Patterns of Delinquent Behavior in the Brazilian Youth Justice SystemAbstract:
Bridging the gap between scientific knowledge and professional practice remains challenging in studies on offending behavior. This research, grounded in the Social and Personal Control Theory of Deviant Behavior, aimed to identify different levels of severity of engagement in offending behavior to support the development of evaluation parameters for adolescent offenders. Semi-structured interviews were conducted with 158 male adolescents (aged 12 to 19) serving legal measures in two regions of Brazil. Severity indicators, descriptive statistics, latent class analysis, and qualitative exploration of the variables were employed. The results revealed three distinct groups with varying levels of severity of engagement in offending conduct, consistent with existing literature. The findings highlight the need for specific guidelines for professional care and suggest the possibility of a fourth class comprising abstinent adolescents or those with minimal offenses, characteristic of the school-aged adolescent population.
Keywords:
psychological assessment; antisocial behavior; juvenile delinquency; forensic psychology
Resumo:
Aproximar o conhecimento científico da prática profissional é um desafio nos estudos sobre a conduta infracional. Esta pesquisa, fundamentada na Teoria da Regulação Social e Pessoal da Conduta Criminal, teve como objetivo identificar distintos níveis de gravidade de engajamento em condutas infracionais, com vistas a fundamentar estudos para definição de parâmetros de avaliação para o adolescente autor de ato infracional. Entrevistas semiestruturadas foram realizadas com 158 adolescentes, sexo masculino (12 a 19 anos), cumprindo medidas socioeducativas em duas regiões do Brasil. Utilizando indicadores de gravidade, empregou-se estatística descritiva, análise de classes latentes e exploração qualitativa das variáveis estudadas. Resultados revelaram três grupos com diferentes níveis de gravidade de engajamento na conduta infracional, alinhando-se à literatura existente. Destaca-se a necessidade de diretrizes específicas para atendimento profissional e a possibilidade de uma quarta classe de adolescentes abstinentes ou com reduzidas infrações, característico da população de adolescentes escolares.
Palavras-chave:
avaliação psicológica; comportamento antissocial; delinquência juvenil; psicologia forense
Resúmen:
Acercar el conocimiento científico a la práctica profesional es un desafío en los estudios sobre conducta delictiva. Esta investigación, basado en la Teoría de la Regulación Social y Personal de la Conducta Delictiva, tuvo como objetivo identificar los diferentes niveles de gravedad de la participación en delitos, para definir los parámetros de evaluación de los adolescentes que han cometido delitos. Fueron realizadas entrevistas semiestructuradas a 158 adolescentes del sexo masculino (12-19 años) que cumplían medidas socioeducativas en dos regiones de Brasil. Se utilizaron indicadores de gravedad, estadística descriptiva, análisis de clases latentes y exploración cualitativa de las variables. Los resultados revelaron tres grupos con diferentes niveles de severidad en la participación en conductas infractoras, en consonancia con la literatura. Se destaca la necesidad de orientaciones específicas para la atención profesional y la posibilidad de una cuarta clase de adolescentes abstinentes o con infracciones reducidas, características de la población adolescente escolar.
Palabras clave:
evaluación psicológica; conducta antisocial; delincuencia juvenil; psicología forense
Although the advancements promoted by longitudinal studies on criminal behavior since the 1980s are undeniable, especially compared to the cross-sectional studies that predominated before then, gaps remain⸻particularly regarding effective interventions to address the problem (Bonta & Andrews, 2023; Le Blanc, 2021). Despite the increased number of publications since the 2000s in the Brazilian context, the impact of scientific knowledge on the assistance provided to adolescents in conflict with the law remains limited.
Such a context is reflected both in the lack of standardized criteria to guide the decisions of Juvenile and Youth Judges and in the absence of a solid foundation to support the assessment and intervention strategies adopted by institutions responsible for implementing legal measures (Cardozo & Maruschi, 2023; Komatsu & Bazon, 2017). The gap between scientific knowledge and professional practice is a global challenge that deserves attention. Concrete proposals and evaluation programs are crucial to assist in the development of effective intervention plans aimed at reducing adolescent involvement in criminal acts and interrupting the behavior of those whose conduct is already evident (Le Blanc, 2021).
Results from longitudinal research databases have made it possible to observe both the persistence and changes in antisocial behavior over time, fostering the development of explanatory theories and practice models that have led to the proposition of interventions. These analyses have expanded the understanding of the mechanisms of initiation, activation, and aggravation of criminal behavior; the identification of different causal factors or factors associated with criminal behavior; the effects of these factors at various stages of development; and the influence of life events and transition phases on criminal behavior (Bonta & Andrews, 2023; Le Blanc, 2021).
The first studies adopting a developmental approach identified that criminal behavior typically begins between 9 and 14, peaks between 16 and 19, and rapidly declines in early adulthood, between 17 and 29 years. However, Gottfredson and Hirschi (1990) explain criminality as a combination of opportunity and low self-control, with the latter being determined during the socialization process in childhood and remaining stable throughout life. Consequently, the age/crime curve would not vary, as it essentially reflects the biological process associated with development. From this perspective, the authors argue that longitudinal studies are unnecessary for investigating the predictors of criminal or offensive behavior’s onset, continuity, and desistance.
Nonetheless, Loeber and Le Blanc (1990) highlighted different ways to analyze quantitative and qualitative changes in the course of criminal behavior, including frequency, variety, and trajectories and correlates that distinguish them. Longitudinal data have demonstrated the crime curve in various forms of antisocial behavior across two generations (1960 and 1980), revealing general patterns and identifying specificities within different trajectories (Le Blanc, 2021). Among other studies that also demonstrate the existence of distinct patterns of criminal or offensive behavior, Moffitt’s (2018) taxonomy stands out, identifying two distinct trajectories: the “adolescence-limited” trajectory, in which criminal behavior appears mainly during adolescence and tends to disappear in adulthood, and the “life-course-persistent” trajectory, in which antisocial behaviors emerge in childhood and persist into adulthood. A review of empirical studies that adopted Moffitt’s taxonomy identified a new group of adolescents characterized by abstinence or a very low number of infractions (Moffitt, 2018).
Research using Cambridge databases, focusing primarily on dynamic risk factors during childhood, identified four distinct trajectories of criminal behavior: (1) Minor recidivists, characterized by theft and generally non-serious but frequent offenses; (2) Serious and versatile recidivists, involving a wide range of offenses, including the use of violence; (3) Rare or non-recidivists, involving property-related offenses; and (4) Rare or non-recidivists, involving violent offenses (Basto-Pereira & Farrington, 2018).
In line with these authors (Basto-Pereira & Farrington, 2018; Moffitt, 2018), the theoretical framework adopted in this study is based on the Social and Personal Control Theory of Deviant Behavior” (Fréchette & LeBlanc, 1987; Le Blanc, 2021), which identifies distinct longitudinal trajectories in adolescents’ criminal conduct. Fréchette and LeBlanc (1987) analyzed a sample of 470 adolescent offenders and, through the indicators “volume,” “diversity,” “age of onset of antisocial behavior,” and “severity of the offense,” presented the first version of a typology of criminal conduct comprising two types of conduct. The first, “common delinquency,” includes 90% to 95% of the adolescent population, regardless of social class. Most of these individuals are involved with at least one type of crime, usually minor; however, as they have social and psychological resources, they continue to lead a life of respect for laws and social norms without significant difficulties in social adaptation. Consequently, some adolescents commit crimes during adolescence but desist in early adulthood. The second type, “distinctive delinquency,” describes the behavior of a minority of adolescents whose social development is hindered by problems that result in antisocial behavior. Although their behavior may resemble that of others, specific developmental deficits contribute to persistent criminal conduct. Abstaining adolescents are estimated to account for around 5% of the general population. In more recent work, Le Blanc (2021) revised the classification, describing the groups as (1) “Abstainers”; (2) “Delinquency limited to adolescence,” which is subdivided into “common” and “transient” and (3) “Persistent Delinquency.”
In general terms, Fréchette and LeBlanc (1987) propose that conventional behavior occurs and persists when there is an adequate level of psychological development, particularly in terms of self-control, provided that strong bonds with society are maintained, external and internal regulation and control are appropriate, and prosocial models prevail over antisocial ones. Conversely, the absence of these conditions increases the likelihood of antisocial behavior. The construct “engagement” refers to the maintenance of criminal behavior over time, resulting from significant problems in psychosocial development, as opposed to occasional offenses typical of the developmental period. The leading indicators of the evolution of the level of severity of engagement in criminal or offensive behavior are precocity, frequency, diversity of offenses, and the severity of the offense, which are particularly useful in identifying the stage of escalation or worsening.
This article uses the construct “antisocial behavior” as defined by Le Blanc (2021), who proposes standardizing the concept in behavioral sciences. The term encompasses so-called “problematic behaviors” (serious childhood actions that threaten the physical or psychological integrity of others) as well as criminal or delinquent conduct committed respectively by adolescents or adults, regardless of official accountability for the act.
Aiming to identify adolescents whose offenses were occasional, in comparison to individuals whose behavior indicates problems in psychosocial development and, consequently, exhibit a higher level of severe engagement in offending, Fréchette developed the Échelle de Gravité des Délits [Scale of Seriousness of Crimes - SSC], for use by the Montérégie Youth Centre in Canada (Fréchette, 1980). Specifically based on behavior, the scale addresses twelve variables associated with offending history and describes the conditions and circumstances that support the planning and execution of offending acts. The instrument allows differentiating three levels of seriousness of engagement in offending: occasional, non-serious behavior; an ongoing offending trajectory; and a level with a high number of indicators of seriousness, including early onset of the behavior, a large volume of offenses, and the presence of offenses against the physical integrity of individuals. The variables Fréchette (1980) adopted in the SSC are: (1) Presence of infractions against people; (2) Number of crimes; (3) Presence of intoxication (alcohol and other drugs) during the infraction; (4) Use of instruments or weapons; (5) Planned organization; (6) Presence of accomplices with police records; (7) Friends or acquaintances with police records; (8) Absence of tension during the action; (9) Utilitarian motivation; (10) Presence of criminal history; (11) Early age at the beginning of antisocial practices; (12) Destruction of objects during the infraction.
Fréchette developed a “typology of delinquency severity” to establish a classification system for adolescents in Quebec (Canada). He classified “distinctive delinquency” (persistent) and, based on this typology, created the practical model known as the Integrated Differential Intervention Model (IDIM). This typology included not only behavioral dimensions but also social and psychological dimensions. “Sporadic marginal” refers to episodic actions that are infrequent and mild in severity, including social disobedience, school indiscipline, and sporadic drug use. “Inadequate Regressive” behavior characterizes offenders whose primary motivation is utilitarian, focused on immediate gain. Hence, such individuals act against property and repeat their mistakes. “Explosive conflictual” behavior involves severe offenses against persons to alleviate tension. This behavior can escalate rapidly, often including violent opposition to authority and the consumption of alcohol and other drugs. Finally, “autonomous structural” behavior is characterized by persistent and serious offenses, often carried out without any sense of tension, frequently involving the use of weapons and association with antisocial peers (Desjardins, 2014; Piché, 2006).
A study conducted using the SSC (Fréchette, 1980) involved its adaptation and validation in Chile as part of an assessment protocol based on the IDIM and the risk, need, and responsiveness model (Bonta & Andrews, 2023), addressing individual, family, school, and social dimensions, in addition to criminal behavior (Chesta et al., 2022). With some specificities—among which the inclusion of a group called “complex delinquency” stands out—the study identified a typology with three forms of criminal behavior: (1) Common and transient delinquency; (2) Persistent delinquency; (3) Complex delinquency, including sexual abusers, individuals with mental health disorders, and serial offenders (Pérez-Luco et al., 2017).
There are no records of studies using the SSC or IDIM to collect data in Brazil. Studies employing other self-reporting strategies to address criminal behavior were conducted to identify subgroups of adolescents: male schoolchildren (Komatsu & Bazon, 2017) and adolescents serving legal measures (Galinari & Bazon, 2021). These studies used parameters such as frequency and diversity of offenses, as well as the precocity in the manifestation of antisocial behavior, alongside other social and psychological variables, to define a typology of adolescents who engage in the practice of offenses.
This study aimed to identify different levels of severity of engagement in criminal behavior to support research efforts to define evaluation parameters for adolescents who commit criminal acts. In the juvenile justice context, such an assessment is essential to differentiate adolescents whose offenses are occasional — committed by young people with a history of respect for the law and adherence to social rules — from those whose involvement in criminal acts reflects problems or gaps in typical development. Additionally, an appropriate assessment is crucial to determine the most suitable socio-educational measure (considering both intensity and duration) according to the individual needs of adolescents within the justice system.
The behavioral dimension was specifically considered in the scope presented here to identify adolescents with some level of engagement in criminal behavior (requiring socio-educational intervention) rather than individuals whose criminal behavior is restricted to isolated acts. Considering that the group of adolescents was heterogeneous—serving sentences in both open and closed environments within the juvenile justice system, with significant variability in the frequency, diversity, and severity of offenses—the hypothesis is that the statistical comparison of the results in variables related to behavior would enable the identification of different classes of adolescents, with varying levels of engagement in criminal behavior.
Method
This article is part of a larger study that integrates the primary author’s doctoral dissertation. It is an empirical study with a cross-sectional approach, employing descriptive-interpretative and probabilistic analyses. A mathematical model was used to estimate the probability of each adolescent belonging to a distinct class, characterized by different levels of engagement in criminal behavior.
Participants
The convenience sample comprised one hundred and fifty-eight (158) adolescents aged 12 to 19 years (mean age = 16.5 years), all male, from two Brazilian regions: Southeast (n = 100) and Midwest (n = 58). In terms of education, they had an average of 8.4 years of schooling. Only 36.1% (n = 57) regularly attended school at the time of the assessment or admission. Among the others, 32.9% (n = 52) had dropped out during the school year, and 31.0% (n = 49) had dropped out more than a year. Regarding employment status, 63.9% (n = 101) did not engage in any pro-social activity; 29.1% (n = 46) had an informal job in legal activities, and only 7.0% (n = 11) were formally employed.
Most adolescents lived with their mothers (53.8%, n = 85), followed by both parents (18.3%, n = 29); grandparents, uncles, or siblings (13.3%, n = 21), or only with their fathers (7.6%, n = 12). Some lived alone, with a partner (5.7%, n = 9), or with a friend (1.3%, n = 2).
The predominant economic level was Class “C” (57.0%, n = 90), followed by Classes “D-E” (27.2%, n = 43), according to criteria from the Associação Brasileira de Empresas de Pesquisa [ABEP] (2021). There were also adolescents belonging to Class “B” (12.0%, n = 19) and one adolescent (0.6%) who belonged to Class “A.” The economic level of five adolescents (3.2%) was not reported.
Before the study was initiated, formal authorization was obtained from institutional guardians and family members. Additionally, the adolescents voluntarily provided their consent. Both adolescents and guardians signed the respective informed consent and assent forms.
Data were collected from October 2021 to January 2023, and cases from nine institutions legally responsible for implementing judicially determined legal measures, four located in the Southeast and five in the Midwest, were invited to participate in the study. All institutions presented similar characteristics in terms of objectives and technical strategies, as recommended by the Juvenile Justice System.
The adolescents were serving probation (n = 72), detention (n = 9), or awaiting the definition of a measure in provisional detention (n = 77) for having committed or being accused of involvement in criminal acts. Most of the adolescents serving probation were charged with “drug trafficking” (41.7%, n = 30), followed by “robbery” (22.2%, n = 16), “theft” (12.5%, n = 9), “homicide” (6.9%, n = 5), and other minor infractions (16.7%, n = 12). “Drug trafficking” predominated among adolescents in detention (55.6%, n = 5), followed by “robbery” (33.3%, n = 3) and “theft” (11.0%, n = 1). Among adolescents in provisional detention, “robbery” was slightly more frequent (45.5%, n = 35), followed by “drug trafficking” (44.2%, n = 34), “theft” (5.2%, n = 4), and one case each of homicide, serious injury, kidnapping, and receiving stolen goods (5.2%, n = 4). Analyses comparing the legal measures applied to adolescents show significant differences according to the region (Southeast and Midwest), confirming studies suggesting a lack of standardized criteria in judicial decision-making (Cardozo & Maruschi, 2023).
Note that measures specifically regarding homicide cases and those serving community-based socio-educational measures refer to the judicial determination at the time data were collected for this study. Therefore, we cannot exclude the possibility that these adolescents had previously served a more severe measure prior to our evaluation.
Instruments
Data were collected through individual in-person interviews with the adolescents using a semi-structured script that included sociodemographic variables, life history, and criminal behavior. The adolescents completed the Questionário de Classificação Econômica Brasil [Brazil Economic Classification Questionnaire] (ABEP, 2021) and a set of psychological assessment instruments. In this study, we highlight the offenses self-reported by the adolescents using the Semi-Structured Interview Script. The primary author specifically adapted this script for the study, based on the semi-structured interview script provided by the Montérégie Youth Center, Quebec, Canada, which was translated and adapted in Brazil by Estevão et al. (Fréchette, 1980). Collaborating researchers with extensive experience in the socio-educational system examined the content and format of this adapted script. The meetings served as “focus groups” to refine the interview script, ensuring its face and content validity (expert panel). A pilot study was conducted with adolescents in conflict with the law to identify potential adjustments and enhance comprehension.
The interview included the presentation of 15 (fifteen) cards, 14 (fourteen) of which contained the names of the most frequent infractions among adolescents, according to data from the Brazilian National Council of Justice [CNJ in Portuguese] (2020). The 15th card was blank and intended to represent infractions reported by the adolescents that were not included on the other cards. Additionally, each infraction card had a description on the back written in language accessible to adolescents (R. S. Cardozo, personal communication, September 23, 2021).
Initially, three variables were used to analyze the distribution of the results reported by the adolescents: (1) Volume of infractions (a weighted value was assigned to each type of infraction according to its severity); (2) Diversity (sum of the different types of infractions); (3) Age at onset of infractions. The following variables were then analyzed in terms of absolute and percentage frequency: (4) History of infractions involving violence against people; (5) Presence of accomplices with police records; (6) Planning; (7) Use of instruments or weapons; (8) Destruction of objects; (9) Alcohol and/or drug intoxication; (10) Utilitarian motivation; (11) Absence of tension; (12) Insufficient feeling of guilt; and (13) Antisocial peers. Items 4 to 13 were scored when the indicator appeared in at least one of the infractions self-reported by the adolescents.
Procedures
Data collection. After consent was obtained, the adolescents were individually assessed at the institution where they were serving their legal measures in a proper room for psychological examination, with only one researcher present. Data were recorded in writing or directly on the LimeSurvey Digital Platform. The meetings where the assessment instruments were applied lasted an average of 90 minutes.
The infractions listed on each of the 15 cards (shown one by one) were laid out on the table in front of each adolescent and read during the semi-structured interviews addressing self-reported infractions. The adolescent was asked if they knew what each infraction meant, and the description of the infractions on the back of the cards was read in case of doubt. After receiving clarification, the respondent was asked to select the cards containing infractions they had already committed, regardless of whether the police had caught them. The adolescents were then encouraged to draw a timeline of these infractions, starting with the first involvement and ending with the most recent. Questions were asked for each infraction to characterize the behavior before, during, and after the infraction.
Data analysis. The information provided during the semi-structured interview was used to determine the presence or absence of descriptors of engagement in criminal conduct, which were scored by the primary author in an Excel spreadsheet. The data were statistically analyzed using JAMOVI 2.6.13, JASP 0.19.1.0, and SPSS 23.0. Significance was set at 0.05.
Latent class analysis (LCA) (Weller et al., 2020) was performed based on the classification into quartiles of the variables related to the infractions reported by the adolescents: volume, diversity, and age at the first infraction. Considering the sample size, the analysis was performed with a permutation (bootstrap) of 1,000 repetitions. The indicators used in the LCA were compared between the groups using Kruskal-Wallis tests, followed by Mann-Whitney tests with Bonferroni correction to verify the characteristics of each class of adolescents.
Once the classes of adolescents were determined, the Chi-square test was performed for each of the other variables describing the severity of engagement in criminal behavior. The distribution of the results was evaluated based on the adjusted and standardized residuals for each variable.
Ethical Considerations
The research project complied with the guidelines of Resolution 466/12 of the Brazilian National Health Council. Additionally, it was submitted to and approved by the Institutional Review Board at the Faculdade de Filosofia, Ciências e Letras de Ribeirão Preto, Universidade de São Paulo - CAAE No. 48685021.2.0000.5407.
Results
Latent class analysis (LCA) indicated an adequate model for both three and four groups of adolescents. However, superior goodness of fit was found for the three-class model, as indicated by BIC, AIC, and CAIC values. The model with three classes presented a log-likelihood of -562, an AIC of 1182, a CAIC of 1300, and a BIC of 1271. Its entropy was 0.731, and it had 34 degrees of freedom. The G² value was 37.7, and p = 0.400. The model with four classes presented a log-likelihood of -554, an AIC of 1187, a CAIC of 1345, and a BIC of 1306. Its entropy was 0.750, and it had 24 degrees of freedom. The G² value was 23.0, and p = 0.740. Based on these goodness-of-fit indices, the three-class model was chosen. The adolescents were classified into one of three clusters.
The groups differed significantly in the three main variables: volume, diversity, and age at the onset of the first antisocial behavior. Group 1, characterized by less severe criminal engagement and composed of 77 adolescents (48.7%), presented a lower volume (M = 53) and diversity (M = 1.7) of infractions, as well as a higher age at the onset of criminal behavior (M = 14.8 years). Group 3, comprising 42 adolescents (24.7%), exhibited the greatest severity, with the highest volume (M = 521) and diversity (M = 6.4) of infractions and the lowest average age at onset (M = 12.0 years) of criminal behavior. Group 2, consisting of 39 adolescents (24.7%), presented intermediate characteristics across the three variables (mean volume = 122, diversity = 3.4, and mean age of onset = 13.9 years). The p-values were less than 0.001 in all analyses (Table 1).
In addition to the main variables, the groups were compared using other indicators of the severity of engagement in criminal conduct. The Chi-square results for each of the other indicators and the distributions assessed from the adjusted and standardized residuals for each variable are presented in Table 2.
Distribution of adolescents (n = 158) into three groups and the results of the Kruskal Wallis test for the three main descriptors of severity of engagement in criminal conduct
The residuals were consistent for all the complementary variables examined. Group 3 had a higher proportion of adolescents involved in severe criminal conduct and a lower number of adolescents with minimal involvement. The opposite pattern was observed in Group 1. Group 2 showed criminal involvement levels close to the general distribution, displaying intermediate proportions.
Discussion
The latent class analysis identified three distinct groups of adolescents with varying severity levels of criminal behavior engagement. These results are significant as they provide a broader understanding of the specific patterns of criminal behavior, highlighting the need for assessment instruments sensitive to adolescents’ developmental particularities.
The results corroborate existing literature (Basto-Pereira & Farrington, 2018; Fréchette & LeBlanc, 1987; Le Blanc, 2021; Moffitt, 2018), highlighting the importance of differentiating adolescents to determine whether legal measures are relevant and, if so, to identify the most appropriate measures to address the needs of individuals engaging in criminal acts.
Based on the Social and Personal Control Theory of Deviant Behavior, it is worth noting that Group 1, characterized by a reduced volume and limited diversity of offenses, as well as a late onset of this behavior (M = 14.8 years), is indicative of transient offenders (or those limited to adolescence), whose criminal behavior is occasional (Le Blanc, 2021). The low use of instruments and minimal destruction of objects indicate a lower level of behavioral planning. The feeling of guilt after the criminal act — more significant than that observed among adolescents in Groups 2 and 3 — and the reduced number of associations with antisocial peers reinforce a behavioral pattern that is more influenced by contextual factors, combined with internal experiences typical of adolescence, which are often disorganized due to the uncertainties inherent to this developmental period.
The characteristics described for Group 1 corroborate the scientific literature on occasional offending in adolescence (Le Blanc, 2021; Moffitt, 2018) and indicate a reduced probability of persistence and worsening of engagement in offending behavior compared to the other two groups. However, it is important to identify developmental gaps or vulnerabilities that may increase the likelihood of recidivist offending in order to address the adolescents’ individual needs and reduce their risk of future offenses.
It is important to note that, despite being isolated, some of the offenses committed by the adolescents in Group 1 can be serious. Fréchette and LeBlanc (1987) argue, however, that the severity of the offense does not necessarily imply a marked engagement in criminal behavior and may represent an isolated act resulting from specific situations or circumstances. Violence associated with situational and individual factors, including deficits in executive functions, high impulsivity, and high levels of anger and hostility, has been highlighted by various authors (Basto-Pereira & Farrington, 2018; Romero-Martínez et al., 2022; Vaughan et al., 2023). Yang et al. (2023) note that high levels of reactive aggression may be associated with depressive symptoms, warranting thorough investigation in each case. On the other hand, proactive violence — generally planned actions used strategically to achieve a specific goal — is considered a stronger predictor of persistence in criminal behavior compared to reactive aggression (Ritchie et al., 2022; Roncero et al., 2018). These findings emphasize the need to include the type of violence observed in adolescents’ assessments as soon as they enter the justice system. Avoiding the analogous naming of different phenomena is essential, as distinguishing between them has significant implications for planning intervention programs and determining whether such a program is needed for each case.
In addition to the early average age of onset (12 years old) and the higher frequency and diversity of offenses, Group 3 differs from Group 1 in all variables. Moreover, Group 3 differs from Group 2 in terms of the following variables: accomplices with police records, use of instruments, destruction of objects, utilitarian motivation, and absence of tension—suggesting the planning of criminal behavior, which characterizes almost all adolescents in Group 3 (92.9%). These results are reinforced by the high frequency of adolescents reporting the presence of all indicators of seriousness of engagement in criminal behavior, with high frequencies for “accomplices with police records” (97.6%), “planning” (92.9%), “utilitarian motivation” (100%), and “insufficient feeling of guilt” (95.2%).
According to Fréchette and LeBlanc (1987) and Le Blanc (2021), the group of adolescents with the most severe engagement in offenses tends to increase both the number and severity of offenses over time. Despite the cross-sectional design of this study, retrospective data, including information from childhood, were collected, revealing indicators of a worsening trajectory of offending behavior among adolescents in Group 3. The diverse self-reported offenses (M = 6.4) were higher than those in Group 2 (M = 3.4) and Group 1 (M = 1.7). Additionally, the high volume of self-reported offenses (M = 521) committed in the last year, far exceeding those in Group 2 (M = 122) and Group 1 (M = 53), combined with the other variables mentioned above, are clear indicators of the worsening of behavior in these cases.
The characteristics identified in the adolescents in Group 3 describe a more severe trajectory, similar to the persistent trajectory group described by Le Blanc (2021) and Moffitt (2018). Therefore, it is essential to include social and psychological variables associated with criminal behavior when evaluating adolescents in the justice system to identify the factors that should be prioritized to interrupt their criminal trajectory.
Group 2 assumes an intermediate position in all variables analyzed, including the volume and diversity of offenses and the age at which this type of behavior began. Despite a lower level of engagement compared to Group 3, Group 2 comprises a high percentage of adolescents who score positively for “planning” (79.5%), “utilitarian motivation” (89.7%), and “insufficient feeling of guilt” (84.6%). These characteristics warrant professional care and increased attention from the justice system, as they indicate a certain level of engagement in criminal behavior.
An analysis of the scientific literature in the field indicates that Group 2 can be compared to transient offenders or those whose offending behavior is limited to adolescence; however, they present a more significant number of offenses than Group 1. Therefore, it is essential to emphasize that contextual factors, such as drug use, labeling resulting from involvement with the justice system, school dropout, and association with organized crime, among other aspects, may favor involvement in antisocial activities and hinder or prevent a healthy transition to adulthood (Le Blanc, 2021; Moffitt, 2018).
In line with the theory underpinning this work, the results are similar to the groups suggested by Le Blanc (2021), except for the abstinent group, which is not represented in the sample. Including school adolescents in the sample, following the logic of the results presented here, would likely result in the emergence of a fourth group — comprising abstinent adolescents or those with very occasional offenses. This abstinent group, even if apprehended by the police, would not necessarily require socio-educational intervention, although the need for professional care in social protection could not be ruled out.
The results highlight the importance of tailored interventions that align with different levels of severity of engagement in criminal behavior and their associated risk and protective factors. Integrative assessments designed to identify priority factors in the intervention process are essential to develop an effective care plan and reduce the likelihood of adolescents reoffending. Furthermore, beyond ensuring effective interventions for adolescents requiring specialized assistance, the initial assessment upon admission to the justice system is crucial in preventing those who have committed occasional offenses from receiving socio-educational measures disproportionate to their needs. As shown in the literature, such excessive measures lack positive effects and may hinder an adolescent’s normal development or exacerbate antisocial behavior (Bonta & Andrews, 2023; Viljoen & Vincent, 2024). Therefore, identifying the different levels of severity of engagement in criminal behavior is vital to support the comprehensive assessment of adolescents.
Finally, although the results align with the international literature, this study presents some limitations. First, the sample consisted of a limited number of adolescents within the juvenile justice system, which restricts the generalizability of the findings. Increasing the number of participants serving legal measures and including students would likely yield more robust and consistent results.
Another important limitation concerns retrospective data, which, although widely used in research, are more susceptible to biases than prospective data, whose information can be more reliably and precisely monitored. Despite the advantages of the prospective method, its high cost and lengthy follow-up periods pose significant challenges in most studies. However, future studies adopting a longitudinal design are desirable, as they can contribute to a greater understanding of the mechanisms of activation, aggravation, and deactivation of criminal behavior. Such knowledge is essential for working with adolescents and promoting public policies focused on primary prevention, thereby reducing the emergence of new cases.
References
-
Associação Brasileira de Empresas de Pesquisa. (2021). Critério de Classificação Econômica Brasil [Brazilian Economic Classification Criterion]. https://www.abep.org/criterio-brasil
» https://www.abep.org/criterio-brasil -
Basto-Pereira, M., & Farrington, D. P. (2018). Advancing knowledge about lifelong crime sequences. British Journal of Criminology, 59(2), 354-377. https://doi.org/10.1093/bjc/azy033
» https://doi.org/10.1093/bjc/azy033 - Bonta, J., & Andrews, D. A. (2023). The psychology of criminal conduct (7th ed.). Taylor & Francis.
-
Cardozo, R. S., & Maruschi, M. C. (2023). A importância da utilização de critérios de avaliação fundamentados em evidências na aplicação das medidas socioeducativas pelos magistrados brasileiros [The importance of the use of evidence-based evaluation criteria in the application of socio-educational measures by Brazilian judges]. Revista CNJ, 7(1), 123-138. https://doi.org/10.54829/revistacnj.v7i1.299
» https://doi.org/10.54829/revistacnj.v7i1.299 -
Chesta, S., Pérez-Luco, R., Alarcón, P., Wenger, L., Concha-Salgado, A., & García-Cueto, E. (2022). Empirical determination of transitory and persistent delinquency in Chilean youth: Validation of the criminal engagement severity scale “EGED”. International Journal of Environmental Research and Public Health, 19(3), 1396. https://doi.org/10.3390/ijerph19031396
» https://doi.org/10.3390/ijerph19031396 -
Conselho Nacional de Justiça. (2020). Reentradas e reinterações infracionais: Um olhar sobre os sistemas socioeducativo e prisional brasileiros [Reentries and infraction reiterations: A perspective on the Brazilian socio-educational and prison systems]. https://www.cnj.jus.br/wp-content/uploads/2020/09/Panorama-das-Reentradas-no-Sistema-Socioeducativo.pdf
» https://www.cnj.jus.br/wp-content/uploads/2020/09/Panorama-das-Reentradas-no-Sistema-Socioeducativo.pdf -
Desjardins, S. (2014). Guide de soutien à lá pratique: Les balises entourant l´expertise prédécisionnelle [Practice Support Guide: The Guiding Principles of Pre-Sentence Expertise]. Centre Jeunesse de Montréal: Institut Universitaire. https://www.unipsed.net/wp-content/uploads/2015/10/balises_entourant_expertise_predecisionnelle_guide_de_soutien.pdf
» https://www.unipsed.net/wp-content/uploads/2015/10/balises_entourant_expertise_predecisionnelle_guide_de_soutien.pdf - Fréchette, M. (1980). Échelle de gravité de l’engagement délictuel [Severity scale of criminal engagement] (R. Estevão, Trans.). Les Centres Jeunesse de La Montérégie.
- Fréchette, M., & LeBlanc, M. (1987). Délinquances et délinquants [Delinquency and delinquents]. Gaëtan Morin Éditeur.
-
Galinari, T. S., & Bazon, M. R. (2021). Criminal behavior and psychosocial risk factors in Brazilian adolescent offenders: An exploratory latent class analysis. International Journal of Environmental Research Public Health, 18(19), 10509. https://doi.org/10.3390/ijerph181910509
» https://doi.org/10.3390/ijerph181910509 - Gottfredson, M., & Hirschi, T. (1990). A general theory of crime. Stanford University Press.
-
Komatsu, A. V., & Bazon, M. R. (2017). Personal differences among Brazilian adolescents with distinct levels of engagement in delinquency. International Journal of Criminology and Sociology, 6, 65-74. https://doi.org/10.6000/1929-4409.2017.06.07
» https://doi.org/10.6000/1929-4409.2017.06.07 - Le Blanc, M. (2021). The development of antisocial behavior and crime: Replication with the Montreal cross sectional and longitudinal studies. Springer.
- Loeber, R., & Le Blanc, M. (1990). Toward a developmental criminology. Crime and Justice, 12, 373-473.
-
Moffitt, T. E. (2018). Male antisocial behavior in adolescence and beyond. Nature Human Behaviour, 2, 177-186. https://pmc.ncbi.nlm.nih.gov/articles/PMC6157602/
» https://pmc.ncbi.nlm.nih.gov/articles/PMC6157602/ - Pérez-Luco, R., Alarcón, P., Zambrano, A., Alarcón, M., Chesta, S., & Wenger, L. (2017). Taxonomía de la delincuencia adolescente con base en evidencia chilena [Chilean adolescent delinquency taxonomy evidence based]. In C. Bringas & M. Novo (Eds.), Psicologia jurídica: Conocimiento y práctica [Legal psychology: Knowledge and practice] (pp. 251-270). Universidad de Sevilla.
- Piché, J. P. (Coord.). (2006). Reintegrating young offenders into the community: Guide to intervention in youth probation. Ministère de La Santé et des Services Sociaux: La Direction des Communication.
-
Ritchie, M. B., Neufeld, R. W. J., Yoon, M., Li, A., & Mitchell, D. G. V. (2022). Predicting youth aggression with empathy and callous unemotional traits: A meta-analytic review. Clinical Psychology Review, 98, 102186. https://doi.org/10.1016/j.cpr.2022.102186
» https://doi.org/10.1016/j.cpr.2022.102186 -
Romero-Martínez, A., Sarrate-Costa, C., & Moya-Albiol, L. (2022). Reactive vs proactive aggression: A differential psychobiological profile? Conclusions derived from a systematic review. Neuroscience and Biobehavioral Reviews, 136, 104626. https://doi.org/10.1016/j.neubiorev.2022.104626
» https://doi.org/10.1016/j.neubiorev.2022.104626 -
Roncero, D., Andreu, J. M., & Peña, M. E. (2018). Efecto de diferentes patrones de agresión sobre la desadaptación institucional y la reinteración delictiva en menores infractores [Effect of different aggression patterns on institutional desadaptation and recidivism in juvenile delinquents]. Revista Española de Investigación Criminológica, 16, 1-24. https://doi.org/10.46381/reic.v16i0.159
» https://doi.org/10.46381/reic.v16i0.159 -
Vaughan, E. P., Speck, J. S., Frick, P. J., Walker, T. M., Robertson, E. L., Ray, J. V., Wall Myers, T. D., Thornton, L. C., Steinberg, L., & Cauffman, E. (2023). Proactive and reactive aggression: Developmental trajectories and longitudinal associations with callous-unemotional traits, impulsivity, and internalizing emoticons. Development and Psychopatology. Advance online publication. https://doi.org/10.1017/s0954579423000317
» https://doi.org/10.1017/s0954579423000317 -
Viljoen, J. L., & Vincent, G. M. (2024). Risk assessments for violence and reoffending: Implementation and impact on risk management. Clinical Psychology: Science and Practice, 31(2), 119-131. https://doi.org/10.1111/cpsp.12378
» https://doi.org/10.1111/cpsp.12378 -
Yang, Z., Luo, Y., Chen, F., Luo, R., Qi, M., Li, Z., & Wang, Y. (2023). Longitudinal associations between reactive and proactive aggression and depression in early adolescence: Between‐ and within‐person effects. Journal of Youth and Adolescence. 53(5), 1186-1196. https://doi.org/10.1007/s10964-023-01896-2
» https://doi.org/10.1007/s10964-023-01896-2 -
Weller, B. E., Bowen, N. K., & Faubert, S. J. (2020). Latent class analysis: A guide to best practice. Journal of Black Psychology, 46(4), 287-311. https://doi.org/10.1177/0095798420930932
» https://doi.org/10.1177/0095798420930932
-
How to cite this article:
Maruschi, M. C., Estevão, R., & Pasian, S. R. (2025). Patterns of delinquent behaviour in brazilian youth justice system. Paidéia (Ribeirão Preto), 35, e3513. https://doi.org/10.1590/1982-4327e3513
-
Article derived from the doctoral thesis of the first author under the supervision of the third, defended in 2024, in the Postgraduate Program in Psychology at the Universidade de São Paulo, Faculdade de Filosofia, Ciências e Letras de Ribeirão Preto. Support: Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (Process No. 88887.816973/2023-00)
Edited by
-
Associate editor:
Clarissa Mendonça Corradi-Webster
Publication Dates
-
Publication in this collection
11 Aug 2025 -
Date of issue
2025
History
-
Received
20 May 2024 -
Accepted
30 Jan 2025 -
Reviewed
18 Nov 2024
