The incidence of parent abuse among young people on the edge of care

At referral 54 per cent (112) of parents (or parental figures) reported that the young people had been violent to them in the past six months. The majority of those reporting this violence were mothers (82 per cent), but fathers (8 per cent), other relatives (4 per cent) and adoptive parents (5 per cent) also reported violent assaults by the young people they were parenting. Earlier studies have also found that mothers are more likely to be victims of parent abuse than fathers (Cornell and Gelles, Reference Cornell and Gelles1982; Agnew and Huguley, Reference Agnew and Huguley1989; Paulson et al., Reference Paulson, Coombs and Landsverk1990; Browne and Hamilton, Reference Browne and Hamilton1998). This has led some researchers to argue that parent abuse should more properly be viewed as mother abuse, given the preponderance of mothers, mainly lone mothers, in their samples (Hunter et al., Reference Hunter, Nixon and Parr2010).

The incidence of parent abuse in our sample was much higher than that reported in international studies. Two North American studies estimated that the annual incidence of parent abuse ranged from 5 per cent to 10 per cent (Agnew and Huguley, Reference Agnew and Huguley1989; Cornell and Gelles, Reference Cornell and Gelles1982), while another found that nearly 14 per cent of a representative sample of 15–16 year olds had physically assaulted their mothers in the previous six months (Pagani et al., Reference Pagani, Tremblay, Nagin, Zoccolillo, Vitaro and McDuff2004). The only large UK study of parent abuse reported that 8.5 per cent of a sample of 469 university students had been violent towards their mothers and 6.1 per cent towards their fathers in the past year (Browne and Hamilton, Reference Browne and Hamilton1998). Although the sample were all students and therefore likely to be middle class, it may nevertheless be reasonably representative, since evidence from the USA suggests that parent abuse is not related to social and economic status (Cornell and Gelles, Reference Cornell and Gelles1982; Agnew and Huguley, Reference Agnew and Huguley1989; Paulson et al., Reference Paulson, Coombs and Landsverk1990). Two UK studies of over 80,000 calls to a telephone helpline similarly reported that 8 per cent of these concerned children's physical aggression to parents (Parentline Plus, 2008, 2010).

Given the limited availability of voluntary support services for families with adolescents, the higher incidence of parent abuse in our own sample is likely to be due to the seriousness of the difficulties of those adolescents who meet the threshold for the provision of family support services. Although the Children Act 1989 envisaged that family support services would be provided to ‘children in need’, in practice open access family support services, such as children's centres, are often targeted at families with younger children. Thresholds for the provision of social work support to families who request help with difficult teenagers, rather than being referred for help with parenting, are very high. Families who do receive such support are often those with very serious problems (Sinclair et al., Reference Sinclair, Garnett and Berridge1995; Triseliotis et al., Reference Triseliotis, Borland, Hill and Lambert1995; Biehal, Reference Biehal2005a).

The nature of the abuse: parent and child accounts

Interviews with parents and young people revealed an astonishing catalogue of violent behaviour from children, as well as verbal aggression and general bullying. Some directed their violence solely at one family member, but others appeared to lash out at everyone around them. Violence was sometimes accompanied by general destructiveness, with one in ten parents referring to their children ‘smashing up the house’, ‘kicking in the door’ or ‘smashing up his room’. Most incidents described involved quite serious assaults, for example:

Severe violence was also directed at siblings, as one carer described:

In a few cases, the aggression appeared to be linked to young people's drug or alcohol abuse:

However, some parents acknowledged that the violence could be two-way:

The voices of young people who are violent to their parents have rarely been heard in research reports. Many of the young people interviewed accepted that their own behaviour was indeed problematic. Some blamed themselves for the difficulties they were experiencing in their families, describing themselves as ‘bad’. A twelve-year-old girl, who said she had been ‘bad’ for the past three years, explained:

Others described their violence as being rooted in their conflicts with parents

Some explained that they were ‘misbehaving’ because they were unhappy, or that they felt ‘angry inside’. A few considered that the roots of their aggression lay in their past experience of abuse, for example:

Others linked it to their current experience of abuse:

It was clear from the accounts of both young people and parents that some considered there was a link between past domestic violence and their own violence, for example:

A few young people also juxtaposed their descriptions of their own difficult behaviour, including violence, with accounts of witnessing domestic violence:

These accounts reveal the often serious nature of parent abuse and provide some insight into the ways that young people and parents explain it. Their accounts indicate that it may occur against a backdrop of parental conflict, domestic violence and child maltreatment. In the rest of this article, we will draw on quantitative data to investigate the histories and circumstances of these young people and families and examine the relationship of parent abuse to these and other factors.

The young people

As other studies with larger, more representative samples have found, young people who were violent to parents were as likely to be female as male (Paulson et al., Reference Paulson, Coombs and Landsverk1990; Browne and Hamilton, Reference Browne and Hamilton1998; Pagani et al., Reference Pagani, Tremblay, Nagin, Zoccolillo, Vitaro and McDuff2004; Parentline Plus, 2010). Just over half (52 per cent) of those violent to parents were girls. Within this sample of eleven- to sixteen-year olds, the likelihood that they would be violent to a parent was not related to age.Footnote ¹

In 63 per cent of cases, the violence went beyond the family, as the young people were also violent to others. This violence was mostly directed at other children, but parents also reported assaults on adults, including teachers. For nearly two-thirds of the young people, therefore, parent abuse was one aspect of a broader pattern of violent behaviour.

Parent abuse took place in the context of a range of other difficulties, including learning difficulties (13 per cent) and, for 12 per cent, attention deficit hyperactivity disorder (ADHD), although it is unclear whether a professional diagnosis had been made in all of these cases. Many parents also had serious concerns about young people's behaviour outside the home, including concerns about their involvement in offending (25 per cent), their friends (52 per cent) and their behaviour at school (76 per cent). Two-thirds (69 per cent) had truanted, 42 per cent had been excluded from school and half had run away from home during the previous six months. However, 65 per cent of the young people said that they were worried about their parents. One quarter of them had previously been looked after on at least one occasion, an indication of previous child and family difficulties.

Around one-fifth (21 per cent) of the parents reported that their children's behavioural difficulties had emerged before they were five years old and had persisted since then. Research has indicated that early onset persistent (EOP) behavioural difficulties may be related to temperamental hyperactivity in early childhood and this may interact with persistent adversity in the child's environment to result in continuing behavioural difficulties (Barker et al., Reference Barker, Oliver and Maughan2010). A Canadian study found that EOP behavioural difficulties increased the odds of child violence to mothers at the age of fifteen to sixteen years by 6.58. Where high levels of physical aggression were observed in early childhood but declined over time, the odds of parent abuse at the age of fifteen to sixteen were lower (3.77), but nevertheless nearly four times higher than the odds for children with low physical aggression in their pre-school years (Pagani et al., Reference Pagani, Tremblay, Nagin, Zoccolillo, Vitaro and McDuff2004).

The family context

Most (82 per cent) of the young people who were violent to parents were living with their mothers and only 24 per cent lived with their fathers. Half of them lived with lone mothers, as shown in Table 1.

Table 1

Young people's family circumstances

Among the total study sample, lone mothers were significantly more likely to report violence from their children (65 per cent) than other parents (48 per cent).Footnote ² Although mothers (alone or with partners) were the most common victims of parent abuse, this association was not statistically significant and was undoubtedly due to the fact that they constituted the vast majority of respondents.

Nearly half (46 per cent) of the parents abused by their children reported continuing conflicts with current or former partners, including many of those who were lone parents. Research has shown that children's conduct problems may be associated with a history of family conflict, and that this conflict may be more damaging than parental separation per se (Rutter et al., Reference Rutter, Giller and Hagell1998; Utting et al., Reference Utting, Monteiro and Ghate2007). It therefore seems likely that the family conflict experienced by many of the young people who abused their parents may have contributed to their own abusive behaviour.

Half of the families in which parent abuse occurred had a history of domestic violence. Young people who were violent to parents were nearly three times more likely to have witnessed partner violence (28 per cent) than those who were not violent to parents (10 per cent).Footnote ³ Among young people violent to parents, those who had witnessed partner violence were more than twice as likely to be currently living with a lone mother (41 per cent) than with both parents (17 per cent) or a parent and new partner (16 per cent).Footnote ⁴ Other studies have found a similar relationship between witnessing domestic violence and parent abuse (Kennair and Mellor, Reference Kennair and Mellor2007). Witnessing partner violence has been found to have negative effects on children's social-emotional development, including their ability to manage emotions during times of stress or anger, and may increase the risk of behavioural problems (Carpenter and Stacks, Reference Carpenter and Stacks2009; Meltzer et al., Reference Meltzer, Vostanis, Doos, Ford and Goodman2009; Munro, Reference Munro2011). The increased likelihood that young people living with lone mothers would have witnessed partner violence may therefore be a partial explanation of why lone mothers were at greater risk of parent abuse.

Many studies have shown that parenting style may also have an effect on child behaviour, including both harsh parenting and weak and ineffective parenting, which fails to set effective boundaries to children's behaviour (Maccoby and Martin, Reference Maccoby, Martin, Mussen and Hetherington1983; Larzelere and Patterson, Reference Larzelere and Patterson1990). One-fifth of parents who experienced abuse were reported by social workers and family support workers to be ‘hardly ever’ warm to the young person and 17 per cent were reported to be harsh ‘most of the time’.

Professionals considered that 44 per cent of the parents ‘hardly ever’ set boundaries to their children's behaviour and there was little difference in this respect between parents reported to be harsh and those who were not. Interviews with parents indicated that many were struggling, unsuccessfully, to set boundaries. They often lacked confidence and expressed a sense of helplessness and despair about their ability to set limits to their children's behaviour. Some found it hard to cope because they were depressed, or parenting alone, feeling at a loss as to how to change their children's behaviour within and outside the family. Other studies of parent abuse have suggested that parent abuse is more likely where parents are inconsistent and overly permissive (Agnew and Huguley, Reference Agnew and Huguley1989; Paulson et al., Reference Paulson, Coombs and Landsverk1990) and one found that weak parental supervision at age ten to twelve significantly increased the odds of physical aggression towards mothers during adolescence (Pagani et al., Reference Pagani, Tremblay, Nagin, Zoccolillo, Vitaro and McDuff2004).

Problems of substance abuse (in most cases alcohol abuse) were reported by 17 per cent of parents who experienced abuse, and, for this group, substance use appeared to contribute to their difficulties in setting boundaries. Professionals reported that 81 per cent of parents with drug or alcohol problems ‘hardly ever’ managed to set boundaries to their children's behaviour, compared to 37 per cent of those without reported problems of substance use.Footnote ⁵ These findings are broadly consistent with those of the above-mentioned Canadian study, which reported that a significant association between parental substance abuse and parent abuse disappeared once parental supervision was taken into account (Pagani et al., Reference Pagani, Tremblay, Nagin, Zoccolillo, Vitaro and McDuff2004).

Abuse and neglect

Social workers indicated that 32 per cent of young people who abused parents had experienced past maltreatment, including 15 per cent who had been physically abused, 9 per cent sexually abused, 5 per cent emotionally abused and 8 per cent who had experienced neglect. There were also concerns about current maltreatment in relation to half of those with past experience of maltreatment. In most cases, these concerns centred on neglect (14 per cent) and emotional abuse (31 per cent), although in a few cases (6 per cent) there was concern about physical abuse. Nearly half (47 per cent) of those who had experienced maltreatment also had experience of domestic violence, an association that has been found in previous studies (Hamby et al., Reference Hamby, Finkelhor, Turner and Ormrod2010).Footnote ⁶

Histories of maltreatment were equally common among the young people in our total sample who had not abused parents. However, studies with population-based samples of young people have found an association between child maltreatment and parent abuse. Browne and Hamilton (Reference Browne and Hamilton1998) reported that students who were violent to parents were nearly three times more likely to have been physically or sexually abused by a family member than those who were not violent. Serious violence was also more commonly found among those who had been physically abused, compared to those not physically maltreated (Browne and Hamilton, Reference Browne and Hamilton1998). The international evidence also suggests that there is an association between this and past or current physical abuse of the child (Brezina, Reference Brezina1999; Pagani et al., Reference Pagani, Tremblay, Nagin, Zoccolillo, Vitaro and McDuff2004). Consistent with this, a study of a nationally representative sample of 1,777 children in the USA found that those who experienced chronic maltreatment had significantly more behavioural problems than those who did not (Jaffee and Kohn Maikovich-Fong, Reference Jaffee and Maikovich-Fong2011). Children, who experience maltreatment, including physical violence, may learn to use violence as a way of dealing with anger and frustration (Patterson, Reference Patterson1982).

Mental health difficulties

Given the histories of family conflict, violence and child maltreatment described above, it is perhaps not surprising that many of the young people and parents had mental health difficulties. There were no significant differences in this respect between those families which reported parent abuse and those which did not. The majority (76 per cent) of parents abused by their children scored over the clinical threshold on the GHQ-12, a standardised measure of mental health which measures anxiety and depression (Goldberg and Williams, Reference Goldberg and Williams1988). It is difficult to know how far parent mental health problems contributed to, or were a consequence of, the child's violence, but it is possible that the relationship between the two was bi-directional, with each reinforcing the other.

Many of the young people who abused parents had emotional and behavioural difficulties, as identified by a parent-completed screening measure of child mental health, the SDQ (Goodman, Reference Goodman1997). This provides scores for five domains, which are summed to give a total score. Table 2 compares their scores to those for eleven- to fifteen-year olds in a representative sample of children and young people in the community (Meltzer et al., Reference Meltzer, Gatward, Goodman and Ford2000). It shows that these young people were more than six times as likely to have scores over the clinical threshold for total difficulties, and many had sub-scale scores indicating clinical problems in specific domains.

Table 2

Per cent with clinically significant scores for emotional and behavioural difficulties on the SDQ

Note: *No data are available for peer problems and pro-social behaviour.

Young people who were violent to parents were five times more likely to have conduct problems and over ten times more likely to have emotional problems than others of a similar age in the wider community. They were also far more likely to show symptoms of hyperactivity, and three-quarters of them had serious problems in their relationships with peers.

Child and parent mental health difficulties were closely associated, as 89 per cent of the young people with clinically significant mental health difficulties (on the SDQ) had parents with GHQ scores indicating that they too had mental health difficulties.Footnote ⁷ As noted earlier, one fifth of parents had first been concerned about their children before they were five years old. These young people had significantly higher SDQ scores both for total difficulties and also on the peer problems sub-scale compared to those whose difficulties apparently emerged later.Footnote ⁸

The incidence of self-harm was also very high, as 23 per cent of the young people who were violent to parents had self-harmed in past year, compared to a rate of only 1.2 per cent for eleven- to fifteen-year olds in the general population. Self-harm has been found to be more common among adolescents in families with a high degree of family discord or where parents have high GHQ scores, both of which were true for many of the young people who abused parents (Meltzer, Reference Meltzer2001).

Explaining parent abuse

There were only two significant differences between families in which parent abuse occurred and those which did not report it. Lone mothers were significantly more likely to be the victims of parent abuse than other parents, and young people who had witnessed partner violence were three times more likely to abuse their parents than those who had not witnessed this. Nearly two-thirds (64 per cent) of young people who were violent to parents had experienced either domestic violence, abuse or both. As previous research has shown, both maltreatment and domestic violence may increase the risk that children will develop conduct disorder or emotional disorders, which may make the emergence of child violence more likely (Trickett and McBride-Chang, Reference Trickett and McBride-Chang1995; Hester et al., Reference Hester, Pearson and Harwin2007; Jaffee and Kohn Maikovich-Fong, Reference Jaffee and Maikovich-Fong2011; Munro, Reference Munro2011).

We used multivariate analysis to investigate the factors which predicted parent abuse, once all relevant variables had been taken into account. This indicated that violence to parents was predicted by witnessing partner violence, higher scores for emotional problems (on the SDQ) and higher scores for parents’ mental health difficulties (on the GHQ). There was also an association with SDQ scores for conduct problems, but this did not quite reach significance (p = 0.054). This model explained 24 per cent of the variance in relation to parent abuse within the total sample. The regression was run again substituting ‘lone mother’ for ‘witnessed partner violence’, as these two variables appeared to cancel each other out when entered together due to the close association between them. In this second model, parent mental health difficulties, children's emotional and conduct problems and living with a lone mother also explained 24 per cent of the variance in parent abuse.Footnote ⁹ These results suggest that young people were more likely to be violent to parents if they had witnessed partner violence, were cared for by lone mothers − particularly those suffering from anxiety or depression − and they themselves had mental health difficulties.

The gender issues here are complex. First, this study, like others before it, found that girls were as likely to be violent to parents as boys. Second, in this sample mothers were not significantly more likely to be victims of this abuse than fathers. However, lone mothers were more likely to experience violence from their children, and their children were more likely than others to have witnessed partner violence. It is possible that in these cases there was an indirect chain effect, with domestic violence increasing the risk that the young person would become violent, leading to family breakdown and leaving mothers, many of whom were experiencing anxiety and depression, to cope with their children's anger and violence alone. Not surprisingly, in this context many of them found it difficult to set effective boundaries to their children's behaviour.

Supporting families experiencing parent abuse

All young people and parents in the study received a whole-family support intervention, which was either a short-term intensive service provided by an adolescent support team or a longer-term, less intensive mainstream social work service (see Biehal, Reference Biehal2005a for details). Both services took an ecological approach, addressing difficulties in different domains of the young people's lives: at the level of the individual child, the family, the peer group, the school and the wider community (Bronfenbrenner, Reference Bronfenbrenner1979). Many parents found the support and advice with parenting provided extremely helpful, and many of the young people welcomed help to change the behaviour that was causing them, as well as others, so much distress. At six-month follow-up, 63 per cent of parents who had reported parent abuse at baseline reported that this was now a less serious problem (or no longer a problem at all), irrespective of which of the two services they received. Positive change was also reported in relation to other child and parent difficulties (Biehal, Reference Biehal2005b). The whole family, ecological approach of these services therefore appeared to contribute to positive change for many families, although there were also other mediators of change (see Biehal, Reference Biehal2008).

Limitations to the study

It is important to note that these factors predicted the likelihood of parent abuse for a sample in which virtually all young people had serious behaviour problems. Further research is needed with a sample representative of the wider population to determine whether other factors might also be identified which would distinguish young people who abuse parents from those who do not.