Revenge and Hate offences among Brazilian and Mexican adolescents: prevalence and associated factors

Delitos de venganza y odio entre adolescentes brasileños y mexicanos: prevalencia y factores asociados

Marina Rezende Bazon¹*, Áurea E. Grijalva-Eternod², Caio dos Santos de Andrade¹, Maria Laura Polizeli Spagnollo¹, Vinicius Frayze David³

¹ Study and Research Group on Psychosocial Development and Intervention (GEPDIP), Faculty of Philosophy, Sciences and Letters of Ribeirão Preto of the University of São Paulo (FFCLRP-USP), Brazil.
² Centre of Social Sciences and Humanities of the University of Guadalajara, Mexico.
³ Faculty of Philosophy, Sciences and Letters of Ribeirão Preto of the University of São Paulo (FFCLRP-USP), Brazil.
* Corresponding author: mbazon@ffclrp.usp.br

Received April 2025 / Accepted July 2025

Abstract
This study is part of the International Self-Report Delinquency Study (ISRD4) and examines data from 4,558 adolescents between the ages of 13 and 17 from Brazil and Mexico. From a cross-cultural perspective, it estimates the prevalence of offences reported as motivated by revenge and hate. In addition, logistic regression models with stepwise variable selection were used to test psychological and social variables, as well as victimisation, on the likelihood of such offences. Prevalence rates were low, but online hate speech stood out with a rate of over 6%. The most frequently associated psychological variable was a pro-vengeance attitude, followed by low self-control. The most prominent social variables were family-related problems, including disruption, conflicts, and abuse/victimisation. Being Mexican reduced the likelihood of involvement in weapon carrying and assault, but increased the likelihood of involvement in online hate speech when considering victimisation. Taken together, the data provided considerable support for the propositions of the General Strain Theory.

Keywords: Revenge, hate crimes, adolescents, ISRD4, Latin America.

Resumen
Este estudio se inscribe en el marco del International Self-Report Delinquency Study (ISRD4), con datos de 4.558 adolescentes de entre 13 y 17 años de Brasil y México. Desde una perspectiva transcultural, se estimó la prevalencia de delitos declarados motivados por la venganza y el odio. Además, se utilizaron modelos de regresión logística con selección de variables por método stepwise para testar los efectos de variables psicológicas y sociales, así como de la victimización, en la probabilidad de cometer tales delitos. Las tasas de prevalencia fueron bajas, pero destacó el discurso de odio en línea, con una tasa superior al 6%. La variable psicológica más frecuentemente relacionada fue una actitud pro-venganza, seguida de un bajo autocontrol. Las variables sociales más destacadas refieren a problemas en la familia (desestructuración, conflictos y abuso/victimización). Ser mexicano contribuyó a disminuir las probabilidades de estar implicado en portación de armas y agresión, pero aumentó las probabilidades de participar en el discurso de odio en línea al considerar las victimizaciones. En conjunto, los datos apoyan en gran medida las propuestas de la Teoría General de la Tensión.

Palabras clave: Venganza, crímenes de odio, adolescentes, ISRD4, Latinoamérica.

1. Introduction

Understanding the offenders’ motivations can help comprehending psychological/emotional needs as tangible aspects that can be addressed in primary and secondary prevention programs, especially in adolescence. It can also help to circumscribe aspects of youth socialisation, considering the stimuli and environmental resources made available to young people (Erofeeva, et al., 2019). Motivation has generally been defined as a system of needs, objectives/goals and motives (Erofeeva, et al., 2019). This refers to the notion of the orientation (drive) that guides social behaviour (desires, interests and needs, whether conscious or unconscious), built throughout the individual’s relational history and internalised as mental schemas (Winter et al., 1998).

The term ‘motivation’ refers to different but complementary notions. Antisocial orientation (or antisocial cognition) embraces negative attitudes towards the law and the justice system, the belief that crime pays off, and the rationalisations that justify it, considered a predictor of criminal conduct (Walters, 2015). Offender’s motivation for treatment, as a counterpoint, is an aspect of "specific responsiveness" (Andrews & Dowden, 2006). A third notion, which is of greater interest to this study, refers to the "motive" (as immediate causes), which impel individuals to engage in certain types of offence as a dynamic factor for specific crimes (Farrington, 1993). Motive is defined as a conscious impulse to commit a specific act of will that poses a social danger and is considered a crime behind criminal law (Tararukhin, 1974, apud Erofeeva, et al., 2019).

Although the importance of studying and understanding why individuals act criminally in certain situations and not in others seems obvious, there are far fewer studies concerning motivation at this level, including in the Criminology field (Koegl & Farington, 2021). Scientific research into the motives behind crimes is still incipient, even though they are recognised in criminal law in relation to sentencing principles (Hessick, 2006).

Some researchers have sought to identify and categorise motives for offences through empirical studies (Koegl & Farington, 2021). Agnew (1990) created a list and submitted it to 1,300 adolescents who had revealed one or more of the 14 offences targeted by the research team. The participants’ answers to the open question "what made you do it?" were coded into 15 categories of motives: 1) self-gratification/pleasure-seeking; 2) utilitarian; 3) curiosity, emotion; 4) retaliation, revenge; 5) conflict, fight; 6) unintentional provocation; 7) anger, rage; 8) avoiding pain or punishment; 9) social pressure; 10) peer esteem; 11) moral justification; 12) boredom; 13) self-esteem; 14) power; and 15) political, ideological. Studies aimed to arrive at a smaller set of categories. Putniņš (2010), from a total of 24 motives, identified five broad categories: 1) immediate gratification; 2) interpersonal hostility; 3) social conformity; 4) material needs and 5) substance use. Gudjonsson & Sigurdsson (2004), likewise, identified four broad categories from 22 motives: 1) conformity; 2) provocation (e.g. losing control, taking revenge); 3) financial; and 4) excitement. Usually, two or more categories appear to underlie an offence, although one category may predominate. In an empirical study by Koegl & Farrington (2021), the strongest associations were between sensation-seeking and pleasure/happiness (r=0.70), escape and tension reduction (r=0.59), anger/jealousy/hate and retaliation/revenge (r=0.58), sadness/loss/despair and mental health (r=0.52), redemption/honour and retaliation/revenge (r=0.47), substance use and social influences (r =0.44), and between ideological motives and anger/jealousy/hate (r=0.39).

Motives were categorised by age too. Luneyev (1980, apud Erofeeva, et al., 2019) points out that in early adolescence, the most recurrent motives include the desire to have fun, to show strength/courage, or to assert oneself in the eyes of one’s peers, and to aspire to something special. In later adolescence greed, revenge and conformism become more prevalent. Farrington (2003), reviewing longitudinal studies, indicates that the reasons given for offending throughout adolescence vary significantly, ranging from the search for excitement/pleasure, to utilitarian motives, and anger (in the case of violent crimes and revenge). In contrast, utilitarian motives become dominant from the age of 20. LeBlanc (2003), drawing on data from a longitudinal study in Canada, assert that the main motivation behind antisocial/delinquent behaviour in adolescence is the hedonistic desire to seek satisfaction for its own sake. However, in distinctive/persistent delinquency trajectories, mixed motives are identified: hedonism, utilitarianism and anger.

The prominence of anger and hostility as a reason/motive for offences in adolescence has increased. In terms of theoretical elaboration, General Strain Theory (GST) argues that the tensions inherent in certain negative/stressful social/relational situations increase the chances of negative emotional reactions, such as anger and frustration. These reactions push individuals towards a form of action as a coping mechanism. It can be manifested as an illegal/violent act (Agnew, 2001; Agnew, 2015).

Anger has been identified as the primary catalyst for revenge offences (Caetano & Bazon, 2023; DiBlasi et al., 2024), which can be defined as a retaliatory response to perceived harm or injustice (McGaughey et al., 2025). Victims of offences have a higher risk of committing a crime in the future (McGaughey, et al., 2025), either against the original perpetrator or in a form of "displaced revenge", which targets a third party who was not involved in the original transgression (Jackson et al., 2019). The influence of gender on revenge is unclear. Some studies show that men experience a stronger desire for revenge than women, while others have found no such effect (Cota-McKinley et al., 2001). Older age has been shown to be associated with a lower desire for revenge (Jackson, et al., 2019). This association seems to be mediated by the development of self-control, which is considered important in avoiding revenge (Cota-McKinley et al., 2001).

Although there are no specific global statistics on anger/revenge offences among young people, it is possible to observe worrying trends. For example, in a study of 18,749 young people tried in Florida, approximately 18% cited anger/vengeance as their primary motive for their offences, ranking second only to impulsivity (36.78%) and financial gain (19.44%) (Craig et al., 2018). In this context, FBI data indicate that 61% of school shootings in the United States were motivated by rage/revenge (Vossekuil et al., 2004).

At this point, it is important to mention the so-called "hate crimes". According to the UN and various international courts, hate crimes are offences motivated by prejudice against the victim’s identity characteristics, such as race, religion, sexual orientation, nationality, disability, among others. These motives can be defined as "illegal acts involving the intentional targeting of a victim based on the perpetrator’s prejudice or bias against the victim’s real or perceived status" (Craig, 2002, p. 86). Such offences can manifest themselves in various ways, including vandalism, verbal attacks (online and offline), threats, physical assault, bodily injury, and even murder (Levin & McDevitt, 2022). At the macro-social level, these individuals are frequently driven by radicalised ideologies that engender intolerance and the dehumanisation of the victims. At an individual level, anger-revenge frequently serve as prominent motives. This phenomenon is especially prevalent in societies characterised by historical inequalities and/or identity tensions (Kivivuori, et al., 2016; MDHC, 2023).

In particular, cyber hate is unfolding in a context of increasing tension between free speech, hate speech, and censorship, driven by the widespread use of social media (Walters, 2023). Young people are considered the most susceptible to radicalisation when exposed to extremist ideas due to psychosocial needs related to identity processes (Jahnke et al., 2020). In addition, they are regular Internet users and are more exposed as “witnesses” and as “direct victims” of hate crimes on the Internet (Simões, 2024).

2. Justification and objectives

Revenge and hate crimes are not a new phenomenon. Despite this, research focusing on offences with such motivations, especially involving young people, is not particularly extensive. The limited availability of statistics hinders the ability to make comparisons between countries regarding the prevalence of such offences and their associated factors. The present study aims to contribute to the existing knowledge on this subject, verifying the rates of adolescents involved in revenge and hate offences in Brazil and Mexico, according to self-report data. It will also analyse the relationships between offences and relevant psychological and social variables, considering similarities and differences between countries. International comparisons are very important in this field, enabling national social and cultural differences to be explored as structural determinants of revenge and hate crimes, among other factors (Fisher, 2024; McGaughey et al., 2025).

Neither Brazil nor Mexico has a law that defines the concept ‘revenge crime.’ However, identifying this motive for different crimes can affect the severity of the criminal response in both countries, as revenge can be considered a despicable motive. Similarly, neither Brazil nor Mexico has a specific law that defines and classifies ‘hate crimes,’ although the term is recognised in doctrine and jurisprudence in both countries and is used to describe crimes motivated by prejudice and discrimination against groups or individuals due to characteristics that are generally associated with socially vulnerable groups.

Latin America is the region most affected by high levels of violence and crime (International Monetary Fund, 2024), and it is associated with structural factors [e.g. highest economic inequality and extreme poverty, ethnic and racial discrimination, significant gender inequality, etc. (World Population Review, 2025)]. Additionally, the region has recently witnessed an increase in the spread of radicalised ideas, causing ideological polarisations, which greatly intensified conflict and social tensions (Sarsfield et al., 2024).

Brazil and Mexico are the largest and richest Latin American countries, with a leading role in world economic and political affairs. Unfortunately, they are also at the top of the violence rankings, and stand out in problems linked to racial and ethnic discrimination. The situation is further compounded by the observation that both countries exhibit high impunity rates (with Brazil registering 33.24 and Mexico 46.19 on a scale of 0 to 100), achieving scores that surpass the global average (31.64) according to the Global Impunity Index 2024 (Le Clercq & Cháidez, 2024). In this scenario, young people are particularly vulnerable (Díaz, 2019; Araújo, 2024). Adolescents are frequently exposed to several forms of violence within their communities, educational institutions, and domestic environments, resulting in chronic stress that can adversely impact their emotional and psychological development (World Bank Group, 2020; ONU, 2020).

Although Mexico and Brazil have many similarities, they also exhibit differences in certain contextual aspects that make their comparison relevant. In terms of social trust, data from Latinobarómetro (2023) indicate that, while interpersonal trust is higher in Mexico (30.8%, compared to only 3.9% in Brazil), Brazil shows greater trust in institutions responsible for security and justice. Specifically, 47.3% of Brazilians reported having some or a great deal of trust in the police, compared to 30.9% in Mexico. This difference may be related to a higher perception of corruption and tolerance toward this practice in Mexico, as indicated by the Rule of Law Index 2024 (World Justice Project, 2024). Moreover, Latinobarómetro (2023) suggests that Brazil demonstrates greater institutional strength, as reflected in its higher support for democracy (45.6%) compared to Mexico (34.9%). However, paradoxically, a larger proportion of people in Mexico perceive themselves as being somewhat or fully protected from crime (34.8%) compared to Brazil (16.5%). Likewise, reports of victimisation in the past year are significantly higher in Mexico (51.6%) than in Brazil (23.6%).

Taking into account the similarities and differences existing between Mexico and Brazil, this study aims to explore the occurrence of revenge and hate motivated offences reported by adolescents in both countries and to examine the psychological and social variables associated with such behaviours. In this way, the study provides comparative evidence that contributes to a broader understanding of the factors shaping youth violence within the Latin American context.

3. Method

3.1. Sample description

The current study employs data from the International Self-Report Delinquency Study (ISRD4) (Marshall et al., 2022), collected from a total of 4,558 adolescents/students from public and private schools in Brazil and Mexico, aged between 13 and 17. Table S1 in the supplementary material summarises the characteristics of the sample. The socioeconomic status of the sample was calculated based on their perceived family and personal financial situations (relative economic position), after applying the percent of maximum possible (POMP) score transformation (Cohen et al., 1999), where 0 represents the lowest perceived economic position and 100 represents the highest.

The subsamples were equivalent in terms of gender distribution and average age. However, significant disparities were observed among the subsamples. In Mexico, there were more immigrant participants with an immigrant father and/or mother than in Brazil, although in both countries the trend was the same, i.e., the majority were non-immigrants. There was also a significant difference in family composition, with a higher proportion of Brazilians reporting living in a two-parent family or with another configuration. Despite this, in both subsamples, the majority reported living in a two-parent family. The average of relative economic position also differed between countries, with Brazilians having a higher average. The size of this difference, however, was small (Cohen´s d = 0,20).

3.2. Measures

The ISRD4 standardised questionnaire was employed, and its items comprised several measures, as detailed in Marshall et al. (2022). In this study, the variables of interest were the offences that the adolescents reported having committed motivated by revenge and hate, in offline contexts—Weapon carrying (carried a weapon, such as a stick, knife, or gun), and Assault (beaten someone up or hurt someone with a stick, knife or gun so badly that the person was injured) - and in the online context - Intimate postings (posted, re-posted or texted an intimate photo or video of someone that he or she did not want others to see), Hacking (hacked into a computer or cracked an account that does not belong to you, to get or destroy data), and Online hate speech (sent hurtful messages or comments on social media about someone’s race, ethnicity or nationality, religion, gender identity, sexual orientation, or for similar reasons). This was assessed using yes-no questions. With the exception of the Online hate speech question, all the others were follow-up questions (e.g. “You said earlier that you had beaten someone up or hurt someone with a stick or knife so badly that the person was injured. When you did this, were you getting revenge for something the person had done or said?”). Online hate speech was captured using a single question: “Have you ever sent hurtful messages or comments on social media about someone’s race, ethnicity or nationality, religion, gender identity, sexual orientation, or for similar reasons? Yes or No?”. Considering the exploratory nature of the study, other ISRD4 measures were tested as predictor variables, including those of a psychological, social/interactional nature, as well as variables related to negative family events and different types of victimisation. Table S2 in the supplementary material presents information about them.

3.3. Procedures

Following the study protocol (Marshall et al., 2022), each national coordinator managed the ethical aspects of the study in their respective country. Data were collected in the schools between September 2022 and February 2024. The questionnaire was administered to groups of 20–30 students and typically took between 30 minutes and 1 hour to complete. In Mexico and Brazil, data were gathered using tablets or computers, as the questionnaire was accessible online via LimeSurvey software.

Scale scores were computed by averaging the items within each scale. Considering the exploratory nature of the study, the original scale values were maintained to facilitate the interpretation of the results. The data was summarised using descriptive analyses. Next, binomial logistic regressions were carried out with the different offences as dichotomous outcomes in two stages. First, psychosocial variables were considered as predictors, then, victimisation variables were considered. In all analyses, the adolescent’s gender, relative economic position, and the country were also included as control variables. As the sample varied according to the offence, analyses were carried out separately for each crime. Additionally, given the large number of predictor variables, a stepwise variable selection procedure was applied. The process started with a full model including all the predictors and iteratively removed variables with the highest p-values, continuing until only predictors significant at the 5% level remained. All variables were treated as covariates, with the exception of negative family events and family conflicts, which were treated as ordinal factors. Odd Ratios (Ors) were used as a measure of effect size, indicating the direction of the observed effect: odds ratios greater than 1 suggest an increased likelihood of the occurrence of the offences as the values of the respective predictors increase, while values less than 1 indicate an increase in the opposite direction. Multicollinearity was tested for the models, considering acceptable tolerance values greater than 0.2. Model quality was estimated using Nagelkerke R² and the area under the curve (AUC). Nagelkerke R² is a measure of the model’s effect size, representing the percentage of explained variance. The AUC reflects the model’s ability to distinguish between the occurrence and non-occurrence of offences. Values above 0.5 indicate stronger predictive performance than chance, while values closer to 1 indicate greater predictive accuracy.

4. Results

The rates of adolescents who reported having committed offences motivated by revenge or hate in the Brazilian and Mexican sub-samples, as well as in the total sample, are summarised in Table S3 of the supplementary material.

The prevalence of teenagers reporting the offences under study is generally low. Considering the sample as a whole, only 0.6% of adolescents in Brazil and 0.5% in Mexico reported carrying a weapon in revenge. However, considering who reported carrying a weapon for different reasons, the proportion in Brazil was 6.42% and in Mexico 3.96%. Assault for revenge was relatively higher in the sample as a whole: 2.2% in Brazil and 1.6% in Mexico. As for who reported assault on someone, the proportions are very high: 55.13% in Brazil and 32.56% in Mexico.

Overall, only 0.7% of respondents in Brazil and 0.8% in Mexico reported intimate content posting to get revenge. From those who reported doing so, 40% in Brazil and 23.5% in Mexico did it for revenge. Hacking to get revenge, overall, accounts for only 1% in Brazil and 0.6% in Mexico. Among those who reported hacking, the proportion was 15.1% in Brazil and 15.5% in Mexico. The ‘hate speech online’ was the most prevalent form of revenge (4.3% in Brazil and 7.5% in Mexico). Table 1 summarises the results concerning psychosocial scales.

Reduced logistic regression models proved to be adequate, according to the Hosmer-Lemeshow test. As well as gender and relative economic position, other variables tested did not show significant relationships: ‘Happiness’, ‘Minority Identity’, ‘Perception of Discrimination’ and ‘Perception of Violence’, ‘Family bond’, ‘Parental Supervision’ and ‘School bond’. Afterwards, binomial logistic regressions were carried out using the victimisation variables (yes or no). As the sample size varied according to the offence, the analyses were carried out separately for each crime. The stepwise technique was also used. Table 2 summarises results.

The variables attack/aggression motivated by ordinary reason and online intimate posting did not show significant relationships with any of the offences focused on. Also, assault and intimate posting for revenge were not associated with any of the victimisations researched. Furthermore, multicollinearity tests did not indicate tolerance values below 0.7 for any variable in any of the tested models. In addition to chi-square statistics and p-values, tables 1 and 2 present the odds ratios (OR) for each variable, the Nagelkerke R², and the area under the curve (AUC).

5. Discussion

This study aims to analyse, from a cross-cultural perspective, self-reported offences motivated by revenge and hate in a large sample of adolescents from two Latin American countries: Brazil and Mexico. The prevalence rates found in the study are generally low, but this should be considered in light of the study’s limitations, which stem from the use of self-report questionnaires that may introduce social desirability bias, particularly in sensitive behaviours such as hate speech and revenge crimes.

Nevertheless, the rate of hate speech found can be considered relatively high (6.1% of the total sample). This offence refers to a rapidly increasing crime, especially in countries experiencing growing ideological radicalisation around social issues. This is partly due to the ease with which it can be committed, considering the accessibility of many young people to computers and the anonymity favoured by the use of the internet. According to the ISRD3, the prevalence of hate crime victimisation among adolescents aged 12 to 16 at the time ranged from 2% to 7% in 15 countries around the world, without specifying if it was online or offline (Marshall et al., 2015). In a more recent study, focusing on a sample of 3,620 students in grades 7 to 9 in schools in Germany and Switzerland, Castellanos et al. (2023) found that 21% of students identified themselves as offline perpetrators and 33% as offline victims, while these percentages were lower for online hate speech (13% and 20%, respectively), which seems somewhat counterintuitive and warrants further investigation. As reported by the FBI Hate Crimes (USA) and the European Union’s Fundamental Rights Agency (FRA), hate crimes can range from 5% to 20%, depending on the type of offence and the country analysed, with our findings consistent with this data. In any case, offline and online hate speech is widespread and highly correlated when it involves teenagers, with very negative repercussions for those involved, as well as for social cohesion (Castellanos et al., 2023).

Secondly, the prevalence of assault by revenge is noteworthy. Fights and assaults are important domains of physical violence among adolescents. Studies indicate that the prevalence of physical violence during adolescence is high, ranging from 34.5% to 63.3% depending on the country, although there is a significant downward trend in rates in several low-income and middle-income countries, according to a recent study conducted with samples from 30 countries in Africa, Asia, and the Americas (Smith et al., 2024). Thus, the rate of assault for revenge (1.86%) and for other reasons (2.68%) in the total sample is apparently low, but it refers to a very serious/violent offence (concerning attacks in which the victim was injured), and is based on a sample of adolescents from the general population. This prevalence rate is analogous to that documented in the study by Näsi et al. (2016), which used a representative sample of Finnish teenagers and found a rate of 2.3% of respondents reporting involvement in hate-motivated aggression or bullying. Direct comparison of the overall prevalence across the sub-samples from Brazil and Mexico is precluded due to the presence of significant sociodemographic differences. However, it is noteworthy that in the Brazilian sub-sample, the ‘revenge’ motive constituted over 50% of the aggression-related responses, whereas in the Mexican sub-sample, this figure was approximately 30%. This suggests that ‘revenge’ may be more prevalent in Brazilian society. However, the non-response rate was consistently higher in the Mexican sub-sample, suggesting a reluctance to disclose certain behaviours. This could potentially lead to an underestimation of the prevalence of assault in Mexico.

Thirdly, the prevalence rates found for intimate posting and hacking for revenge are noteworthy, as both exhibit almost the same prevalence in the total sample (0.7%), yet diverge significantly when considering adolescents who reported these same offences with other motivations. Considering only the subsample of those who stated that they had already posted intimate content, 40% cited revenge as a motive in Brazil, whereas in Mexico this figure was 23.5%. The results together suggest that certain forms of offence are more recurrent when it comes to taking revenge and/or expressing anger, in line with a global trend (Gudjonsson & Sigurdsson, 2007). This specific action has been referred to in specialist literature as ‘revenge pornography’ and is considered a global internet phenomenon with a significant negative impact on victims. These victims are predominantly children and women, and the perpetrators are predominantly men, highlighting a gender issue (Hall & Hearn, 2017).

By contrast, only 1% of respondents in Mexico cited revenge as a motive for hacking, and 0.6% did the same in Brazil, although hacking carried out for other reasons reached a rate of 4.15% in the overall sample. Similarly, weapon carrying for revenge exhibited the lowest overall prevalence in the total sample (0.5% - 0.61% in Brazil and 0.46% in Mexico), although the rates of the offence itself, with other motivations, were the highest among the offences surveyed (10.2% of the total sample; 8.9% in Brazil; 11.2% in Mexico). Weapon carrying remains a matter of concern, according to the Gun Violence Archive (2023), because the majority of teenagers involved in the use of weapons die or are injured.

Focusing on the psychological and social variables as predictors, ‘vengeful attitude’ emerged as one of the variables most strongly associated with the offences in question, remaining significant in four of the five models tested. In all of them, it was found to increase the likelihood of committing the offence. The relationship between pro-revenge attitudes and revenge and hate crimes can be considered tautological. In other words, ‘vengeful attitude’ is one of the dimensions of criminal behaviour. Nevertheless, this result is important in terms of intervention and prevention, and reiterates findings such as those reported by Lowe et al. (2020), with a large sample of adolescents, which found that attitudes supporting revenge emerge as a significant predictor of involvement in violence among adolescents, even after controlling for other factors such as violent victimisation and gang affiliation. Therefore, assessment of pro-revenge attitudes and activities to reduce beliefs that favour violence, such as support for revenge, seems very important in the context of violence-reduction intervention, as well as promoting values that emphasise peaceful resolution of interpersonal problems could be an effective strategy to counteract the social acceptability of vengeful attitudes (Collyer et al., 2010). Furthermore, vengeful attitudes are very likely to be reduced if individuals trust that someone else, such as legal authorities, will act to restore balance when harm has been done. Thus, trust in authorities may be a key factor in ensuring that these attitudes or motivations do not lead to action (Asif et al., 2024).

Unusually, the vengeful attitude was not associated with weapon carrying for revenge. The study conducted by Copeland-Linder et al. (2012), which aimed to examine the effect of retaliatory attitudes on subsequent violent behaviour in young people, also found no significant relationship with weapon carrying. These authors state that the lack of association between these variables is not surprising, because it can be an act of aggression or self-protection. This offence was only significantly (and inversely) related to future expectations. Carrying a weapon is a risk-based offence. It is not a violent offence in itself, although it does involve taking a risk, because it significantly increases the likelihood of a violent outcome. Given this, the inverse relationship with future expectations seems to make sense. It is important to consider that the sample analysed consists of school-attending adolescents, who, to some extent, have greater future expectations precisely because they remain within institutional control structures. This raises concerns about the situation of young people without access to education, who have poorer future expectations, as is the case of many young people in Latin America.

Low self-control has been identified as a significant variable for assault and intimate posting for revenge. It was found to more than double the likelihood of intimate posting for revenge. In the literature, the association between low self-control and hate-motivated offences appears to be weaker than the association between low self-control and non-hate-motivated offences (Näsi et al, 2016). However, according to a meta-analytic review conducted by Burnette et al. (2013), there is evidence of a strong link between self-control and forgiveness (or more specifically, resistance to retaliation).

It is noteworthy that this particular offence - intimate posting for revenge - was also predicted by time spent on the internet (not for study purposes), suggesting the significance of young people’s routine and leisure time spent on unstructured activities (Osgood & Anderson, 2004; Olecká et al., 2022). Surveys of the general population demonstrate that the daily time spent using the internet is approximately 9 hours in Brazil and 7 hours in Mexico (considering all devices), with the global average being approximately 6 hours per day (Petrosyan, 2025).

Regarding online hate speech, it was the offence that showed significant relationships with the most variables. On the one hand, we identified negative associations with morality, measured by moral beliefs, and anticipated shame. On the other hand, positive associations were observed between online hate speech and various family-related variables, such as family disruption or conflict. The more frequently negative family events occur, the more likely adolescents are to be implicated in this offence. A recent systematic review on the subject highlighted factors associated with involvement in cyber-hate during adolescence, as a perpetrator, notably the normalisation of violence – with consequent moral disengagement – with a positive association, and closeness to family, with a negative association (Jiménez-Díaz & Del Rey, 2025).

Our finding suggests that online hate speech can serve as a way of releasing family tensions that cannot be expressed in any other way (Erzar, 2024). The relationship between theft and hate speech on the Internet can perhaps be interpreted in the same way. Being a victim of crime in the community, particularly theft, may be linked to feelings of insecurity and anger, which seem to increase the expression of these feelings in online hate speech. These findings appear to align with the propositions of GST (Agnew, 2001). Consequently, rather than focusing exclusively on censorship or prohibition, a more effective strategy for reducing hate speech on social media might be to encourage discussion and moral reflection among young people in different community settings and strengthen close personal relationships.

It is important to note that certain psychosocial variables were not pertinent when controlling for the characteristics of the sample, including gender, relative economic position, and country of origin. Exploratory analyses using ‘minority identity’ and ‘perceived discrimination’ were conducted with different hypotheses about their positive or negative relationships with revenge and hate crimes. ‘Happiness’, ‘family bonding’, ‘parental supervision’, and ‘school bonds’, were tested on the hypothesis that they would decrease the probability of revenge and hate crimes. ‘Perceptions of violence’ and ‘online identity’ were hypothesised to enhance the probability of reported crimes, particularly those committed online. However, none of these hypotheses were confirmed.

In a similar study, Näsi et al. (2016) found that frequent fights between parents, low parental supervision, low social control and male gender were significant factors. However, the variables of economic disadvantage, parental alcohol consumption, and low self-control were not significant, with the latter being associated with non-hate crimes only. The differences with the present study, at least with regard to self-control, may be due to differences in the behaviours examined. The authors only cited bullying and aggression. However, the similarities are noteworthy: the socio-economic aspect was not significant in either study and problems/conflicts within the family were significant for the ‘hate’ motivation. The authors comment that the results support classic theories such as Social Control (Hirshi, 1969) and GST (Agnew, 2001), with an emphasis on the latter.

Focusing on victimisation analyses, it is noteworthy that internet threats exerted a profound influence on the probability of offences, including hacking perpetrated for vindictive motives and the dissemination of online hate speech. The latter exhibited an increase of up to 200%. These findings highlight the intricate web of influences and factors that shape this behaviour, in line with the GST. The role of familial issues in this context is also highlighted, with adverse or disruptive events, parental conflict and abuse, both major and minor, being particularly salient factors.

Under these conditions, a very interesting finding refers to the fact that being female significantly increases the likelihood of being involved in online hate speech by 200%. This is an innovative result, as it captures a nuance that other studies have not captured. In most research conducted on cyberhate, being male has been shown to be a risk factor (Jiménez-Díaz & Del Rey, 2025). This suggests that women may be more exposed (or perceive themselves to be more exposed) to this phenomenon, either as targets or perpetrators of online attacks, and respond reactively. This finding appears to align with the prevailing notion that the stresses associated with victimisation in close or intimate relationships constitute a predominant factor in female delinquency, thereby providing further support for GST. Additionally, female tension manifesting as externalising behaviour appears to have gained greater traction in the online domain than in public spaces, which are more commonly associated with the expression of male tension.

Unusually, minor parental victimisation was negatively associated with weapon carrying for revenge, significantly decreasing the likelihood of such behaviour. The item that assesses this refers to being hit with an object, being punched or kicked forcefully or being beaten up. These actions, although not defensible, are very common parental ‘educational’ practices in the Brazilian and Mexican sociocultural contexts and may be indicative of family discipline (Bazon & Komatsu, 2023). Negative emotions associated with minor victimisation may be expressed by other antisocial behaviours (Chavira et al., 2022; Gadagnoto et al., 2022; Peterle et al., 2022), such as online hate speech.

It is important to note that, for most of the analysed offences, being Mexican decreases the likelihood of being involved in revenge and hate-motivated crimes (specifically in weapons carrying and assault), except in the case of online hate speech, where victimisation variables were considered predictors. This finding is counterintuitive, given that institutional trust is higher in Brazil, suggesting that revenge may be less necessary for achieving justice in that country. One possible explanation is that the higher level of interpersonal trust in Mexico may serve as a coping mechanism, reducing the need for revenge among adolescents. Another possible explanation is the greater presence of organised crime in Mexico, which exerts a form of social control (Mendoza, 2022). In future studies, maybe with mixed methodological approaches, a broader range of offences related with revenge/hate could be explored, according to the opinion of the adolescents themselves.

As for the limitations of the study, in addition to the aforementioned social desirability bias, sometimes raised by self-report questionnaires on sensitive topics, it is necessary to highlight the fact that this is a cross-sectional study. This constrains the interpretation of the relationships between social/contextual variables and revenge and hate crimes from being treated as causal. Regarding the chosen method of analysis, it is well known that the stepwise method can lead to model overfitting or difficulties in replicating results. Therefore, although evidence of associations between predictors and outcomes was identified, these findings should be interpreted with caution, particularly for offences with smaller sample sizes. However, given the considerable variation in sample size across groups, this approach appeared to be the most appropriate for maintaining consistency in the analyses of the different offences studied.

In the same line, it is worth noting that some of the independent variables tested showed weak reliability indicators. Particularly, the fact that Family Bond and School Bond have low α values (0.48 and 0.62, respectively) and do not show any significant relationship with the offences assessed in the sample suggests considerations, as the constructs in question do not seem to have been correctly captured by the items in the questionnaire. Nevertheless, the explorations carried out and the findings can guide decisions regarding the improvement of measures, as well as new studies with a design that allows for the testing of causal relationships.

References

Agnew, R. (1990). Adolescent Resources and Delinquency. Criminology, 28(4), 535 - 566. https://doi.org/10.1111/j.1745-9125.1990.tb01338.x

Agnew, R. (2001). Building on the foundation of General Strain Theory: Specifying the types of strain most likely to lead to crime and delinquency. Journal of Research in Crime and Delinquency, 38(4), 319-361.

Agnew, R. (2015). Using General Strain Theory to Explain Crime in Asian Societies. Asian Criminology, 10, 131-147. https://doi.org/10.1007/s11417-014-9198-2

Andrews, D. A., & Dowden, C. (2006). Risk principle of case classification in correctional treatment: A meta-analytic investigation. International Journal of Offender Therapy and Comparative Criminology, 50, 88-100.

Araújo, P. I. de O. (2024). Massacres enquanto signos de poder: A governança criminal por Los Zetas em Tamaulipas, México. SciELO Preprints. https://doi.org/10.1590/SciELOPreprints.9888

Asif, M., Fazal, H., Sarfaraz, M., & Ul Hassan, A. (2024). Investigating the Moderators in the Relationship Between Righteous Anger and Support for Lynching. Asian Journal of Criminology, 19, 573-592. https://doi.org/10.1007/s11417-024-09444-y

Bazon, M. R., & Komatsu, A. V. (2023). Desenvolvimento Social e da Personalidade e os Diferentes Padrões de Conduta Infracional na Adolescência. Editorial Pedro & João.

Burnette, J. L., Davisson, E. K., Finkel, E. J., Van Tongeren, D. R., Hui, C. M., & Hoyle, R. H. (2013). Self-Control and Forgiveness: A Meta-Analytic Review. Social Psychological and Personality Science, 5(4), 443-450. https://doi.org/10.1177/1948550613502991

Caetano, L. A. O., & Bazon, M. R. (2023). Ato Infracional na Adolescência: Revisão Integrativa dos Estudos na Perspectiva da Teoria Geral da Tensão. Cadernos de Psicologia, 3, p. 01-21.

Castellanos, M., Wettstein, A., Wachs, S., Kansok-Dusche, J., Ballaschk, C., Krause, N., & Bilz, L. (2023) Hate speech in adolescents: A binational study on prevalence and demographic differences. Frontiers in Education, 8, 1076249. https://doi.org/10.3389/feduc.2023.1076249

Chavira, D. A., Ponting, C., & Ramos, G. (2022). The impact of COVID-19 on child and adolescent mental health and treatment considerations. Behaviour Research and Therapy, 157, 104169. https://doi.org/10.1016/j.brat.2022.104169

Cohen, P., Cohen, J., Aiken, L., & West, S. G. (1999). The problem of units and the circumstance for POMP. Multivariate behavioral research, 34(3), 315-346.

Collyer, C. E., Johnson, K. L., de Mesquita, P. B., Palazzo, L. A., & Jordan, D. (2010). Sensitivity to Violence Measured by Ratings of Severity Increases after Nonviolence Training. Perceptual and Motor Skills, 110(1), 48-60. https://doi.org/10.2466/pms.110.1.48-60

Copeland-Linder, N., Johnson S. B., Haynie, D. L., Chung, S. E., & Cheng, T. L. (2012). Retaliatory attitudes and violent behaviors among assault-injured youth. Journal of Adolescent Health, 50(3), 215-20. https://doi.org/10.1016/j.jadohealth.2011.04.005

Cota-McKinley, A. L., Woody, W. D., & Bell, P. A. (2001). Vengeance: Effects of gender, age, and religious background. Aggressive Behavior, 27(5), 5. https://doi.org/10.1002/ab.1019

Craig, J. M., Wilson, A. D., Baglivo, M. T., Wolff, K. T., Piquero, A. R. & Epps, N. (2018). Why they did it? Sex differences in the impact of mental health and substance use on motivations for offending. Crime & Delinquency, 64(6), 723-757. https://doi.org/10.1177/0011128717718489

Craig, K. M. (2002). Examining hate-motivated aggression: A review of the social psychological literature on hate crimes as a distinct form of aggression. Aggression and Violent Behavior, 7, 85–101.

Díaz, A. F. (2019). A zona cinza: ordem criminosa e autodefesa armada no México. Tempo Social, 31(1), 277–299. https://doi.org/10.11606/0103-2070.ts.2019.141533

DiBlasi, T., Wydo, M., & Smith, K. (2024). An exploratory analysis of vengeful episodes in prisoners. Journal of Rational-Emotive & Cognitive-Behavior Therapy, 42(4), 930–944. https://doi.org/10.1007/s10942-024-00555-0

Erofeeva, M. A., Grinenko, A. V., Stanovova, L. A., Kosolapova, N. V., Mikhaylovsky, M. N., Rybina, I. A., & Kochetkov, I. G. (2019). Motivation and motives of juvenile delinquents. EurAsian Journal of BioSciences, 13, 135-140.

Erzar, T. (2024). Expressing Anger and Hatred Online: Effects, Consequences, Motivations and the Limits of Apparent Connectedness. Theological Quarterly, 84(4), 945-953. https://doi.org/10.34291/BV2024/04/Erzar

Farrington, D. P. (1993). Motivations for conduct disorder and delinquency. Development and Psychopathology, 5(1-2), 225-241.

Farrington, D. P. (2003). Developmental and Life-Course Criminology: Key Theoretical and Empirical Issues - The 2002 Sutherland Award Address. Criminology, 41(2), 221-255. https://doi.org/10.1111/j.1745-9125.2003.tb00987.x

Fisher, J. (2024). The Culture of Hate. Open Journal of Social Sciences, 12, 1-10. https://doi.org/10.4236/jss.2024.127001

Gadagnoto, T. C., Mendes, L. M. C., Monteiro, J. C. S., Gomes-Sponholz, F. A. & Barbosa, N. G. (2022). Emotional consequences of the COVID-19 pandemic in adolescents: Challenges to public health. Revista da Escola de Enfermagem da USP, 56, e20210424. https://doi.org/10.1590/1980-220X-REEUSP-2021-0424.

Gudjonsson, G. H., & Sigurdsson, J. F. (2004). Motivation for offending and personality. Legal and Criminological Psychology, 9(1), 69-81

Gudjonsson, G. H. & Sigurdsson, J. F. (2007). Motivation for offending and personality. A study among young offenders on probation. Personality and Individual Differences, 42(7). https://doi.org/10.1016/j.paid.2006.10.003

Gun Violence Archive. (2023). Gun violence archive. https://www.gunviolencearchive.org/

Hall, M., & Hearn, J. (2017). Revenge pornography: Gender, sexuality and motivations. London: Palgrave Macmillan. https://doi.org/10.4324/9781315648187

Hessick, C. B. (2006). Motive’s role in criminal punishment. Southern California Law Review, 80(1), 89–149.

Hirschi, T. (1969). Causes of Delinquency. Berkeley: University of California Press

International Monetary Fund (2024). Crime and violence in Lati America and the Caribbean: Facts, costs, and policy options. IMF. https://doi.org/10.5089/9798400288470.087

Jackson, J. C., Choi, V. K., & Gelfand, M. J. (2019). Revenge: A multilevel review and synthesis. Annual Review of Psychology, 70(1), 1. https://doi.org/10.1146/annurev-psych-010418-103305

Jahnke, S., Schröder, C. P., Goede, L.-R., Lehmann, L., Hauff, L., & Beelmann, A. (2020). Observer sensitivity and early radicalization to violence among young people in Germany. Social Justice Research, 33(3), 308–330. https://doi.org/10.1007/s11211-020-00351-y

Jiménez-Díaz, O., & Del Rey, R. (2025). Cyberhate in adolescents and youths: A systematic review of labels and associated factors. Aggression and Violent Behavior, 81, 1–11. https://doi.org/10.1016/j.avb.2024.102023

Kivivuori, J., Savolainen, J., & Aaltonen, M. (2016). The revenge motive in delinquency: Prevalence and predictors. Acta Sociologica, 59(1), 69–84. https://doi.org/10.1177/0001699315607969

Koegl, C., & Farrington, D. P. (2021). Advancing Knowledge about Motivations for Criminal Offending. Victims & Offenders, 17, 1-22. https://doi.org/10.1080/15564886.2021.1895940

Latinobarómetro. (2023). Latinobarómetro 2023: Percepción de la democracia en América Latina. https://www.latinobarometro.org/lat.jsp

Le Clercq, J. & Cháidez, A. (2024). Índice Global de Impunidad 2024. Midiendo el alcance de la impunidad global. Universidad de las Américas Puebla.

Leblanc, M. (2003). La conduite délinquante des adolescents: son développement et son explication En Leblanc, M. & Ouimet, M. et Szabo, D. Traité de criminologie empirique. Montréal: Les Presses de l’Université de Montréal, 2003, 779.

Levin, J. & McDevitt, J. (2022). Hate crimes. In L. R. Kurtz (Ed.), Encyclopedia of violence, peace, & conflict (3rd ed., pp. 28-34). Academic Press. https://doi.org/10.1016/B978-0-12-820195-4.00040-6

Lowe, C. C., Ellonen, N., Lee Miller, B., & Peltonen, K. (2020). Best Served Cold: Vengeful Attitudes and Violence among Finnish Adolescents. Deviant Behavior, 43(2), 163–178. https://doi.org/10.1080/01639625.2020.1789379

Marshall, I. H., Birkbeck, C., Enzmann, D., Kivivuori, J., Markina, A., & Steketee, M. (2022). International Self-Report Delinquency (ISRD4) Study Protocol: Background, Methodology, and Mandatory Items for the 2021/2022 Survey. (ISRDTechnicalReport#4), Northeastern University, Boston, MA.

Marshall, I. H., Enzmann, D., Hough, M., Killias, M., Kivivuori, J., & Steketee, M. (2015). Youth victimization and reporting to police: First results from the third round of the International Self-Report Delinquency Study (ISRD3) (Vol. 3). ISRD3 Technical Report Series. https://www.researchgate.net/publication/281411530_Youth_Victimization_and_Reporting_to_Police_First_Results_from_the_Third_Round_of_the_International_Self-Report_Delinquency_Study_ISRD3

McGaughey, K., Delaney, R., McGlinchey, E., Hanna, D., & Armour, C. (2025). The Psychosocial and Contextual Predictors of Revenge Desire and Attitudes in Crime Victims: A Scoping Review. Psychological Reports, 0(0). https://doi.org/10.1177/00332941241313032

Mendoza, M. (2022). The Tyranny of Narco-Power: Political Rule and Austere Domination in Michoacán, Mexico. The Journal of Latin American and Caribbean Anthropology, 26(3-4), 408-426. https://doi-org.wdg.biblio.udg.mx:8443/10.1111/jlca.12579

Ministério dos Direitos Humanos e da Cidadania (MDHC). (2023). Relatório de Recomendações para o Enfrentamento do Discurso de Ódio e o Extremismo no Brasil. Brasília. Disponível em: GOV.BR https://www.gov.br/mdh/pt-br/assuntos/noticias/2023/julho/mdhc-entrega-relatorio-com-propostas-para-enfrentar-o-discurso-de-odio-e-o-extremismo-no-brasil. Accessed on: 18/03/2025.

Näsi, M., Aaltonen, M., & Kivivuori, J. (2016). Youth hate crime offending: the role of strain, social control and self-control theories. Journal of Scandinavian StudieS in criminology and crime Prevention, 2016, 17(2), 177–184. http://dx.doi.org/10.1080/14043858.2016.1260332

Olecká, I., Pospisil, J., Trochtová, L., & Pospíšilová, H. (2022). Social Networking as Leisure: An Attempt to Conceptualize Liquid Leisure. Sustainability, 14(9), 5483. https://doi.org/10.3390/su14095483

Organização das Nações Unidas (ONU). (2020). Relatório sobre o Desenvolvimento Humano 2020. Disponível em: https://hdr.undp.org/es/content/informe-sobre-desarrollo-humano-2020

Osgood, D. W., & Anderson, A. L. (2004). Unstructured socializing and rates of delinquency. Criminology, 42(3), 519-550. https://doi.org/10.1111/j.1745-9125.2004.tb00528.x

Peterle, C. F., Fonseca, C. L., Freitas, B. H. B. M., Gaíva, M. A. M., Diogo, P. M. J., & Bortolini, J. (2022). Emotional and behavioral problems in adolescents in the context of COVID-19: A mixed method study. Revista Latino-Americana de Enfermagem, 30(spe), e3745. https://doi.org/10.1590/1518-8345.6273.3745

Petrosyan, A. (2025). Daily time spent online worldwide 2024. In: https://www.statista.com/statistics/1380282/daily-time-spent-online-global/

Putniņš, A. L. (2010). An exploratory study of young offenders’ self-reported reasons for offending. The Journal of Forensic Psychiatry & Psychology, 21(6), 950–965. https://doi.org/10.1111/j.1751-9004.2010.00322.x

Sarsfield, R., Moncagatta, P., & Roberts, K. M. (2024). Introduction: The New Polarization in PutniņšLatin America. Latin American Politics and Society, 66(2), 1-23. https://doi.org/10.1017/lap.2024.15

Simões, R. (2024). Género, violência e ódio online: Conceitos e representações. Imprensa da Universidade de Coimbra. https://doi.org/10.14195/978-989-26-2334-4

Smith, L., López Sánchez, G. F., Oh, H., Jacob, L., Kostev, K., Rahmati, M., Butler, L., Keyes, H., Barnett, Y., Yon, D. K., Shin, J. I., & Koyanagi, A. (2024). Temporal Trends of Physical Fights and Physical Attacks Among Adolescents Aged 12-15 years From 30 Countries From Africa, Asia, and the Americas. Journal of Adolescent Health, 74(5), 996-1005. https://doi.org/10.1016/j.jadohealth.2023.12.005.

Tararukhin, S. A. (1974). Criminal behavior Social and psychological traits. Moscow: “Legal literature”.

Vossekuil, B., Fein, R. A., Reddy, M., Borum, R., & Modzeleski W. (2004). The final report and findings of the safe school initiative: Implications for the prevention of school attacks in the United States. United States Secret Service and United States Department of Education.

Walters, B. (2023). The Future of Free Speech. Palgrave Macmillan.

Walters, G. D. (2015). Antisocial cognition and criminal thinking: Theoretical interpretations and biosocial implications. In M. DeLisi & M. G. Vaughn (Eds.), The Routledge international handbook of biosocial criminology (pp. 516–529). Routledge/Taylor & Francis Group.

Winter, D. G., John, O. P., Stewart, A. J., Klohnen, E. C., & Duncan, L. E. (1998). Traits and motives: Toward an integration of two traditions in personality research. Psychological Review, 105, 230-250. https://doi.org/10.1037/0033-295X.105.2.230

World Bank Group. (2020). Violence and Development. Disponível em: https://www.worldbank.org/en/topic/violence-and-development

World Justice Project (2024). Rule of Law Index. https://worldjusticeproject.org/rule-of-law-index/

World Population Review (2025, March 24). Gini Coefficient by Country 2025. https://worldpopulationreview.com/country-rankings/gini-coefficient-by-country

Open Science badges