Introduction

Interest in the concept of recidivism has grown significantly in criminological research in recent years. The study of this concept aims to understand the criminological and psychosocial reality of individuals who, once identified as offenders, are sentenced, serve the corresponding sentence, and then repeat their previous criminal behaviors or engage in new types of offenses (^{King & Elderbroom, 2014}; ^{Nakamura & Bucklen, 2014}). Under this perspective, criminal psychology has dedicated many decades to the study of recidivism with the aim of analyzing the processes of community re-entry and rehabilitation of offenders, the effectiveness of prison treatment programs and the preventive performance of recidivism risk assessment instruments, among other aspects (^{Bonta, Law, & Hanson, 1998}; ^{Costopoulos, Plewinski, Monaghan, & Edkins, 2017}).

The most widely used measure in empirical studies is criminal recidivism, re-entry to prison or re-imprisonment; including re-entry to prison for a new custodial sentence due to a failure to comply with standards of conduct, preventive imprisonment or the breach of a suspended sentence (^{Andres-Pueyo, 2015}). Many of these studies analyze recidivism in relation to determinants and variables such as risk factors, treatment programs or personal variables, which provide more exact knowledge of recidivism to help design appropriate prevention and intervention strategies (^{Andres-Pueyo, 2015}).

Although the comparison of recidivism rates at international level is complex due to factors such as differences in methodologies, follow-up times and criminal and prison laws, access to data across countries gives us a global vision of the scope of this phenomenon.

Few empirical studies on recidivism have been conducted in Spain, with the exception of the region of Catalonia, where a study on the prevalence of mental disorders in Spanish prisons (^{Vicens et al., 2011}) reported that 54% of prisoners were re-offenders, and a report monitoring recidivism in Catalonia from 201 to 2013, which found a re-offending rate of 30.2% (various authors, 2015). In 2012, a study was conducted on criminal recidivism and published by the Ministry of the Interior (^{Graña, 2012}). The study included a sample of 811 individuals and calculated the recidivism rate based on a detailed analysis of prison inmates with a criminal record. A total of 31.6% of the sample were reoffenders, of whom 10.1% had committed violent offenses and 8.5% sexual offenses. Subsequently, and replicating the study by ^{Vicens et al.}, ^{Zabala (2016)} studied the relation between recidivism and mental disorder in a sample of 184 inmates, finding a criminal recidivism rate of 41.8%.

In light of the above, it must be considered that the high rates of criminal recidivism in Spain have a significant impact, generating high social and economic costs. Hence, recidivism risk assessment represents an important challenge.

Regarding the predictors of recidivism, the findings of various studies focus on personal and socio-environmental variables, which, in many cases, precede the first arrest and should be taken into account when determining individuals with a high risk of recidivism (^{Andrews & Bonta, 2006}; ^{Bertone, Domínguez, Vallejos, Muniello, & López, 2013}). Over recent decades, research has established a series of risk factors to determine the characteristics that identify individuals with a higher likelihood of reoffending. These factors are related to sociodemographic characteristics, criminal history, antisocial behavior, violence-linked attitudes, treatment adherence and presence of psychopathology, especially antisocial personality disorder (^{Rodríguez, Gómez, Fernández, & Reyes, 2013}; ^{Yang, Wong, & Coid, 2010}). In addition, early substance abuse could be considered a “meta-variable” linked to other factors which trigger such behavior, including unemployment, a dysfunctional family environment, dropping out of school or lack of a social support network (Bertone et al., 2013; ^{Håkansson & Berglund, 2012}). The complexity of analyzing these diverse variables corroborates the need for empirical studies that contribute to determine the relations between the different factors and their direct association with recidivism (^{Grann, Danesh, & Fazel, 2008}; ^{Plattner et al., 2009}).

With regard to aggressiveness, there is evidence that the frequency of past violent behaviors is a robust predictor of future violent conduct (^{Fazel, Buxrud, Ruchkin, & Grann, 2010}). In this sense, the literature also suggests an association between premediated aggression and psychopathy, as individuals with a personality disorder are characterized by insensitivity, manipulativeness and the use of interpersonal violence (^{Hare, 2003}). A pioneering article published in 2015 (^{Swogger, Walsh, Christie, Priddy, & Conner, 2015}), using a one-year follow-up study of a cohort of offenders, compared impulsive and premeditated aggression as possible predictors of criminal recidivism. The findings indicated that premeditated, but not impulsive, aggression predicted violent reoffending, suggesting that assessment of the type of aggression may provide relevant information beyond the simple frequency of aggressive behaviors (Swogger, Walsh, Christie, Priddy, & Conner, 2015).

Findings on the association between psychopathy and recidivism are inconsistent. Although a significant negative relationship has been found between the presence of components of psychopathology in offenders and the prediction of future criminal behaviors, an individual’s criminal history has, however, been found to be the variable most strongly associated with criminal recidivism (^{Bonta et al.,1998}; ^{Costopoulos et al., 2017}; ^{Gendreau, Little, & Goggin, 1996}). Establishing a negative relationship between psychiatric diagnosis and recidivism would help mitigate the stigma surrounding persons with a mental health disorder and the supposed dangers inherent in their condition (^{Fazel & Yu, 2011}). However, it is important to note that other studies on the relationship between the presence of mental health problems and recidivism have found that inmates with a psychiatric disorder are more prone to reoffending and repeat incarceration compared with individuals with a psychiatric disorder but no previous incarceration (^{Baillargeon, Binswanger, Penn, Williams, & Murray, 2009}; ^{Pflueger, Franke, Graf, & Hachtel, 2015}; ^{Segeren, de Wit, Fassaert, & Popma, 2017}).

In view of the inconsistent findings in the previous literature, the aim of the present study is to shed some light on the role of factors such as psychopathy and aggression in criminal recidivism, with the hope of finding a positive association or results that contribute to the prediction of the risk of repeat offending, and the design of prevention and intervention strategies.

Method

Results

The study was conducted with 110 male prison inmates, of whom 55.8% (n = 53) were reoffenders, and 77.2% (n = 85) were under second-degree imprisonment. As regards the type of offense committed, the most frequent was robbery with violence, accounting for 30.5% of inmates (n = 29), followed by murder, 13.7% (n = 13), and homicide, 12.6% (n = 12). Most of the offenders were single or divorced, 48.6% (n = 53) and 16.5% (n =18), respectively, with no statistically significant differences between the reoffender and non-reoffender groups (χ ² = 1.472, p = .689, gl = 3).

Regarding education, the most common levels were primary (58.2%) and secondary level (31.8%), with no significant differences in the distribution of this variable between the two groups (χ ² = 4.862, p = .182, gl = 3).

Mean age was M = 41.4, SD = 9.05 years (Range: 24-69) with statistically significant differences according to recidivism (t = -2.219, p = .029, gl = 93), and a difference of 4.1 years between non-reoffenders and reoffenders. Non-reoffenders had 7 years’ more work experience than reoffenders (t = - 3.507, p = .001, g = 89).

The inmates with mental disorder were aged M = 41.8, SD = 9.54 years, while those without mental disorder were aged M = 41.0, SD = 8.60 years. There were no significant differences between the groups (t = -0.478, p = 6.38, gl = 108).

A total of 50% (n = 55) presented a mental disorder, but there was no statistically significant association with recidivism (χ ² = 3.261, p = .071, gl = 1). However, there was a higher proportion of participants with a mental disorder among reoffenders, 66% (n = 35). No association was found with the type of pathology, given that the most frequent problem was personality disorders with 58.2% (n = 32), followed by adaptive disorders, 23.6% (n = 13), and psychotic disorders, 18.2% (n = 10). The participants with mental disorder had received fewer years of schooling, M = 7.9, SD = 3.31 years, compared to M = 10.0, SD = 3.10 years for those without mental disorder TM (t = 3.312, p = .001, gl = 107). The individuals with mental disorder presented a higher level of reimprisonment, M = 2.6, SD = 1.66, compared to those without a disorder, M = 1.9, SD = 1.43 (t = -2.122, p = .036, gl = 93).

As regards the relationship between aggression and recidivism, we found statistically significant differences (p < 0.05) in all the dimensions studied, with values being higher in the reoffender group, except for Boldness and Impulsive Aggression, where no significant differences were revealed (Table 1).

Table 1.  Analysis of the differences (Student’s t-test) in the study variables between reoffenders and non-reoffenders. 

The individuals with mental disorder presented similar results to those for recidivism, with statistically significant differences (p < .05) in all the dimensions, and higher values in the mental disorder group, with the exception of Boldness, Verbal Aggression and Impulsive Aggression, where no significant differences were found (Table 2).

Table 2.  Analysis of the differences (Student’s t-test) in the mental disorder and aggression variables between reoffenders and non-reoffenders. 

Of the reoffenders (n = 53), 63.6% had a mental disorder TM (n = 35). A total of 39.6% (n = 21) of the individuals had been incarcerated twice, with a mean repeat incarceration rate of M = 3.3, SD = 1.44 (Range: 2-7). There were significant differences between those with and without a mental disorder (Table 3). Statistically significant differences (p < .05) can be observed for Physical Aggression, Anger, Hostility, Meanness and Disinhibition, with higher values in the mental disorder group.

The multivariate logistic regression analysis with recidivism as dependent variable revealed an Odds Ratio (OR) for meanness of 1.163 (IC 95%: 1.078-1.254). In other words, the higher the level of meanness, the greater is the likelihood of reoffending (Table 4). This model predicts correctly in 71.4% of the cases, with the area under the ROC curve being 78.4% for the predicted values (IC 95%: 68.7-88.1%) (p < .05).

Table 3.  Aggression in reoffenders and mental disorder. 

Table 4.  Variables in the logistic regression equation. 

Discussion

The aim of this work was to determine the relationship between aggression, psychopathy and criminal recidivism in two groups of inmates, one with mental disorder and one without.

The sample comprised 110 male inmates proportionally divided into two groups according to presence or absence of mental disorder. The majority of the participants were sentenced for crimes against persons (26.3% for homicide or murder) with a criminal recidivism rate of 55.8%. Few studies have been conducted in Spain on recidivism and mental disorder among prison inmates. The samples used present certain differences from that used in the present study since they were individuals who presented a series of risk factors increasing the likelihood of recidivism, linked to sociodemographic characteristics, criminal history, antisocial behavior, violence-related attitudes, treatment adherence and the presence of psychopathology, especially antisocial personality disorder (^{Rodríguez, Gómez, Fernández, & Reyes, 2013}; ^{Yang, Wong, & Coid, 2010}).

Our study revealed no differences in criminal recidivism according to the presence or absence of mental illness; that is, recidivism was not associated with inmates suffering from mental disorder. This result coincides with the findings of ^{Zabala (2016)}. The same result was also reported in studies conducted in other countries such as that by ^{Hall, Miraglia, Lee, Chard-Wierschem and Sawyer (2012}) on mentally ill persons leaving prison in New York State, which reported a similar risk of recidivism in inmates with and without mental disorder.

However, although we found no relationship between recidivism and mental disorder, our results did confirm a higher rate of repeat imprisonment among reoffenders with mental disorders compared to those without mental health problems. This coincides with previous works such as that by ^{Cloyes et al. (2010}), who conducted a study on Utah State prisoners with and without serious mental illness, finding that the median time to return to prison for prisoners with mental disorder was 3485 days, while for those without mental illness, this period was 743 days. Hence, persons with mental disorder reoffend sooner than those without, facilitating the higher number of repeat incarcerations among the former.

Personality disorders are the most common mental problem among the inmates in our study (58.2%), although this percentage is lower than that reported in other studies conducted in Spain (^{López-Barrachina et al. 2007}; ^{Vicens et al., 2011}; ^{2007 Zabala 2016)}. Other studies have also focused on the analysis of the associations between recidivism and specific disorders. ^{Zabala (2016)}, for example, reported that criminal recidivism was unrelated to personality disorders, while, ^{Fazel and Yu (2011}), found individuals with psychotic disorders presented a higher risk of reoffending compared to persons without mental disorder.

The analysis of the psychological variables under study showed that the reoffenders scored significantly higher on trait aggression, anger, hostility, physical and verbal aggression and premeditated aggression. With regard to the psychopathy factors, reoffenders scored higher on meanness and disinhibition. Meanness is the only psychological variables of the measures used in our research that predicted inmates’ recidivism.

These findings partly coincide with those of a study conducted in Spain by the Forensic Science Institute of the Autonomous University of Madrid, which measured post-treatment criminal recidivism among intimate partner offenders between 2010 and 2015¹. The study examined whether there existed pre-treatment differences in aggression between the reoffenders and non-reoffenders, finding higher levels of physical aggression in the reoffenders before the psychological intervention, but no differences in verbal aggression, anger and hostility. In addition, it is worth noting that in our study, the assessment of the psychological variables revealed the same results for inmates with mental disorder and reoffenders.

Finally, regarding the sociodemographic characteristics of the participants, the reoffenders were younger and had less work experience than the non-reoffenders, which highlights the need to focus attention also on individuals’ personal circumstances before their first arrest in order to prevent recidivism (^{Andrews & Bonta, 2006}; ^{Bertone, Domínguez, Vallejos, Muniello, & López, 2013}). In the case of inmates with mental disorder, we found that they had fewer years of schooling compared to those without disorder, coinciding with ^{Zabala (2016)}, whose study found that prison inmates with mental disorder had low levels of education.

In conclusion, the presence of mental disorder is not a determinant of an inmate’s likelihood of reoffending. However, inmates with a mental disorder do reoffend more frequently and score higher on all the aggression factors, as well as on disinhibition and meanness, with the latter being a predictor of recidivism.

The main limitations of our research lie in the sample. The sample size was small, there was a considerable age difference between the study groups (reoffenders and non-reoffenders), and the sample consisted only of male inmates, who were all incarcerated in the same prison. Thus, the lack of heterogeneity in our sample complicated a deeper study of recidivism according to the different mental disorders and types of offenses committed. Other difficulties we found were the diversity of the criteria for recidivism in this field, which complicates comparisons between studies at both national and international level, and the shortage of studies on recidivism and mental disorders, particularly studies including personality variables in the analysis of the inmates. We believe future research should tend towards further studies in Spain on criminal recidivism and psychopathologies, focusing in detail on the specific disorders that correlate with reoffending and including a larger number of personality variables in the study design, with the aim of improving recidivism prevention strategies.

In our opinion, the present work has important practical implications in the field of criminal psychology, as it identifies and analyzes measures of aggression and psychopathy involved in criminal recidivism, underlining the significance of the meanness variable in predicting prison inmates’ likelihood of reoffending. This finding, we believe, is important because it opens up a new line of intervention in the search for solutions to the prevention of risk behaviors in this population. In addition, it is worth highlighting the importance of the sociodemographic variable of low educational level. Identifying these variables will facilitate the development of new methods of intervention, which can be implemented both in prisons, to prevent recidivism, and in schools, as a measure of primary prevention.