What Are the Mechanisms Underpinning Intergenerational Transmission of Violence Perpetration? A Realist Review

CMO 1: Normalization of Violence and Harmful Gender Norms

If the parents were violent in the presence of or toward their child, in contexts characterized by high levels of violence and gender inequality, (C) (G1), then the child normalizes and justifies violence and adopts harmful gender norms (G2; M), which might result in violence perpetration in both adolescence and adulthood (O) (G2) (Figure 2).

OPEN IN VIEWER

Figure 2.

Final CMO configurations for the underlying mechanisms of the transmission of intergenerational violence perpetration.

Seven included studies supported this CMO focused on the cognitive phenomenon of “normalization,” where one’s schema about violence and the acceptability of it is altered by the broader context in which one lives (Domoney & Trevillion, 2021; Gage, 2016; Kimber et al., 2018; Li et al., 2021; Martinez, 2020; Reyes et al., 2015; Stith et al., 2000). Of these, three studies focused on homotypical continuity of IPV (Gage, 2016; Kimber et al., 2018; Martinez, 2020), and three studies focused on heterotypical continuity of violence in the following combinations: IPV (G1)/child maltreatment perpetration (G1) to youth violence (G2)(Li et al., 2021), IPV (G1)/Child maltreatment perpetration (G1) to IPV perpetration (G2) (Domoney & Trevillion, 2021; Reyes et al., 2015; Stith et al., 2000). Normalization is best explained by power theory, which posits that normative attitudes and social acceptance of violence, gender inequality, and family conflict intersect with each other at the structural level to increase tensions within a family. Economic hardships, power imbalances between partners and family members, and high levels of stress increase risk for use of violence, and witnessing or experiencing violence results in children learning that violence is acceptable to resolve conflicts (Straus et al., 1980).

A cross-sectional survey examined exposure to parental spousal violence and physical, psychological and sexual dating violence perpetration with 342 high school students in Port-au-Prince, Haiti, which has seen frequent coups d’états and civil unrest as well as street violence and domestic violence amongst other human rights violations, in addition to high levels of poverty, inequality and rapid urbanization (Gage, 2016). The authors state that violence is transmitted through the belief system where perceived peer tolerance to violence was a stronger factor associated with IPV perpetration than one’s own tolerance of violence. Therefore, children who observe violence in their families of origin and think their peers endorse violence may grow up to believe that violence is an appropriate way to resolve conflict in dating relationships (Gage, 2016). This is in support of both cultural cognitive theory, which posits that children who grow up in a culture that endorses violence, and experience parental violence, then also use violence to resolve conflict (Nisbett, 2003) and power theory, which sees the roots of violence transmission in both culture and family structure (Straus et al., 1980). Similarly, in a qualitative study with six Latino young adults who had witnessed domestic violence as children, Martinez (2020) found that witnessing domestic violence normalizes the use of violence in relationships and distorts views of what healthy relationships look like. As such, harmful gender norms were adopted, and violence was inflicted on intimate partners (Martinez, 2020).

In their longitudinal study of 1,100 adolescents in a Chinese metropolis, Li et al. (2021) found that children who witnessed family violence were more likely to engage with violent peers, which increased their risk for engaging in adolescent violence. These pathways were fully mediated by normative beliefs. The authors argue that children growing up in households characterized by violence would adopt behaviors carried out by their parents to solve conflicts; however, more impactful on adolescents’ use of violence is that those who experience violence frequently approve of it and associate with peers who carry the same normative beliefs and this is particularly true for boys (Li et al., 2021).

CMO 2: Modeling and Imitation

If parents engage in violence in the presence of their children (C) (G1), then their children have fewer chances to learn non-aggressive responses, and they may imitate their parents’ behavior (G2) (M), which can result in the perpetration of violence (O).

This CMO was supported by 12 included studies focused on the behavioral actualization of the schema where violence is acceptable (Augustyn et al., 2017; Fite et al., 2008; Gage, 2016; Goncy, 2020; Goncy et al., 2021; Kimber et al., 2018; Li et al., 2020; Murshid & Murshid, 2018; Smith et al., 2011; Spapens & Moors, 2020; Stith et al., 2000; van Dijk et al., 2019). Of these, ten focused on the homotypical continuity of violent criminal behavior (Augustyn et al., 2017; Spapens & Moors, 2020; van Dijk et al., 2019) and IPV (Fite et al., 2008; Gage, 2016; Goncy, 2020; Kimber et al., 2018; Murshid & Murshid, 2018; Smith et al., 2011; Stith et al., 2000), two on the heterogenetic continuity of parent-child aggression (G1) and IPV perpetration (G2) (Goncy et al., 2021; Li et al., 2020). This CMO is best explained by social learning theory, which posits that violence is used as an habitual response to conflict because of learned behaviors within the family where violent behavior modeled by adults provides scripts of violent behavior and teaches that violence is an acceptable mode to resolve problems (Bandura, 1977).

For example, Fite et al. (2008) used data from the longitudinal Child Development Project in the United States of 585 children recruited when entering reception class and their caregivers and found higher risk for aggression toward romantic partners in children of parents with violent relationships. They found that children whose parents have high levels of relationship conflict are modeled more aggressive and fewer nonaggressive responses to relationship conflict and therefore are more aggressive toward provocative situations in their own relationships. This is because they did not learn that a less aggressive response could achieve a similar or better outcome (Fite et al., 2008). Similarly, in a cross-sectional study of Haitian adolescents, the authors observed that in a context where violence and gender inequality are normalized at home, individuals learn to engage in interpersonal violence through observation, imitation, and modelling of significant others (Gage, 2016).

A global meta-analysis involving 98 unique samples found a small to moderate global effect for homotypical intergenerational transmission of IPV perpetration (Goncy, 2020). The authors used social learning theory to explain how individuals learn and replicate behaviors from their family of origin in their own relationships (Bandura, 1977). “Perhaps witnessing multiple forms of aggression, including physical forms such as hitting, as well as psychological forms, such as yelling or controlling, sends a powerful message that aggression is acceptable and normative in close relationships” (Goncy, 2020, p. 20). Although studies included in this review came from different contexts, little information was available about the contextual factors, for example, gender norms endorsed in the contexts studied, other than that domestic abuse in the parent generation occurred, and normalization of violence and harmful gender norms may have also been a mechanism.

Another global meta-analysis of 87 effect sizes found a small global effect for the intergenerational transmission of child maltreatment (G1) and IPV perpetration (G2), which was moderated by gender and stronger for men (S. Li et al., 2020). The authors explain this using the General Aggression Model (Anderson & Bushman, 2002), which posits that those with child maltreatment experience normalize the utility of violence and “form aggressive scripts” that they use to deal with stressors and frustrations in their intimate relationships, thereby resulting in the perpetration of IPV. This happens most commonly in the context of little family support and low resilience in the individual. Similarly, data from the Bangladesh Demographic and Health Survey 2007 of 3,396 men showed a significant association between childhood exposure to marital violence and marital violence perpetration in G2 (Murshid & Murshid, 2018). The authors observed that witnessing parental violence in childhood provides a justification for marital violence, therefore normalizing it and allowing it to perpetuate across generations, in line with social learning theory (Bandura, 1977). In the Netherlands, in a multi-generational qualitative study of eight families involved in violent crime, using archival data, criminal investigation files, and interviews with the police, social workers and probation offers, Spapens and Moors (2020) found that children learn that violence is an “acceptable strategy to dissolve disputes.” By observing their parents using violence in their personal conflicts, they understand that violence is efficient in producing a desired result through intimidation (Spapens & Moors, 2020).

In their analysis of the Rochester Youth Development Study and the Rochester Intergenerational Study, Augustyn et al. (2017) analyzed data from 371 parent–child dyads in which G3 children were at least 17 years old. They found that gang membership in fathers increases the likelihood of gang membership in their sons, moderated by the extent of father-son contact in late childhood with the highest risk of son’s involvement in gangs for those who see their fathers for at least an hour per week and those who live with their fathers (Augustyn et al., 2017). The relationship between parental and son gang membership is mediated by a father’s maltreatment of his son, which accounts for nearly 40% of the total effect of intergenerational gang membership. For mothers and daughters, there was a direct effect of gang membership in G2 on gang membership in G3. There was no evidence of intergenerational continuity of gang membership for father—daughter and mother-son dyads. Following Bandura’s social learning theory (Bandura, 1977), the authors hypothesize that abusing one’s child may model that physical violence is acceptable to address conflict hence increasing a child’s risk for violence continuity. In addition, they suggest that child neglect promotes indirect transmission of violence as it models those traditional responsibilities of parenthood, that is, caring for a child, are not important.

In a meta-analysis of 63 effect sizes examining the association between growing up in a violent home and perpetrating IPV, the authors found that association effects were stronger for men. They hypothesize that children model the behavior of their same-sex parent; therefore, boys mimic their father’s behavior of inflicting violence while girls observe their mothers as victims (Stith et al., 2000).

CMO 3: Emotion Dysregulation

If a person has experienced CM or IPV throughout their lifespan, but particularly in early childhood, (C; G1), then they develop poor emotion regulation (M; G1), which may result in them perpetrating violence against their children or partners (O; G2).

Nine included studies supported this CMO (Bozzay et al., 2020; Choi et al., 2023; Domoney & Trevillion, 2021; Greene et al., 2020; Kim et al., 2009; Li et al., 2021; Martinez, 2020; Narayan et al., 2013; Reyes et al., 2015), three for the homotypical transmission of child maltreatment (Greene et al., 2020) and IPV (H. K. Kim et al., 2009; Martinez, 2020) and six for the heterotypical transmission of CM/IPV victimization (G1) to CM perpetration of G2 (Choi et al., 2023; Domoney & Trevillion, 2021), CM/IPV victimization (G1) and IPV perpetration (G2) (Narayan et al., 2013; Reyes et al., 2015) and family violence (G1) to youth violence (G2) (Bozzay et al., 2020; Li et al., 2021).

For example, Choi et al., 2023 examined the risk of inflicting child abuse among women who have experienced child abuse or IPV in adulthood using the Korean National Survey of Domestic Violence including n = 1,122 married women with children under the age of 18. The authors argue that the high levels of violence experienced in the mother generation, and in particular the IPV victimization, are harmful to the mother–child relationship due to women survivor’s difficulties in regulating their own emotions, which thereby increases the risk for abuse against their children. This highlights the importance of mental health support for women who have been victims of family violence (Choi et al., 2023).

In their analysis of the Safe Dates randomized trial with 2,343 adolescents in the United States, Reyes et al. (2015) investigated whether the relationship between witnessing interparental violence or experiencing child maltreatment and subsequent dating aggression was mediated by anger dysregulation or depression (Reyes et al., 2015). They found that anger dysregulation mediated the relationship between child maltreatment experience and dating aggression, but there was no evidence for an indirect effect of witnessing interparental violence through anger dysregulation on dating aggression. This is consistent with a body of literature that describes how violence experiences can interfere with emotional development and in particular emotion regulation (Jouriles et al., 2012). An inability to control emotions in relation to conflict can therefore increase the risk of dating aggression (Reyes et al., 2015). Depression was not a mediator for witnessing interparental violence nor experiencing child maltreatment on dating aggression in this study. Evidence on depression as a mediator between childhood violence experience and dating aggression is mixed. This supports the hypothesis that depression is more closely related to internalizing behaviors such as self-injurious behavior, rather than externalizing behaviors, such as dating aggression. Such a relationship could be gendered with men suffering from internalizing problems more likely to be violent toward women, and both men and women suffering from externalizing behavior more likely to engage in violence (Ellis et al., 2009; Yu et al., 2019).

Additionally, timing of experience of parental IPV seems to be critical in children’s expectations of social relationships and their ability to develop control of their negative behaviors and emotions as found in a longitudinal study of 168 low-income families in the United States. Experience of parental IPV in early childhood substantially increases the risk for dating violence in early adolescence, more so than the continuity or persistence of parental IPV experience through early and middle childhood (Narayan et al., 2013). The authors explain that this is because children are first learning how to regulate their emotions and control their behaviors in infancy and toddlerhood and thus have fewer positive experiences to guide how emotion regulation and interactions in relationships should look like.

Bozzay et al. (2020) investigated how personality traits may help explain relationships between witnessing family violence or experiencing child abuse and youth violent aggression in a sample of caregivers and adolescents (n = 237) in the United States. They found that witnessing family violence was linked with poorer impulse control and negative emotionality, defined as stress reaction and alienation, which increased levels of violent behavior. They also found that the experience of child abuse was linked with negative emotionality, which increased levels of violent behavior (Bozzay et al., 2020). The authors argue that children’s cognitive control processes are impacted by chronic and salient stressors such as family violence and that personality traits may be influenced by genes and their interaction with the environment. Similar findings were discussed by other authors who argued that witnessing parental IPV causes emotional distress, in the form of post-traumatic stress, anxiety, anger, and sadness, which then results in increased risk for violence perpetration (Martinez, 2020) and youth violence perpetration (Li et al., 2021).

In their analyses of data from the Oregon Youth Study and the Oregon Youth Couples Study, Kim et al. (2009) investigated the underlying mechanisms from parental emotion dysregulation via parental IPV to son’s IPV perpetration among 190 young men. They found that parental emotion dysregulation increased IPV, parental poor discipline and the son’s emotion dysregulation thereby increasing the son’s risk for IPV. There was an indirect effect of parent’s emotion dysregulation to son’s relationship conflict via son’s emotion dysregulation, and an indirect effect of parent’s emotion dysregulation to son’s emotion dysregulation through poor parental discipline (Kim et al., 2009). The authors therefore argue that emotion dysregulation is a major mechanism underlying the intergenerational transmission of relationship conflict. This is because people with poor emotion regulation rely on inappropriate strategies when with their partners and may also have poor self-regulatory abilities, which increase their risk for relationship conflict. As this study shows, emotion dysregulation may impair parenting and lead to inconsistent discipline practices, which further increase the risk of offspring emotion dysregulation (Kim et al., 2009).

CMO 4: High Vulnerability

If parents experience a multitude of risk factors such as unemployment, illicit substance use, alcohol use, and mental health problems, they are more likely to engage in violent behaviors (C; G1), then successive generations may also experience these risk factors (M; G2) resulting in these children perpetrating violence (O, G2).

This CMO was supported by four studies (Besemer et al., 2017; Frisell et al., 2011; Jaffee et al., 2013; van Dijk et al., 2019). Three focused on the homotypical continuation of violent crime (Besemer et al., 2017; Frisell et al., 2011; van Dijk et al., 2019) and one on the continuation of child maltreatment/IPV (G1) and subsequent child maltreatment (G2) (Jaffee et al., 2013). High vulnerability is best explained through the risk factor model where many risk factors relating to family dysfunction for example, unemployment within the child’s environment, poverty, household instability at the structural level intersect and co-occur with psychosocial risk factors (e.g., substance use, poor parental mental health) and exacerbate each other creating an intergenerational vulnerability which can manifest in intergenerational violence continuity (Farrington, 2011).

For example, in a qualitative in-depth analysis of police, justice department, and child welfare files of 25 organized crime offenders in Amsterdam and their children, van Dijk et al. (2019) found intergenerational transmission of violent crime to be particularly pertinent from father to son. The authors highlighted that intergenerational continuity is a problem in high-crime families due to multiple mechanisms, one of which is where men with criminal backgrounds found families with women of similar backgrounds, also called “assortive mating” (Farrington, 2011). As the father is largely absent from the family due to being in prison or his involvement in criminal activity, the mother is tasked as the sole primary caregiver in the family. However, due to her own involvement in criminal activity and experience of instability and stress throughout her development, a stable home environment is not provided to the children by father or mother. This instability is then passed on, compromising positive development outcomes for the next generation (van Dijk et al., 2019) in line with resilience theory (Fergus & Zimmerman, 2005).

In their analysis of convictions for violent crime among 12.5 million individuals covering 32 years in the Swedish Multi-Generation Register, Frisell et al. (2011) found strong evidence for intergenerational transmission from parent to child with an odds of 3.5 and equally strong for other first-degree relatives, and somewhat lower odds for second-and third- degree relatives (Frisell et al., 2011). Interestingly, children who were given up for adoption also had increased odds for violent offending OR 1.9, as did children who were adopted into the family OR 1.5. This suggests that shared environment plays an important part in intergenerational violence transmission, but that there must also be a genetic effect. The odds for intergenerational transmission of violent offending in this study were much higher for women, and particularly high for mother–daughter pairs OR 6.3 and grandmother–grand-daughter pairs OR 2.0. Families in high socio-economic strata had higher odds of intergenerational violence transmission OR 5.8, explained by the fact that violent crime more commonly occurs in lower socio-economic strata, hence, shared familial risk factors are particularly salient and family ties more important to influence criminal behaviour in higher socio-economic strata. The authors suggest that the gender effects are explained by female-specific environmental (e.g., anti-social behavior by the mother) and female-specific genetic effects. This is supported by a high degree of assortive mating found in this study (Farrington, 2002), with convicted violent offenders more likely to marry women who were convicted violent offenders OR 5.2 (Frisell et al., 2011).

In their meta-analysis of 23 studies investigating the intergenerational transmission of violent behavior, Besemer et al. (2017) engage with the theory of the criminogenic environment (Farrington, 2011). This stipulates that people who offend are more likely to experience a multitude of risks in other areas of their lives such as poverty, substance use, antisocial features, and ill-health while living in an unstable home environment. Their children may therefore “be entrapped in poverty, have disrupted family lives, may experience single and teenage parenting, and may live in the most deprived neighborhoods” (Farrington, 2010) (p. 205). Criminal behavior is therefore not directly transmitted across generations but through a “continuity of a constellation of antisocial and criminogenic features” (Besemer et al., 2017) (p. 163). Besemer et al. (2017) also found that the intergenerational transmission is stronger from mothers to their offspring compared to fathers. As most caregiving tasks traditionally have been carried out by mothers, maternal imprisonment is more disruptive for children because children are more likely to move to live with other relatives leading to instability in their lives (Besemer et al., 2017).

Jaffee et al. (2013) in their analysis of 1116 parent–child dyads from the Environmental Risk Twin Study (E-Risk) in the UK found that mothers who had experienced childhood maltreatment were distinguished from those who had not by several factors. These factors included socioeconomic disadvantage, poor mental health, poor partner’s mental health, domestic partner violence, and low social support. These same factors also applied to their children who were physically maltreated. Intergenerational violence cycle maintainers reported higher levels of substance abuse, depression, social disadvantage, antisocial behavior, and domestic partner violence (Jaffee et al., 2013).

CMO 5: Impaired Relationships

If parents experienced IPV or child maltreatment (C; G1), then they struggle to form safe and nurturing relationships with others including their partners and children (M; G2) resulting in intergenerational transmission of violence perpetration (O, G2).

Six studies investigated impaired relationships as underlying mechanisms for intergenerational transmission of violence perpetration (Adshead & Bluglass, 2001; Augustyn et al., 2017; Chapple, 2003; Jaffee et al., 2013; Schofield et al., 2013; Thornberry et al., 2013). All studies focused on the homotypical transmission of IPV, child maltreatment, or gang membership.

Schofield et al. (2013) conducted a meta-analysis of five studies focused on the intergenerational transmission of maltreatment and potential moderators. They investigated the role of safe, stable, and nurturing relationships in line with interactional theory, which posits that those who have experienced or engaged in violence will struggle forming positive relationships in adulthood (Thornberry et al., 2013). They found a strong association between child maltreatment in G1 and subsequent maltreatment of G2. However, they then went on to identify safe, stable, and nurturing relationships as protective moderators on the intergenerational continuity of maltreatment despite different measures and conceptualizations of safe, stable, and nurturing relationships in the included studies. For example, in their analysis of 1,116 families in the Environmental Risk (E-Risk) Longitudinal Twin Study in the United Kingdom, Jaffee et al. (2013) found substantial continuity of abuse from G1 to G2. The odds for children of mildly and severely maltreated mothers to experience maltreatment were 3.55 times and 5.31 times those of non-maltreated mothers. For families that maintained the intergenerational cycle of violence, mothers reported fewer warm relationships with their children and less trusting relationships with their partners.

Similarly, in their study of 980 grade 9 to 11 students in Arkansas, Chapple (2003) investigated social bonding/social control theory among boys and girls which states that inherently all of humanity has natural urges to act in aggressive and selfish ways and only attachment and bonds formed with pro-social values, people and institutions prevents acts of violence and aggression (Hirschi, 1969). The study found that adolescents who had witnessed parental violence reported lower mean levels of parental attachment and monitoring and higher mean levels of dating violence, and this was particularly true for girls. The study also found that parental attachment did not moderate the relationship between witnessing interparental violence and dating violence. Parental monitoring was however significantly associated with a reduction in violent offending in G2. Chapple argues that strong parental attachment and monitoring discourage dating violence perpetration, while weak parental monitoring and poor attachment are underlying mechanisms for intergenerational violence transmission of violence and aggression perpetration. This was however disputed in another study, which found that attachment and inconsistent discipline did not act as mediators of intergenerational gang membership between fathers and sons nor mothers and daughters, although mothers involved in gangs were more likely to parent their daughters inconsistently (Augustyn et al., 2017).