Abstract
Correctional literature on determinants of prisoner misconduct is largely focused on the situation in the USA or West European countries. This study expands the research in this field by presenting findings from Romania, an East European country whose prison system faces severe problems, among which overcrowding and poor confinement conditions are of the utmost concern. Therefore a survey was conducted on a sample of 280 adult male inmates in four large Romanian prisons. These exceptional data were supplemented with information from official records drawn from the prison administration’s databases. A unique combination of importation and deprivation factors, with a particular focus on several prison deprivations perceived by inmates as problematic, is examined in relation to the prevalence of four types of inmate misbehaviour: total misconduct, contraband infractions, violence and defiance. The findings show a clear relationship between importation and deprivation characteristics and prison misconduct. Also a differential impact of these characteristics is shown, depending on the type of prison infraction examined. The study concludes by advancing a set of policy recommendations to reduce the incidence of institutional misconduct in Romanian prisons.
Introduction
The correctional literature abounds in studies aiming at identifying the determinants of inmate misconduct. Guided by broader theoretical debates on whether behaviour in prison is ‘imported’ from the outside and/or ‘indigenously’ activated, these studies provide important theoretical and practical-oriented knowledge about individual and prison characteristics associated with misconduct (Cao et al., 1997; Jiang and Fisher-Giorlando, 2002; Steiner and Wooldredge, 2008).
The current literature, however, has a serious limitation: its focus is mostly restricted to the situation in the United States and its unique prison context (Cunningham and Sorensen, 2007; DeLisi, 2003; Innes, 1997; Steiner, 2009; Steiner and Wooldredge, 2008; Tewksbury et al., 2014). Moreover, the few studies on prison misconduct in Europe have almost all been conducted in West European countries (for example, Arbach-Lucioni et al., 2012; Beijersbergen et al., 2014). Thus, there is a regrettable paucity of knowledge on determinants of prison misconduct in East European countries.
The present study attempts to address this limitation by focusing on the specific case of Romania, a former Communist country that started to incorporate the notion of human rights, including prisoners’ rights, on its political and legal agendas only after 1990. Although important penal reforms have been adopted since then, the country’s prison system continues to be plagued by high imprisonment rates, severe overcrowding, under-staffing, small budgets and poor confinement conditions.
More precisely, the study assesses the importance of various ‘indigenous’ and ‘importation’ characteristics of prisoners for explaining variations in patterns of prison misconduct among a sample of male Romanian inmates. Thereby this study builds on the existing international literature, and it advances it in the specific Romanian context. First, it makes progress because the effects of a unique combination of importation and deprivation characteristics are examined. Some of these characteristics have been examined in prior studies (for example, inmate’s age, time served, prior incarcerations etc.), but others have only seldom been explored in existing literature (for example, involvement in working activities). Second, the study will take a deeper look into how certain prison deprivations, as perceived by inmates, correlate with misconduct, thus contributing to a better assessment of the deprivation model for prison misconduct. By focusing on prisoners’ subjective evaluations, the intensity with which particular deprivations are felt by the prisoners can be related to patterns of prison misconduct.
The current study contributes not only to the international literature but also to the Romanian correctional literature. So far no study has specifically linked ‘importation’ and ‘deprivation’ characteristics to misconduct in Romanian prisons. Some notable studies have been published describing prison conditions and inmates’ perceptions of these conditions in Romania (see, for example, ANP, 2006, 2010). Furthermore, a few studies have described inmates’ aggressive behaviour in Romanian prisons, where some surveyed the internal aggression of prisoners – that is, their self-harming and suicidal behaviour (ANP, 2013; Pripp, 2004; Țogoie et al., 2004) and/or inmate-to-inmate assaults and inmate-to-staff assaults (ANP, 2013). However, no prior study has discerned which importation and deprivation characteristics are actually correlated with the occurrence of misconduct in Romanian prisons. So, the current study will offer important new knowledge on determinants of misbehaviour in prisons, and aims to give insights that are helpful in implementing strategies for reducing misconduct in Romanian prisons.
Theoretical expectations and prior research
Scholarship on prisoner misconduct is inspired by two well-known theories of adaptation to imprisonment: the importation and the deprivation models. The importation theory argues that the prisoner characteristics are the main factor explaining misconduct in prison (Irwin and Cressey, 1962). In short, it assumes that misconduct is the result of the same characteristics that in general predispose individuals towards anti-social behaviour and crime (Berg and DeLisi, 2006; Innes, 1997). Typically, these include the socio-demographic characteristics of prisoners (for example, age, marital status, parenthood) and their prior offending characteristics (for example, conviction offence, violent offending past, prior incarceration). Accordingly, it is argued that older inmates who cohabit or are married and have children are less inclined to misbehave in prison, since in general they have lower tendencies towards crime (Jiang and Winfree, 2006; Steiner and Wooldredge, 2008; Tewksbury et al., 2014). Also, it is claimed that in prison, as in the outside world, recidivists and violent offenders are more susceptible to engaging in violent and disobedient behaviour (Arbach-Lucioni et al., 2012; DeLisi et al., 2004; Griffin and Hepburn, 2006).
The deprivation theory offers an alternative view. It suggests that prisoner misbehaviour is not determined by prisoner characteristics but rather by prison characteristics (Clemmer, [1940] 1966; Sykes, [1958] 2007). Its main propositions are rooted in the classic sociological works of Sykes ([1958] 2007) and Goffman (1961) and focus on correlates associated with prison deprivations. The model’s basic assumption is that the more deprived (physically, socially, materially) the prisoners’ experience of imprisonment is, the higher will be their psychological suffering. To alleviate these ‘pains of imprisonment’, prisoners team up and form their own social system based on a subcultural code favouring rule violation and prison misconduct (Sykes, [1958] 2007). Various prison aspects have been used as proxies for measuring the degree of prison deprivation that inmates are exposed to. For example, it is assumed that prisoners who spend more time in detention, who are sentenced to longer terms of imprisonment, and who are held under more restrictive regimes have prolonged experiences of deprivation and, thus, have higher probabilities of engaging in prison misbehaviour (Cao et al., 1997). On the other hand, it is expected that participants in prison work programmes have lower chances of misbehaving, since prison work has ameliorative effects on the deprivations of imprisonment (Gover et al., 2008). Along with these objective characteristics, inmates’ subjective evaluation of prison conditions, relationships with staff and perceived justice have also been put forward as influencing misbehaviour in prison. Although small in number, the studies that have incorporated these perceptual measures clearly documented their importance in relation to misconduct (Beijersbergen et al., 2014; Reisig and Mesko, 2009; Rocheleau, 2013).
Together, the importation and the deprivation models have been tested empirically several times (see, for example, Arbach-Lucioni et al., 2012; Cao et al., 1997; Dhami et al., 2007; Sorensen et al., 1998) in relation to various forms of misconduct. As such, they have been used to explain specific types of misconduct (for example, inmate-to-inmate assaults; inmate-to-staff assaults) or more generic constructs (for example, total misconduct, violence, contraband infractions). However, the findings of prior studies are mixed at best and, with a few exceptions (for example, age), no definite conclusions have been drawn so far regarding the various explanations of prison infractions. This stresses the need for more empirical research in this field. The current study takes a step in this direction by looking into the relationships between several factors of both the deprivation and importation models and inmates’ misconduct in the particular context of Romanian prisons.
Data, measures and method
The present study is part of a research project conducted with the broader purpose of investigating the effects of imprisonment on Romanian offenders’ lives (Dâmboeanu, 2012, 2014). Methodologically, it used a cross-sectional research design and employed a quantitative approach to data collection based on a survey as well as on administrative data.
Research sites
The study was conducted in four Romanian prisons for adult male offenders: Giurgiu Prison, Craiova Prison, Timisoara Prison and Tulcea Prison. All are among the 10 largest correctional facilities in Romania, each holding over 1000 incarcerated individuals, including both pre-trial detainees and convicted offenders. Timisoara Prison is officially classified as an open and semi-open prison and the other three are classified as maximum security and closed regime prisons. However, in practice there are also prisoners held in an open and semi-open regime, although at a lower rate. All four prisons are confronted with the problem of overcrowding: Craiova Prison had the highest occupancy level (183 percent), followed by Giurgiu Prison (127 percent), Tulcea Prison (123 percent), and Timisoara Prison (100 percent). None of these accommodate inmates in single cells. In fact, only Giurgiu Prison, a new and modern institution operating since 2004, houses the prisoners in cells of two, four and a maximum of -six beds. The other three have various accommodation types, including large dormitories, some equipped with over 30 beds. According to the findings of a study conducted by the Administrația Națională a Penitenciarelor (ANP, 2013), in 2012, the rate of inmate-to-inmate assaults was 8.4 per 100 prisoners in Giurgiu Prison, 4.4 in Craiova Prison, 3.3 in Timisoara Prison and 2.1 in Tulcea Prison. The differences in self-harm rates are even more pronounced: Giurgiu having a rate of 11.2 offences per 100 prisoners, Craiova of 6.5, Tulcea of 1.5 and Timisoara of 1.3, respectively.
Participants
In this study it was decided in advance that a total of 320 prisoners would be interviewed. Participants for the study were randomly selected from lists of all inmates with a final conviction incarcerated in the four above-mentioned prisons. The lists contained the names of the prisoners and extra information on them, and were organized according to the time already spent in detention: (a) long-serving prisoners, who had spent more than five years of their current sentence in prison; (b) medium-serving prisoners, who had served between two and five years of their term; and (c) short-serving prisoners, who had served less than two years of their sentence. Long-serving prisoners were over-represented in the samples of Giurgiu and Craiova. At Tulcea Prison this was the only category of prisoners selected (for more information on the sampling strategy, refer to Dâmboeanu, 2014).
From the 320 interviewed persons, 41 were excluded from this study: 15 people did not give permission for their electronic files in which information on their misconduct in prison was registered to be consulted; in 26 interviews, some incompatibilities were found between official data and prisoner questionnaires on important socio-demographics, so their answers were not reliable and they were dropped from the sample. The final sample comprises 280 inmates for whom both questionnaires and official data were available and consistent.
In the current study, we extracted and used two sub-samples from this original final sample: (a) prisoners incarcerated for at least one year at the time of the interview (N = 240); and (b) those incarcerated for at least two years at the time of the interview (N = 179) in order to assess the contribution of deprivation and importation factors to misconduct during the previous year and the previous two years of imprisonment, respectively (see also below). The number of cases for the two-year analyses is smaller, because 61 inmates who had been incarcerated for one year had not (yet) been incarcerated for a period of two years.
Misconduct in prison
Information on the inmates’ misconduct in prison was obtained from official electronic files of the prison administration’s database in each of the four prisons. Consequently, prison misconduct was defined as any violation of a prison rule officially sanctioned by the prison’s administration. It includes both major and minor rule infractions, but no offences to which the Criminal Code and other special penal laws apply and for which the inmates would receive a new criminal conviction.
Based on the information in the electronic files, four outcome variables were constructed for distinct types of misconduct: total misconduct, contraband infractions, violence, and defiance (see Siennick et al., 2013). Total misconduct indicates whether or not the inmate committed any sort of misconduct. Violent misconduct measures whether or not the inmate was involved in assaults on other inmates/staff, in self-mutilation, and/or was in possession of knives or other weapons. Contraband infractions indicate whether the inmate had mobile phones, drugs/alcohol, and/or other prohibited items in his possession. Defiant behaviour indicates whether the inmate violated the prison’s internal regulations and/or behaved disrespectfully towards prison staff.
Each of the outcome variables was constructed twice, that is, for two distinct periods of time: (a) the previous year; (b) the previous two years. The decision to limit the analyses to the previous (two) year(s) was made in order to avoid the potential biases that might have arisen from the use of measure(s) over the entire period the inmates have been incarcerated: that is, the detention period of the respondents differs substantially and the prisoners might have been in different units and even different prisons during their time in prison.
Table 1 shows that slightly over half (53 percent) of the 179 interviewed inmates incarcerated for at least two years had been involved in a misconduct incident during the previous two years. In total, 296 incidents of misconduct were registered for this group. The average frequency of misbehaviour was 1.7 incidents in the previous two years. The highest number of misconduct incidents accumulated by one prisoner was 16 in the previous two years. The two-year prevalence rates for specific types of misbehaviour were: 37 percent for contraband infractions, 22 percent for violence, and 22 percent for defiance. The inmates in Giurgiu Prison have the highest participation rates in total misconduct as well as in the particular types of misconduct.
Descriptive statistics of misconduct and importation and deprivation characteristics, by prison (N = 240).
The picture is very similar when the prevalence of misconduct during the previous year is analysed: 35 percent of the 240 inmates interviewed had a recorded misconduct, 21 percent for contraband violations, 13 percent for violent behaviour and 11 percent for defiant behaviour. Again, Giurgiu Prison had the highest rates of misconduct. The sub-sample of 240 inmates committed 171 infractions in total; the peak number of violations was 9, and the average was 0.7.
Deprivation characteristics
Based on the deprivation model, we hypothesized that inmates in more restricted prison regimes, inmates who have spent relatively longer periods in detention, and those given lengthy sentences are more likely to misbehave in prison. On the other hand, inmates involved in work activities are less likely to misbehave. Information on these four deprivation characteristics was also obtained from the prisons’ electronic databases. They have all been included in analyses as dichotomous variables: length of stay (5 years or more = 1; otherwise = 0); length of sentence (5 years or more = 1; otherwise = 0); prison regime measured at the time of the interview (maximum security and closed regime = 1; otherwise = 0); being involved in work activities (yes, if the inmate worked at least 20 days = 1; otherwise = 0). In addition, the prisons in which the study was conducted were recoded and incorporated as dummy variables in the logistic regression models (with Timisoara Prison as the reference).
Table 1 shows that more than one-third of those incarcerated for at least one year at the time of the interview had served five years or more of their current sentence. Almost three-quarters had been sentenced to five or more years of imprisonment. Sixty percent were held in closed and maximum security regimes, and just over half worked in prison for a minimum of 20 days. There are clear differences when comparing the four prisons: Timisoara Prison has inmates with the shortest prison sentences or stays and the highest rate of inmates working in prison.
Also based on the deprivation model, it can be expected that inmates’ subjective evaluation of prison conditions is relevant for their (disobedient) behaviour in prison. To measure subjective deprivation, several variables were constructed based on the inmates’ answers to 13 questions phrased as: ‘To what extent is [topic X] problematic for you in prison?’ The topics mentioned cover the following aspects of prison life: sharing a cell with other inmates, the quality of the food, a crowded cell, the hygiene conditions, the strictness of prison rules, the harsh attitude of prison staff, a lack of support from prison staff, spending too much time in the cell, the monotony of everyday life in prison, the violent behaviour of other inmates, a lack of privacy, a lack of access to work opportunities, and a lack of access to prison programmes. These prison aspects were selected for inclusion in the questionnaire based on two sources: (a) prior studies (see, for example, Liebling, 2004, Rocheleau, 2013); (b) the findings of a (qualitative) exploratory study also conducted in this postdoctoral research project. The respondents could indicate the extent to which they experienced these topics as problematic on a 1–5 scale (to a very small extent = 1; to a very great extent = 5); 13 dichotomous variables were created (4 or more = 1; 3 or less = 0).
Summary measures based on the 13 subjective deprivation indicators were also constructed. First, an overall scale of prison deprivation was created by averaging all 13 topics (Cronbach’s alpha = .86), and then dichotomizing this variable (4 or more = 1; 3 or less = 0). Second, in order to test the relationships between specific types of deprivation and specific types of misconduct, five particular subscales were constructed and then recoded into dichotomous variables: (a) physical conditions of confinement – based on the average scores of five items: sharing a cell with other inmates, a lack of privacy, the quality of the food, a crowded cell, and hygiene conditions (Cronbach’s alpha = .69); (b) treatment by staff – based on the average of three items: the strictness of prison rules, the harsh attitude of prison staff, a lack of support from prison staff (Cronbach’s alpha = .83); (c) prison routine – based on the average of two items: spending too much time in the cell, the monotony of everyday life in prison (Cronbach’s alpha = .76); (d) violence – based on a single item: the violent behaviour of other inmates; and (e) social valued activities – based on the average of two items: a lack of access to work opportunities and a lack of access to prison programmes (Cronbach’s alpha = .69).
Table 1 indicates that the prison deprivations most intensely experienced by inmates are those associated with crowded cells and the quality of the food, followed by the lack of work opportunities and the long periods of time they have to spend in their cell. Clearly, the inmates held at Giurgiu Prison were the most discontented; their scores on the overall subjective deprivation scale, as well as on all except one (the physical prison conditions) of the particular subscales, were high.
Importation characteristics
The general expectation is that socio-demographic variables are good indicators of the relevant importation characteristics and are related to misconduct. The socio-demographic variables included in this analysis are all based on official data files from the four prison administrations’ electronic databases. We included three socio-demographic variables, and coded these as dichotomous variables: age at interview (35 years old or more = 1; otherwise = 0); being married or cohabiting at incarceration (yes = 1; no = 0); having children at incarceration (yes = 1; no = 0).
Other relevant importation characteristics for misconduct are related to inmates’ criminal history. We therefore included three additional importation variables: prior incarceration, a measure representing the dichotomous coding of four response options (once/the current one = 1; 2–3 times = 2; 4–5 times = 3; more than 5 times = 4) to the question: ‘How many times have you served a prison sentence?’ (2 or higher = 1; 1 = 0); history of violent offences is a recoded dummy variable (yes, if at least one violent offence was committed = 1; no = 0) constructed from an 11-item scale with two response options (yes or no), measuring the prevalence of various types of offence committed by inmates over the life course; and currently imprisoned for a violent offence (yes, that is, for homicide, physical injury, robbery, and rape = 1; no = 0). The first two variables were based on self-reported data, while the last one was based on official data.
Data in Table 1 reveal that almost half of the participants were 35 years or over at the time of the interview; over half were married or cohabitating at the time of the interview; a similar proportion had children; 57 percent had been convicted of a violent offence, 50 percent had served one or more prior prison terms, and 41 percent reported having a history of violence. Noticeable differences existed between prisons: Timisoara Prison had the fewest inmates convicted of a violent crime and with prior incarcerations.
Method
The present study assesses the importance of various ‘deprivation’ and ‘importation’ characteristics for differences in prison misconduct among a sample of Romanian male inmates. To do this, first, descriptive analyses are carried out in order to examine the bivariate relationships between the inmates’ participation in total misconduct on the one hand and each importation and deprivation characteristic separately on the other hand. The results are presented in Table 2. These bivariate analyses are done for misconduct in the previous year and the previous two years, where the first analysis includes 240 inmates and the second analysis only 179 inmates, because 61 of the inmates had not (yet) been incarcerated for a period of two years.
Misconduct by prisoners in previous year and previous two years (prevalence rates), by importation and deprivation characteristics (yes/no).
p < .10; **p < .05; ***p < .01.
Second, logistic regression analyses were done where the prevalence of misconduct during both time periods was treated as the dependent variable and various measures of ‘deprivation’ and ‘importation’ characteristics as independent variables. Since the individual measures of the subjective deprivation characteristics were highly correlated, two models were performed for each dependent variable: Model A tests the effects of the overall subjective deprivation scale, while Model B assesses the effects of the five specific deprivation subscales. The results of the regression analyses are presented in Table 3.
Logistic regression results for importation and deprivation correlates of total misconduct during previous year/previous two years.
p < .10; **p < .05; ***p < .01.
To test whether the results are robust for distinct types of misconduct, all analyses were also done for three separate types of misconduct: violent misconduct, contraband infractions and defiant behaviour. These results are not shown, but discussed in the main text.
Research limitations
The current research has a number of limitations. First, only a few independent variables have been included in the analyses. As a consequence of the relatively small sample size, adding more predictors, although theoretically relevant, could have been proved to severely limit the validity of the findings. More indicators of inmates’ social and economic support and criminal lifestyle (e.g. drug use, membership of a criminal clan) definitely would have added value to the current investigation. Second, as already mentioned, the sample over-represents inmates incarcerated for more than five years; consequently, it over-represents inmates convicted of violent offences, as well as those sentenced to longer prison terms. The results might thus be biased by these over-representations. Third, because the data were collected in only four prisons, the findings should be interpreted with caution and not generalized to other facilities. Replications of this study among representative samples of inmates drawn from all Romanian prisons are definitely needed to confirm the current research findings.
Results
Importation characteristics and misconduct
According to the importation model, prisoners’ socio-demographic characteristics, in particular their age, marital status and parenthood, can be assumed to be inversely related to inmates’ misconduct. The results of the bivariate analyses (Table 2) show strong corroboration of these expectations. First, prisoners aged 35 years or older are significantly less often involved in misconduct in the previous year and the previous two years, respectively. Married (or cohabiting) prisoners also have a lower likelihood of misbehaviour than non-married prisoners. Yet parental status significantly differentiates inmates’ likelihood of committing prison infractions only during the previous year: inmate fathers have half the likelihood of inmates without children of committing prison infractions.
The results of the logistic regression analyses on total misconduct (Table 3), as well as on the three specific measures (not shown), indicate that inmates’ age holds an inverse and significant relationship with all types of misconduct except defiance. Irrespective of whether misconduct was studied for the previous year or for the previous two years, and for Model A or Model B, older inmates were found to be less likely to commit total misconduct, contraband violations or violent misconduct. Being married and having children provide more diverse findings depending on the time period and type of misconduct examined. Both are negatively related to committing total misconduct (although the results are mixed for the one- or two-year period). However, being married is not significantly correlated with any of the three particular types of misconduct in any of the time periods or models. Having children is significantly but, contrary to our expectations, positively related only with contraband violations in the previous two years.
Based on the theoretical ideas of the importation model it can also be expected that inmates with more violent offending patterns and prior incarcerations are more likely to misbehave. The results of the bivariate analyses (Table 2) provide little support for this assumption. Only having a history of violent offences significantly distinguishes between inmates’ chances of misconduct in both time periods: as anticipated, inmates having a history of violent offences have a higher likelihood of engaging in misconduct. Prior incarceration shows no significant differences, while being currently locked up for a violent crime discriminates between inmates’ likelihood of committing prison infractions only in the previous year.
The logistic regression results for both total (Table 3) and particular types of misconduct show further weak corroboration of our hypotheses. Inmates having a history of violence are indeed more likely to participate in total misconduct and defiance, but not in contraband violations and violence. But inmates convicted of a violent offence are more likely to engage only in contraband infractions. Yet, highly counter-intuitively, prior incarceration was found to be consistently unrelated to any type of misconduct.
Deprivation characteristics and misconduct
According to the deprivation model, it can be expected that inmates held in units with higher security are more likely to misbehave than those confined at lower prison security levels. Also in line with the deprivation model, it can be expected that inmates who work in detention have a lower prevalence of prison infractions. The bivariate analysis presented in Table 2 indicates that these expectations are validated. Prisoners held at closed and maximum security regimes have a higher chance to misbehave than those confined with less strict security regimes. On the other hand, prisoners involved in work activities have significantly lower chance to misbehave than prisoners who do not work.
The regression analyses for total (Table 3) and specific types of misconduct (not shown) also show that, both in the previous year and in the previous two years, those incarcerated in closed and maximum security regimes are far more likely to participate in total misconduct, contraband violations and violent misconduct. However, the relationships must be interpreted with caution as some inmates might be confined in these regimes precisely because of their misbehaviour. Surprisingly, prison regime does not significantly correlate with the likelihood of defiant behaviour. Work in prison is negatively correlated with misconduct, as previous literature has shown, but its influence is limited to total and violent misbehaviour in the previous two years.
The deprivation model also implies that inmates who have spent more time in detention and those sentenced to longer prison terms have higher probabilities of misconduct. Table 2 shows that, contrary to these expectations, neither of these two indicators has significant relationships with misbehaviour. The results of logistic regressions largely support the results of the bivariate analysis (Table 3). Length of stay has no significant relationship with total and defiant misconduct, but it has significant and strongly positive relationships with violent infractions in both time periods, and with contraband infractions only in the previous year. On the other hand, length of sentence is inversely correlated with defiance, but also with total misconduct and contraband infractions, although for total misconduct and contraband infractions the relationship is more sensitive to the time period and model under investigation.
Based on the deprivation model, it can be expected that – in addition to objective deprivation characteristics – inmates’ subjective evaluations of prison deprivation would also affect their misbehaviour in prison. Data reported in Table 2 reveal substantial differences with respect to subjective deprivations. In both time periods, prisoners who rated all 13 prison aspects as problematic/very problematic, as well as those who rated their treatment by staff, the violent behaviour of other prisoners and the opportunities for social valued activities in a similar way, engage in total misconduct significantly more often than their counterparts. The differences are also significant for all specific indicators of these three subscales. On the other hand, prisoners who assess the physical prison conditions as problematic/very problematic have a similar likelihood of misbehaving as those for whom prison conditions were not an issue; however, there are significant differences between prisoners who perceived two specific indicators of this subscale as problematic/very problematic and those who did not: the quality of the food and the hygiene conditions. Furthermore, prisoners who rated dealing with prison routine as being problematic/very problematic are only slightly more often involved in misconduct than those who were not bothered by this aspect; yet the differences are significant, although none of the specific indicators yields such significant differences.
In addition, the logistic regressions presented in Table 3 – Model A – show that the overall subjective deprivation scale has a significant and strongly positive relationship with the prevalence of total misconduct in both time frameworks. In relation to distinct types of misconduct (not shown), the overall subjective deprivation scale shows more mixed results. Table 3 – Model B – indicates that only two of the five specific subscales of prison deprivation, namely the violence and social valued activities subscales, are significantly and positively associated with total misconduct in the previous year. Interestingly, the two subscales turned non-significant for misconduct during the previous two years; however, the physical prison conditions subscale gains significance.
The findings for particular types of misconduct show that the violence subscale is linked with violent misconduct in both periods studied, and also with defiant infractions in the previous two years. The treatment by staff subscale is strongly correlated with defiance during the two time periods, but also with violence in the previous two years. The social valued activities subscale holds a moderately significant, albeit strong, relationship with contraband infractions in the previous year only.
Discussion and conclusions
The current study adds to a well-established tradition of research in correctional literature, one that examines the factors associated with prison misconduct. Its main goal was to expand the existing scholarship, largely dominated by US studies, by exploring the relationships between certain importation and deprivation characteristics and the prevalence of various types of misbehaviour (that is, total, violent, contraband and defiant), using a sample of Romanian inmates. The study demonstrated a clear relationship between importation and (both objective and subjective) deprivation characteristics on the one hand and prison misconduct on the other hand, as well as a differential impact of these characteristics according to the type of prison infraction examined.
Overall, the study supported many of the expectations derived from the theoretical importation and deprivation models, and validated many early empirical findings based on US and/or West European data. For example, the study is in line with research that found strong relationships between inmates’ age and total and major forms of misconduct (for example, violence and contraband infractions) (for example, DeLisi et al., 2004; Morris et al., 2010; Tewksbury et al., 2014). However, contrary to this literature, the current study found defiance not to be influenced by age. This suggests that, unlike the more serious forms of misconduct, which have higher prevalence among younger prisoners, minor infractions such as defiant behaviour may occur with the same likelihood across all age categories in Romanian prisons. A possible explanation of this surprising finding may be found in the ‘procedural justice’ theory: younger and older participants in our sample may show comparable disobedient behaviour simply because they share similar negative perceptions of the legitimacy and fairness of prison rules/treatment by staff.
Yet our study also challenges other outcomes that have been previously reported in the importation literature. For example, the characteristics of inmates’ offending history were found to be less clearly linked to misconduct than previous Anglo-Saxon research documented (DeLisi et al., 2004; Drury and DeLisi, 2010; Morris et al., 2010; Sorensen and Cunningham, 2010). The most intriguing finding was the consistent absence of significant relationships between prior incarceration and any of the four types of misconduct included in the study. One potential explanation is that prison recidivists are more familiar with the prison environment and, therefore, they may adjust more easily to prison. But it should be kept in mind that measures of misconduct were constructed based on official data. It is also possible that misconduct committed by recidivists goes largely undetected and unreported because they have improved their ways of avoiding apprehension.
Congruent with prior literature are the findings on the relationship between misconduct and objective deprivation characteristics, mainly the prison regime (Arbach-Lucioni et al., 2012; Innes, 1997; Worrall and Morris, 2011). We found that prisoners incarcerated in higher security regimes are more likely to misbehave, in general, and to commit violence and contraband infractions, in particular. As the deprivation literature suggests, this may be directly linked to the restrictions and deprivation that characterize these regimes (for example, more time spent locked in the cell; fewer visits; and fewer opportunities for participation in prison programmes). But there are other explanations as well. First, prisoners confined in these regimes may in general be more prone towards misconduct and violence and some might even be transferred from less strict prison regimes precisely because of their insubordination and aggressive behaviour. Second, it is also possible that prisoners held in higher security regimes are more closely monitored by prison staff (for example, more frequent cell searches) and therefore their misconduct is more often detected and officially registered. Surprisingly, defiance was not related to prison regime as the deprivation model assumes. This implies that minor rule violations are more ubiquitous in Romanian prisons, having similar prevalence across prison security levels. One potential explanation also lies in the procedural justice approach indicating that, disregarding the prison regime, prison rules and treatment by staff in Romanian prisons might be seen as unfair and unjust; this in turn may affect prisoners’ noncompliance with rules. The relationships between length of stay/length of sentence and misconduct provide mixed, albeit interesting, findings. Length of sentence was linked to contraband infractions and defiance, but not to total and violent misconduct. On the other hand, length of stay was correlated only with violence, thus supporting several prior research findings, which showed that long-terms prisoners have a higher likelihood of violent misconduct (Berg and DeLisi, 2006; Dhami et al., 2007; Steiner and Wooldredge, 2008). According to the deprivation argument, the longer the prisoners are incarcerated, the more intensely they feel the prison’s deprivations; consequently, they will be deeper immersed in prison subculture and will have a higher probability of acting violently.
Unique results were found in relation to the subjective deprivation factors. The overall subjective deprivation scale and all but one of the specific subscales (the prison routine subscale) were found to be strongly correlated to misconduct, although their influence was largely dependent on the period of time and type of misconduct examined. The study showed, for example, a consistent relationship between treatment by staff and defiance. In line with the deprivation model, one of the most likely explanations of this finding lies in the general oppositional attitude towards staff that is deeply entrenched in the prisoners’ subculture (Sykes, [1958] 2007). But a psychosocial explanation can also be employed, since concerns about harsh and unsupportive staff – the core elements of this subscale – might provoke stress and anger in inmates, who in turn act defiantly (Colvin, 2007). A procedural justice approach can explain this relationship as well. In addition, the relationship between -the perception of violence in prison and violent misconduct was straightforward and may be linked to Toch’s (1977: 157) fight–flight premise and to the fact that some prisoners’ need to counter the perceived threat of violence with violence. All these findings indicate, however, that inmates’ subjective evaluations of prison deprivations could be robust empirical constructs and therefore provide important contributions to the understanding of misconduct.
More research is definitely needed to further explore the relationship of subjective deprivation factors to misconduct and to clarify the mechanism through which particular prison deprivations are related to particular prison misconduct in Romania. First, the current study focused only on individual-level predictors and put great emphasis on several perceptual measures of deprivation. Future studies might also incorporate macro-level (objective) characteristics of prison deprivation in order to contrast their effect on misconduct with the effect yielded by subjective deprivations.
Second, the study used a cross-sectional research design and measured misconduct over two distinct periods of time. Further research could also embrace a longitudinal approach in order to test whether the deprivation factors are influencing prisoner misconduct differently depending on the phase of imprisonment (for example, early, middle and late); see, for example, Wheeler’s (1961) analysis on prisonization.
Third, the data we used did not allow us to consider the informal experiences of the prison regime. We therefore encourage subsequent research to take a step in this direction and to analyse how experiences of confinement in different regimes influence prisoners’ misconduct. Using Crewe’s (2011) unique framework of analysis focused on prison ‘depth’, ‘weight’ and ‘tightness’ would provide invaluable insights in this regard.
Fourth, the data of the current study did not allow us to unravel the underlying mechanisms through which particular characteristics of prisoners and prisons are (or are not) related to particular prison misconduct. Qualitative or mixed-methods research should be further employed in order to do that and, thus, to better understand who resorts to misconduct, why and in what prison circumstances (deprivations). For example, we strongly believe that the lack of a relationship between defiant misconduct on the one hand and prisoners’ age and the prison regime on the other hand deserves further research attention, especially in the Romanian prison context. This research impetus is emphasized by findings on the relationship between perception of treatment by staff and defiant and violent misconduct. As already suggested, studies conducted from a procedural justice approach could and should clarify these results.
Policy recommendations
Research on the importation and deprivation correlates of misconduct is highly relevant for prison administrators preoccupied with finding the best ways of increasing prison order and safety via reducing the risk of prison misconduct. The findings of the current study can provide valuable insights in this regard, especially to Romanian correctional practitioners. First, the risk classifications in prison and the assignments to various prison regimes are usually based on prisoners’ ‘imported’ characteristics. The finding that inmate age is a strong predictor of three out of four types of misconduct (total misconduct, contraband infractions and violence) should therefore inform prison administrators that they have to allocate additional treatment resources to younger inmates. Their misconduct may be the consequence of a myriad of factors such as impulsivity, social immaturity, a strong orientation towards (anti-social) peer groups, and fewer conventional ties. All these potential triggers must be properly addressed and tackled in order to prevent and/or reduce younger prisoners’ misconduct.
Second, it is important for prison administrators to consider not only the importation factors but also the deprivation factors related to misconduct. As such, further attention should be given by correctional authorities to the inmates held in closed or maximum security regimes, because their misconduct might be employed as a means of satisfying needs and dealing with restrictions. In line with the social prevention model (Bottoms et al., 1995: 193), efforts should be concentrated on lessening the deprivations of the prison environment by providing prisoners with the basic necessities and offering opportunities for personal autonomy.
Third, the relationships between the four types of misconduct and the overall and various subtype scales of deprivation could also be highly informative for prison authorities. Although, as mentioned in the introduction, the conditions of confinement in Romanian prisons are very poor, the prisoners’ perceptions of prison conditions did not stand out as a powerful correlate of prison misconduct. The opposite is true about prisoners’ perceptions of their treatment by staff: their strong association with defiant infractions should warn the prison administrations that additional efforts are needed in order to improve the quality of staff–prisoner relationships, including training staff members in skills related to human relationships and/or working with various categories of offenders (see also DiIulio, 1987).
In the same way, the relationship between concerns about the violent behaviour of other inmates and violent misconduct should make prison officials more attentive to prisoners prone to violence. One step has already been taken by the Administrația Națională a Penitenciarelor, which has developed a national strategy to reduce aggressive behaviour in prisons. The document provides, among other things, the elaboration of a tool for estimating the risk of violence. However, after identifying these inmates, further consideration should also be given to how they would best be managed during the entire period of incarceration.
Footnotes
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by a grant from the Romanian Ministry of Education and Research, CNCS – UEFISCDI, project number PN-II-RU-PD-2012-3-0116: ‘Effects of Imprisonment on Romanian Offenders’ Lives’, 2013–2015. This work was also supported by a grant from the Netherlands Institute for Advanced Study in the Humanities and Social Sciences (NIAS).