilities (i.e., counting prior imprisonments or/and time served for the actual sentence in other facilities¹⁰; continuous in months, mean-centered), and prior prison infractions (count variable, mean-centered).

The LSI-R is a risk and needs survey for offenders aged 16 and older that includes a structured interview and is professionally administered, taking between 30 and 45 min to administer (Andrews & Bonta, 1995). It includes 54 items grouped into 10 subscales: Criminal History (10), Education/Employment (10), Financial (two), Family/Marital (four), Accommodation (three), Leisure/Recreation (two), Companions (five), Alcohol/Drug Problems (nine), Emotional/Personal (five), and Attitudes/Orientation (four). The total score is used to classify individuals according to their institutional risk (i.e., high, medium, or low). The LSI-R has been validated and exhibited good/adequate psychometric properties (Andrews & Bonta, 1995). However, in this study, the Cronbach’s alpha was modest (α = .57), and the scales including fewer items showed poor reliability. Therefore, scales with alpha below .50 were excluded. As so, we only make use of the total score plus the Criminal History (α = .69), Education/Employment (α = .68), Alcohol/Drug Problems (α = .76), and Emotional/Personal (α = .69) Scales. The LSI-R is translated to Portuguese, but there are no adapted cut scores yet.

BSI and LSI-R variables were treated as continuous and mean-centered for analyses. Data on the independent variables at the first month are presented in Table 1.

Table 1.

Predictor Variables at the First Month.

Variable	M/prop.	SD	Range
Socio-demographic
Age	19.15	1.40	17-22
Education	6.85	2.35	1-12
Marital status (single)	0.84	0.36	0-1
Nationality (Portuguese)	0.59	0.50	0-1
Racial identification (White)	0.44	0.50	0-1
Visits in prison	5.32	6.66	0-30
Clinical
Drug abuse history	0.80	0.40	0-1
Mental treatment history	0.37	0.49	0-1
BSI GSI (≥ 1.0)	1.03	0.63	0.06-2.53
Somatization (≥ 0.6)	0.75	0.80	0.00-3.43
Obsessive/Compulsive (≥ 1.6)	1.21	0.71	0.00-2.83
Interpersonal sensitivity (≥ 1.1)	1.03	0.82	0.00-3.25
Depression (≥ 1.2)	1.27	0.77	0.00-3.50
Anxiety (≥ 1.3)	0.92	0.78	0.00-2.83
Hostility (≥ 1.1)	0.97	0.78	0.00-4 .00
Phobic anxiety (≥ 0.6)	0.66	0.67	0.00-2.60
Paranoid ideation (≥ 1.3)	1.22	0.68	0.00-3.00
Psychoticism (≥ 0.9)	1.10	0.78	0.00-2.80
Criminological
Age at first prison	18.43	1.48	16-21
Crime
Drug-related	0.13	0.34	0-1
Property	0.71	0.46	0-1
Violent	0.16	0.37	0-1
Penal status (sentenced)	0.40	0.49	0-1
Prior prison infractions	1.24	2.51	0-13
Prior time served (months)	7.45	9.31	0-42
LSI-R total score	26.59	5.73	12-40
Criminal history	2.72	1.89	0-8
Education/employment	7.73	1.57	1-10
Alcohol/drug problems	3.84	2.15	0-7
Emotional/personal	1.89	1.56	0-5

Note. M = mean; prop. = proportion; SD = standard deviation; BSI = Brief Symptom Inventory; GSI = Global Severity Index; LSI-R = Level of Service Inventory–Revised. Only visits, age, and penal status vary over time in prison. Age and penal status can just vary one time (and one unit) across waves. Values in parenthesis after BSI scales indicate their cut score for the general population.

Procedure

All inmates who entered the institution between March 2011 and December 2011 were eligible to participate in the study. Only those not understanding the Portuguese language were excluded. During this period, each month, the first author met with the selected newcomers in small groups (two to five), generally in the cafeteria of their unit. The inmates were informed about the objectives of the study, the confidentiality of the data, and that participation was voluntary. All accepted to participate.

After signing an informed consent form, participants filled out the BSI and engaged in the LSI-R interview. For those who had reading problems and who were not able to fill out the BSI (e.g., foreigners or illiterates), the researcher read the questions and recorded their chosen answers. Although in small groups, the LSI-R interview was carried out individually in a more distanced part of the room, and only the inmates and the researcher were present during the assessment protocol. The writing pen was given to the inmates as a reward.

Data were collected at four time periods during the first year of inmates’ detention in the current facility: at the first month (n = 75), third month (n = 67), sixth month (n = 60), and 12th month (n = 55). This time frame was chosen because it is known that adjustment problems are more frequent in the beginning of the sentence and dissipate over time. Twenty inmates dropped out during the study because they were released or transferred to another facility. The potential effect of selection bias was explored, but attrition appears to be random.¹¹ The outcome data represent the number of events observed during each observation period, that is, one month for Wave 1, two months for Wave 2, three months for Wave 3, and six months for Wave 4. For the purpose of this study, data on the LSI-R and BSI were treated as invariant (i.e., the scores from Wave 1 were used over the remaining periods).

Information on socio-demographic and clinical characteristics was based on inmates’ self-reports during the interviews. Criminological variables and outcome data were gathered from institutional files. This information was retrieved from the Prison Information System in the correctional facility, an electronic database containing general information on the inmates, including their infraction record and number of visitations.

Analysis

Multilevel regression analyses in Stata (Version 13) were performed to explore predictors of inmates’ infractions and changes over time. Because the outcomes are count variables with a non-normal distribution, we used Negative Binomial (NB) regressions, which are best suited to account for overdispersion in the data (Hilbe, 2011).¹² As the same individuals were measured repeatedly across moments of time, multilevel random-effect (RE) analyses were made to accommodate this within-cluster dependence and to capture variation over time between individuals (Rabe-Hesketh & Skrondal, 2008).¹³ Also, as the waves are not equally spaced (and the opportunity to infract is higher in longer periods), an exposure variable was included in the fixed part of the models, with coefficient constrained to 1, thus controlling for different waves’ length.¹⁴

More specifically, to answer our first research question (differences in infraction along time), the multilevel analyses were performed with time in prison (categorically coded with the first month as the reference category) as the only predictor of each type of infractions. Outcomes’ means at different waves were then calculated based on predictions of the model (predicted means). When the omnibus Wald test revealed significant mean differences, and after inspecting the distribution of the data, user-defined orthogonal contrasts with Bonferroni’s correction were made to test mean differences between specific waves.

To answer our second research question (predictors of inmates’ infractions over time), the effect of different predictors on each outcome was initially explored through bivariate analyses (RE NB regressions). Significant predictors were then added to time in prison in multivariate models. Trends in the data were explored through polynomials of time (quadratic term only). LSI-R and BSI scales were included in a second step to test their incremental validity. Besides the linear effect of time, and due to our small sample size, only (marginally) significant predictors were kept in the final models, in a maximum number of five.

Before analyses, missing data on BSI items (4%) were manually imputed based on predicted probabilities of regression models (there were no information missing at random in other variables).¹⁵ While modeling the data, the multivariate models were controlled for specification error, multicollinearity, and influential observations.¹⁶ Robust standard errors were calculated in all estimations to deal with other minor statistical concerns, including overdispersion (Hilbe, 2011; Rabe-Hesketh & Skrondal, 2008).

Results

Changes Over Time

Figure 1 shows changes in institutional infractions over time in prison. The level of total infractions sharply increased from the first to the third months and decreased thereafter, although differences across waves were not significant. The longitudinal course of severe infractions was rather irregular, with increases and decreases across the waves. The Wald tests of equality of means revealed that there were marginally significant differences between measurements, χ²(3) = 6.23, p = .10. Severe infractions, measured at the third month, were higher than in other months, as confirmed by orthogonal contrasts (p = .016). Minor infractions increased until the sixth month and decreased thereafter, but no significant mean differences across waves were observed. The prevalence of minor infractions was lower than the prevalence of severe infractions at the third month only, and their mean values were rather similar at the 12th month (analyses available on request).

Figure 1.

Predicted infractions mean by type over time in prison. Expected counts are based on the fixed part of the model only.

Predictors

Bivariate analyses regressing infractions on the independent variables indicate that younger age, being single, non-White, having a lower number of visits, earlier onset imprisonment, an arrest history before the age of 16, being sentenced, accused of property crimes, having more prior prison infractions, and a longer time served in prison were (marginally) associated with higher rates of some form(s) of infractions. Drug offenders were less likely to infract. Regarding the BSI, the GSI Scale, interpersonal sensitivity, depression, anxiety, and psychoticism were negatively associated with minor infractions. This indicates that prisoners with more mental problems were less likely to misbehave. However, those with higher hostility levels were more likely to be involved in severe infractions. The LSI-R total score did not predict inmates’ infractions. A more developed criminal history and educational problems were associated with more misconduct. Emotional/personal problems on the LSI-R were related with fewer incidents (analyses available on request).

Predictors of each type of infractions were then combined into multivariate analyses, keeping only significant ones in the final models. The results are presented in Table 2.

Table 2.

Multivariate Models Predicting Disciplinary Infractions Over Time in Prison.

Predictor	Total	Severe	Minor
Predictor	IRR (SE)	IRR (SE)	IRR (SE)
Intercept	0.05 (0.02)***	0.04 (0.02)***	0.02 (0.01)***
Time	0.99 (0.03)	1.03 (0.03)	0.97 (0.03)
Visits	—	0.95 (0.02)*	—
Single	2.03 (0.75)†	—	4.53 (3.07)*
White	—	0.44 (0.16)*	—
Age at first imprisonment	0.84 (0.06)*	—	—
Prior prison infractions	—	—	1.06 (0.04)†
Property crime	1.77 (0.51)*	2.83 (1.17)*	—
Sentenced	—	—	1.72 (0.51)†
BSI hostility	—	1.73 (0.31)**	—
RE intercept variance^a	0.17	0.29	0.03
Wald χ²	16.45**	31.74***	12.43*

Note. IRR = incidence rate ratio (exp(b)); SE = robust standard error around IRR; RE = random-effect parameters; BSI = Brief Symptom Inventory; —not included in the model. Scores on BSI Hostility were mean-centered for analysis. Observations = 257, n = 75.

Significance levels based on likelihood-ratio tests comparing panel and pooled NB regression models.

†

p < .10. *p < .05. **p < .01. ***p < .001, two-tailed.

Time did not predict infractions during inmates’ first year in prison. Regarding socio-demographics, more visits were associated with fewer severe infractions, confirming the protective effect of social support. For each additional visit, an inmate’s severe infraction rate was expected to decrease by 5%, given a constant value for model’s covariates. Being White was also associated with less severe infractions (incidence rate ratio [IRR] = 0.44). In addition, inmates who were single had more minor and total infractions than those having a lasting relationship (IRR = 4.53, 2.03, respectively), although the effect on total infractions were marginally significant.

Among clinical variables, when added in a second step into the multivariate model, hostility was the strongest predictor of severe infractions (IRR = 1.73). Attending to criminological variables, property offenders were expected to commit 2.8 and 1.8 times more severe and total infractions (respectively) than inmates accused of other crimes. A younger age at first prison was related with more total infractions. For each additional year in age at onset imprisonment, an inmate total infractions rate was expected to decrease by 16%. In addition, prior prison infractions and being sentenced were risk factors for minor infractions (IRR = 1.06, 1.72, respectively), but these effects were only marginally significant.

Discussion

Adjustment to prison among young offenders is an understudied topic that deserves further attention from research and practice. The present study extends knowledge by testing social-support theory, different assessment tools, and several covariates of young prisoners’ infraction during the sentence, an important indicator of their institutional adjustment, and an important factor for inmate classification and safety in prisons.

Regarding changes in institutional infractions over time, the level of total infractions peaked at the third month, and subsequently began to decline. This peak seems to occur sooner than reported in previous studies on adults (see Cochran, 2012; Toch & Adams, 2002), which may evidence that young prisoners have a different adjustment pattern. The pattern of minor infractions is more in line with the adult offender literature. In the present study, minor infractions increased during the first 6 months of imprisonment, and declined thereafter. In contrast, the pattern of severe infractions was irregular and was highest when the young inmates were in prison for 3 months. A possible explanation for this pattern was provided by Palmer and Farmer (2002). They reported that being in custody for more than a month was related to self-reported victimizing behaviors among young prisoners, which they argue could be used as a coping strategy to establish status after they feel adjusted to the prison regime. Perhaps this may apply to our data as well.

When we look at the role of the major predictors explored in this study (i.e., social support, mental problems, and institutional risk), it turned out that visits were a protective factor for the most severe form of misconduct. This underscores the importance of having social support in prison and extends social-support theory to young prisoners. Cochran (2012) suggested several explanations for the protective effect of visits on infractions. First, visits may help inmates to cope with the strains of imprisonment, like social isolation, reducing stress. Second, visits may help maintaining social bonds with family and friends, and consequently, the informal social control provided by those networks. Third, visits may improve prisoners’ perceptions and attitudes toward the prison system (see Cochran, 2012). In addition, prisoners with fewer visits may have less material support (e.g., money, food, tobacco) and thus feel pushed to break the rules for getting such goods (Colvin, 2007).

Because visits have the potential to reduce severe infractions, prison systems could reduce inmates’ misconduct by taking measures to improve possibilities for visits. For instance, inmates under isolation sanctions could be allowed to receive some visits because they may be those who need it most. Also, the prison system may take simple cost-effective measures to improve inmates’ social support, such as more flexible visitations hours, developing policies to make visitation a more pleasant experience for visitors and inmates, and creating ways to develop inmates’ social networks throughout the sentence (see Cochran, 2012; Monahan et al., 2011).¹⁷ Furthermore, visits in prison have been related to reductions and delay in recidivism after release (Bales & Mears, 2008; Cochran, 2014), accentuating the beneficial influence of social support in the long run.

Prisoners with more mental problems were less likely to engage in misconduct, especially in minor infractions. Although contrary to the literature in general, this result is in line with a few prior studies that observed no or negative associations between mental problems and misconduct among young prisoners (McReynolds & Wasserman, 2008; Van der Laan & Eichelsheim, 2013). However, because no inmate in our sample was diagnosed with a severe mental disorder, our data do not exclude the possibility that more severe clinical cases would indeed be associated with more misconduct.

In line with prior studies (Butler, Loney, & Kistner, 2007; DeLisi, Caudill, et al., 2010; DeLisi, Drury, et al., 2010), the results for hostility indicate that thoughts, emotions, and behaviors typical of angry states (Canavarro, 1999) predict disruptive behaviors among young prisoners. This scale achieved incremental validity in the multivariate model for severe infractions, being the strongest predictor. Therefore, the BSI Hostility scale may be useful for risk classification. Inmates with higher levels of hostility could be identified by prison staff and enrolled in appropriate programing (e.g., anger management). The results also suggest that it may be informative to look at individual scales of screening tools rather than just at their total score, as different mental problems may have a different influence on inmate behavior.

Focusing on institutional risk, the LSI-R total score had modest reliability and did not predict infractions in our sample. Although designed for offenders aged 16 and older, the LSI-R emphasizes adult concerns and may not be so reliable for young prisoners (Shields & Simourd, 1991). Thus, the LSI-R should not be used to classify young prisoners into custody levels, at least in its actual format. More adapted instruments are necessary for classification purposes among this population. Criminal history and education/employment problems appear to be more relevant predictors but did not achieve incremental validity. Differently, the results for emotional and personal problems, which are related to mental problems, seem to corroborate that such mental health symptoms tend to predict less infractions among incarcerated youth.

Regarding other covariates, this study confirms the applicability of some predictors found among adults. Specifically, although most were single, young prisoners with a lasting relationship seem to be more compliant with prison rules. The results regarding type of crime and racial identification have been inconsistent. Yet, as we found here, recent literature reviews evidenced that property offenders tend to be more associated with prison misconduct than those convicted for other crime, as tend to be non-White inmates when compared with Whites (Gonçalves et al., 2014; Schenk & Fremouw, 2012). Also in line with prior reviews, the younger the inmates were when first sent to prison, the more likely they were to misbehave, evidencing the development of criminal trajectories in prison. It also appears that prior misbehavior tends to perpetuate over time in prison and that misconduct is exacerbated after young inmates get sentenced. However, both variables were only marginally significant predictors of minor infractions and, therefore, these results need further validation.

Limitations and Future Directions

A major limitation of this study is its small sample size, which affects statistical power and precision of the analyses. Nevertheless, power is increased by the repeated measures design, and simulation studies document that multilevel models with a sample size higher than 50 at Level 2 may produce accurate estimates (see Maas & Hox, 2005).

It is, however, difficult to interpret how the results can be generalized to other young prisoners in Portugal and other countries. Because most young prisoners in Portugal are imprisoned among adults, the environment may exercise a different influence on those inmates (Tasca et al., 2010). Also, the ethnic composition of our sample is mixed, contrary to other regions of the country, where most inmates are Caucasian. In addition, although this study constitutes 36% of the youths, most foreigners coming from Romania were excluded due to language barriers, being therefore underrepresented. In 2011, Romanians constituted 6% of the prison population. Thus, studies replicating our research among young prisoners assigned to either prisons for adults or prisons for young prisoners, or among young prisoners in other countries and correctional settings, are necessary to generalize our findings.

This study leaves several questions unanswered. For instance, more research on the effects of inmates’ mental problems on infractions is required to further explain our results. Also, more research is needed to disentangle the relationship between social support and young inmates’ behavior. For instance, visits are just a proxy for social support; other variables are also important to assess this construct (e.g., support from staff). Finally, more studies using psychometric and institutional risk tools should be done in the prison context to test their potential utility for purposes of classification among young prisoners.

Despite limitations, this study expands theoretical and empirical knowledge on the course of young prisoners’ behavior, having implications for correctional practice. Although most results found among adults seem to apply to juvenile inmates, others appear more specific of this population, warranting specialized methods of classification and treatment. More longitudinal research among young offenders is needed to deepen current knowledge on their adjustment to prison and optimize institutional resources aiming at their rehabilitation.

Footnotes

Acknowledgements

The authors thank the General Directorate of Reintegration and Prisons (DGRSP) for the study authorization, Daniel Gonçalves for helping in the preparation of the database, and Arjan Blokland from Netherlands Institute for the Study of Crime and Law Enforcement (NSCR) for his assistance in the methods of this study.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research is part of an individual PhD supported by the Portuguese Foundation for Science and Technology with co-financing of the European Social Fund, grant number SFRH/BD/66987/2009.

Notes

Author Biographies

Leonel C. Gonçalves, PhD, is a researcher at the Canton of Zurich, Office of Corrections, Department of Mental Health Services, Switzerland. His research interests include forensic psychology and criminology, with a special focus on adjustment to prison life, the risk assessment of offenders, and their rehabilitation.

Anja J. E. Dirkzwager, PhD, is a senior researcher at the Netherlands Institute for the Study of Crime and Law Enforcement. Her research interests include the physical and psychosocial well-being of prisoners and their family members, prisoners’ perceptions of their conditions of confinement, and the effects of imprisonment on the further life course.

Carla Martins, PhD, is an assistant professor at the School of Psychology, University of Minho, Portugal, where she currently assumes the role of vice president. Her research interests include quantitative methods in psychology, with a special focus on meta-analysis, and socio-cognitive development from infancy to school age.

Rui A. Gonçalves, PhD, is an associate professor at the School of Psychology, University of Minho, Portugal, where he also works as a forensic psychologist at the Counseling Unit of Forensic Psychology. His research interests include prison life issues and violent offenders, namely, psychopaths, sexual offenders, and spouse abusers.

Peter Van der Laan, PhD, is a senior researcher at the Netherlands Institute for the Study of Crime and Law Enforcement and professor at the Faculty of Law of the Vrije University, Amsterdam. His research interests include criminal justice in general and the impact of interventions in particular, with a special interest in issues regarding juveniles.

References

Andrews

D. A.

Bonta

(1995). The Level of Service Inventory–Revised. Toronto, Ontario, Canada: Multi-Health Systems.

Bales

W. D.

Mears

D. P.

(2008). Inmate social ties and the transition to society: Does visitation reduce recidivism? Journal of Research in Crime & Delinquency, 45, 287-321.

Boessen

Cauffman

(2013). Moving from the neighborhood to the cellblock: The impact of youth’s neighborhoods on prison misconduct. Crime & Delinquency, 62, 200-228.

Butler

M. A.

Loney

B. R.

Kistner

(2007). The Massachusetts Youth Screening Instrument as a predictor of institutional maladjustment in severe male juvenile offenders. Criminal Justice and Behavior, 34, 476-492.

Campbell

M. A.

French

Gendreau

(2009). The prediction of violence in adult offenders: A meta-analytic comparison of instruments and methods of assessment. Criminal Justice and Behavior, 36, 567-590.

Canavarro

M. C.

(1999). Brief Symptom Inventory: BSI (text in Portuguese). In Simões

M. R.

Gonçalves

Almeida

L. S.

(Eds.), Testes e Provas Psicológicas em Portugal (Vol. II, pp. 87-109). Braga, Portugal: SHO/APPORT.

Cesaroni

Peterson-Badali

(2010). Understanding the adjustment of incarcerated young offenders: A Canadian example. Youth Justice, 10, 107-125.

Clark

T. A.

(2001). The relationship between inmate visitation and behavior: Implications for African American families. Journal of African American Men, 6, 43-58.

Cochran

J. C.

(2012). The ties that bind or the ties that break: Examining the relationship between visitation and prisoner misconduct. Journal of Criminal Justice, 40, 433-440.

10.

Cochran

J. C.

(2014). Breaches in the wall: Imprisonment, social support, and recidivism. Journal of Research in Crime & Delinquency, 51, 200-229.

11.

Cochran

J. C.

Mears

D. P.

Bales

W. D.

Stewart

E. A.

(2014). Does inmate behavior affect post-release offending? Investigating the misconduct-recidivism relationship among youth and adults. Justice Quarterly, 31, 1044-1073.

12.

Coid

Petruckevitch

Bebbington

Jenkins

Brugha

Lewis

. . . Singleton

(2003). Psychiatric morbidity in prisoners and solitary cellular confinement, I: Disciplinary segregation. Journal of Forensic Psychiatry & Psychology, 14, 298-319.

13.

Colvin

(2007). Applying differential coercion and social support theory to prison organizations: The case of the penitentiary of New Mexico. The Prison Journal, 87, 367-387.

14.

Cortina

J. M.

(1993). What is coefficient alpha? An examination of theory and applications. Journal of Applied Psychology, 78, 98-104.

15.

Cullen

F. T.

(1994). Social support as an organizing concept for criminology: Presidential address to the Academy of Criminal Justice Sciences. Justice Quarterly, 11, 527-559.

16.

Day

J. C.

Brauer

J. R.

Butler

H. D.

(2015). Coercion and social support behind bars: Testing an integrated theory of misconduct and resistance in US prisons. Criminal Justice and Behavior, 42, 133-155.

17.

DeLisi

Beaver

K. M.

Vaughn

M. G.

Trulson

C. R.

Kosloski

A. E.

Drury

A. J.

. . . Wright

J. P.

(2010). Personality, gender, and self-control theory revisited: Results from a sample of institutionalized juvenile delinquents. Applied Psychology in Criminal Justice, 6, 31-46.

18.

DeLisi

Caudill

J. W.

Trulson

C. R.

Marquart

J. W.

Vaughn

M. G.

Beaver

K. M.

(2010). Angry inmates are violent inmates: A Poisson regression approach to youthful offenders. Journal of Forensic Psychology Practice, 10, 419-439.

19.

DeLisi

Drury

A. J.

Kosloski

A. E.

Caudill

J. W.

Conis

P. J.

Anderson

C. A.

. . . Beaver

K. M.

(2010). The cycle of violence behind bars: Traumatization and institutional misconduct among juvenile delinquents in confinement. Youth Violence and Juvenile Justice, 8, 107-121.

20.

DeLisi

Trulson

C. R.

Marquart

J. W.

Drury

A. J.

Kosloski

A. E.

(2011). Inside the prison black box: Toward a life course importation model of inmate behavior. International Journal of Offender Therapy and Comparative Criminology, 55, 1186-1207.

21.

Derogatis

L. R.

(1993). Brief Symptoms Inventory (BSI): Administration, scoring and procedures manual. Minneapolis, MN: NCS Pearson.

22.

General Directorate of Reintegration and Prisons (DGRSP) (n.d.). Annual Report 2011 (text in Portuguese) (Unpublished report). Available at http://www.dgsp.mj.pt/

23.

Drury

A. J.

DeLisi

(2010). The past is prologue: Prior adjustment to prison and institutional misconduct. The Prison Journal, 90, 331-352.

24.

Fernandez

K. E.

Neiman

(1998). California’s inmate classification system: Predicting inmate misconduct. The Prison Journal, 78, 406-422.

25.

Gadon

Johnstone

Cooke

(2006). Situational variables and institutional violence: A systematic review of the literature. Clinical Psychology Review, 26, 515-534.

26.

General Directorate of Reintegration and Prisons (DGRSP) (n.d.). Annual report 2011 (Unpublished report). Retrieved from http://www.dgsp.mj.pt/

27.

Gendreau

Goggin

C. E.

Law

M. A.

(1997). Predicting prison misconducts. Criminal Justice and Behavior, 24, 414-431.

28.

Gonçalves

L. C.

Gonçalves

R. A.

Martins

Dirkzwager

A. J. E.

(2014). Predicting infractions and health care utilization in prison: A meta-analysis. Criminal Justice and Behavior, 41, 921-942.

29.

Graeve

C. M.

DeLisi

Hochstetler

(2007). Prison rioters: Exploring infraction characteristics, risk factors, social correlates, and criminal careers. Psychological Reports, 100, 407-419.

30.

Harvey

(2007). Young men in prison: Surviving and adapting to life inside. Cullompton, UK: Willan.

31.

Heckman

J. J.

(1979). Sample selection bias as a specification error. Econometrica, 47, 153-161.

32.

Hilbe

J. M.

(2011). Negative binomial regression (2nd ed.). Cambridge, UK: Cambridge University Press.

33.

Irwin

Cressey

(1962). Thieves, convicts, and the inmate culture. Social Problems, 10, 142-155.

34.

Jiang

Fisher-Giorlando

(2002). Inmate misconduct: A test of the deprivation, importation, and situational models. The Prison Journal, 82, 335-358.

35.

Jiang

Winfree

L. T.

(2006). Social support, gender, and inmate adjustment to prison: Life insights from a national sample. The Prison Journal, 86, 32-55.

36.

Kuanliang

Sorensen

J. R.

Cunningham

M. D.

(2008). Juvenile inmates in an adult prison system: Rates of disciplinary misconduct and violence. Criminal Justice and Behavior, 35, 1186-1201.

37.

Lahm

K. F.

(2008). Inmate-on-inmate assault: A multilevel examination of prison violence. Criminal Justice and Behavior, 35, 120-137.

38.

Lee

S. J.

Edens

J. F.

(2005). Exploring predictors of institutional misbehavior among male Korean inmates. Criminal Justice and Behavior, 32, 412-432.

39.

Liebling

(1999). Prison suicide and prisoner coping. Crime and Justice, 26, 283-359.

40.

Long

J. S.

Freese

(2005). Regression models for categorical outcomes using Stata (2nd ed.). College Station, TX: Stata Press.

41.

Lovell

Jemelka

(1996). When inmates misbehave: The costs of discipline. The Prison Journal, 76, 165-179.

42.

Maas

C. J. M.

Hox

J. J.

(2005). Sufficient sample sizes for multilevel modeling. Methodology: European Journal of Research Methods for the Behavioral and Social Sciences, 1, 86-92.

43.

McShane

M. D.

Williams

F. P.

(1989). The prison adjustment of juvenile offenders. Crime & Delinquency, 35, 254-269.

44.

McReynolds

L. S.

Wasserman

G. A.

(2008). Risk for disciplinary infractions among incarcerated male youths: Influence of psychiatric disorder. Criminal Justice and Behavior, 35, 1174-1185.

45.

Monahan

K. C.

Goldweber

Cauffman

(2011). The effects of visitation on incarcerated juvenile offenders: How contact with the outside impacts adjustment on the inside. Law and Human Behavior, 35, 143-151.

46.

Morris

R. G.

Worrall

J. L.

(2014). Prison architecture and inmate misconduct: A multilevel assessment. Crime & Delinquency, 60, 1083-1109. doi:10.1177/0011128710386204

47.

Palmer

E. J.

Farmer

(2002). Victimizing behaviour among juvenile and young offenders: How different are perpetrators? Journal of Adolescence, 25, 469-481.

48.

Rabe-Hesketh

Skrondal

(2008). Multilevel and longitudinal modeling using Stata (2nd ed.). College Station, TX: Stata Press.

49.

Schenk

A. M.

Fremouw

W. J.

(2012). Individual characteristics related to prison violence: A critical review of the literature. Aggression and Violent Behavior, 17, 430-442.

50.

Shermer

L. O.

Bierie

D. M.

Stock

(2013). Endogeneity in prison risk classification. International Journal of Offender Therapy and Comparative Criminology, 53, 1248-1274.

51.

Shields

I. W.

Simourd

D. J.

(1991). Predicting predatory behavior in a population of incarcerated young offenders. Criminal Justice and Behavior, 18, 180-194.

52.

Steiner

Wooldredge

(2009). The relevance of inmate race/ethnicity versus population composition for understanding prison rule violations. Punishment & Society, 11, 459-489.

53.

Sykes

(1958). The society of captives. Princeton, NJ: Princeton University Press.

54.

Tasca

Griffin

M. L.

Rodriguez

(2010). The effect of importation and deprivation factors on violent misconduct: An examination of Black and Latino youth in prison. Youth Violence and Juvenile Justice, 8, 234-249.

55.

Toch

Adams

(2002). Acting out: Maladaptive behavior in confinement. Washington, DC: American Psychological Association.

56.

Trulson

C. R.

(2007). Determinants of disruption: Institutional misconduct among state-committed delinquents. Youth Violence and Juvenile Justice, 5, 7-34.

57.

Trulson

C. R.

DeLisi

Caudill

J. W.

Belshaw

Marquart

J. W.

(2010). Delinquent careers behind bars. Criminal Justice Review, 35, 200-219.

58.

Trulson

C. R.

DeLisi

Marquart

J. W.

(2011). Institutional misconduct, delinquent background, and rearrest frequency among serious and violent delinquent offenders. Crime & Delinquency, 57, 709-731.

59.

Van der Laan

Eichelsheim

. (2013). Juvenile adaptation to imprisonment: Feelings of safety, autonomy and well-being, and behaviour in prison. European Journal of Criminology, 10, 424-443.

60.

Wooldredge

Griffin

Pratt

(2001). Considering hierarchical models for research on inmate behavior: Predicting misconduct with multilevel data. Justice Quarterly, 18, 203-231.

61.

Wright

K. N.

(1985). Improving correctional classification through a study of the placement of inmates in environmental settings. Washington, DC: U.S. Department of Justice.

62.

Zamble

(1992). Behavior and adaptation in long-term prison inmates: Descriptive longitudinal results. Criminal Justice and Behavior, 19, 409-425.

63.

Zamble

Porporino

F. J.

(1988). Coping, behavior, and adaption in prison inmates. New York, NY: Springer.