Contents

• Introduction
• Australian data
  • Incidence
  • Australian research
• Single parent families
  • Physical abuse
  • Neglect
  • Sexual abuse
  • Poverty
  • Linking poverty and family structure
• Stepfamilies
  • Physical abuse, emotional abuse and neglect
  • Child homicides
  • Sexual abuse
• Prevention
  • Family support programs
  • An expanded role for schools in prevention
• Conclusion
• References

Introduction

In the last 25 years significant changes in family structure have been experienced in countries around the world, particularly in the developed nations (Clulow 1993). Official statistics show that Australia, the United States and most European countries have experienced higher rates of divorce, falling marriage rates, and an increase in the number of couples living together outside of marriage (Edgar 1988, Clulow 1993).

There has been considerable public concern about the changing shape of the family, particularly the effects that changes in family structure may have on parenting and children's behavior (Tower 1989). Edgar (1991) contended that the place of children in the family, and the meaning and roles of parenthood have become increasingly complex and problematic. Children are now significantly more likely to be raised by defacto couples, single parenthood is on the increase and large numbers of children live in reconstituted or blended families [Note 1] (Corby 1993). These changes to family structure have taken place at a time where there are also increasing pressures on families. All of these combine to make society's expectations of adequate family functioning even more difficult to achieve (Tower 1989).

A large body of research has been produced on the causes of child maltreatment (Ammerman and Hersen 1990). Since the 1970s there has been increased recognition of the importance of the socio-cultural context of child maltreatment (U.S. National Research Council 1993), with researchers investigating the effects of parent, child and environmental factors on child maltreatment. Marital status and/or family structure have been frequently investigated as social factors which may have a bearing on child maltreatment (Daro 1988), particularly in the context of a common finding that non-nuclear families are over-represented in child maltreatment cases (Tomison 1994).

While investigations of single variables like family structure have contributed to the identification of key variables or 'risk factors' associated with child maltreatment, they have neither produced a causal explanation of child maltreatment, nor enabled the identification of causal relationships between associated factors (Browne 1988). Overall, it is generally acknowledged that no single factor can fully explain why maltreatment occurs (Browne 1988, National Research Council 1993).

Although single factor approaches are limited in this way there is still some value in providing a brief overview of the available research investigating the relationship between child maltreatment and family structure. Single parent and blended families merit particular attention in this context. Gaps in the research knowledge which may shed light on future directions for research can then be highlighted.

Australian data

Incidence

In the most recent available Australian national statistics on children and families, (Australian Bureau of Statistics 1995), it was reported that in 1992 approximately 81 per cent of children under the age of 15 years resided with their natural parents, 4 per cent resided in stepfamilies (one natural parent and a married or defacto partner), 14 per cent with a single parent, and less than 1 per cent resided in some other type of household (e.g. with extended family members).

In 1992-93 there were 23,199 substantiated cases of child abuse or neglect for children aged 18 years or under (Angus and Zabar 1995). Because the child maltreatment statistics were based upon children aged under 18 years, while the ABS statistics were based on dependent children under 15 years, it was not possible to make a direct comparison between the ABS and child maltreatment statistics. However, from the child maltreatment data it was apparent that stepparents or defacto parents were over-represented as maltreaters, especially in cases of physical, sexual and emotional abuse. In neglect cases, where the vast majority of abusers were natural parents, step and defacto parents were under-represented.

The latest national Australian statistics on child abuse and neglect 1993-1994 (Angus and Woodward 1995) revealed similar trends. It was reported that 12 per cent of substantiated reports of maltreatment involved a perpetrator who was a stepparent, defacto parent, foster parent or guardian. The over-representation of stepparents or defacto parents as perpetrators was clearest in cases of sexual abuse where 14 per cent of cases were reported to involve a stepparent or defacto partner.

However, the national child maltreatment statistics suffer from a number of limitations and should thus be treated with caution. First, it should be noted that the data constitute a conservative estimate of the extent of child maltreatment across Australia. They do not include any cases of child maltreatment not reported to the state/territory child protection units, or cases which were not substantiated. In addition, there are significant gaps in the data because of state and territory differences in collecting and managing case statistics.

Finally, Angus and Zabar (1995) excluded a substantial number of cases because the relationship of the child to the maltreater was not recorded, or was not known.

Second, the figures may be an under-representation of the extent to which biological parents are involved as the perpetrators of maltreatment. For example, a Victorian examination of the community's attitudes to child sexual abuse (Wallis 1992), concluded from the responses to hypothetical scenarios, that the closer the relationship between a sexually abused child and a perpetrator, the less likely they would be reported by non-offending family members. Only 47 per cent of parents stated that they believed a report should be made to the 'authorities' if a woman discovered that her husband was sexually abusing her child. The most common reaction was that the relationship with the perpetrator should be terminated (54 per cent). Of those people who advocated termination of the relationship, only a minority of 4 per cent indicated that they would report the abuse to a professional source.

In addition, the data may be confounded by a labelling bias which affects official statistics and clinical reports (Gelles 1975, as cited in Gelles and Harrop 1991). Medical professionals may expect that non-biological parents are more likely to maltreat children in their care, and thus, injured children with a non-biological parent may be more likely to be diagnosed and reported as abused.

Australian Research

While a number of Australian studies have considered the effects of family structure on child maltreatment, most merely refer to structure as part of the family demographic information, noting the over-representation of non-nuclear families in their samples (Hiller, Goddard and Diemer 1991, Goddard and Hiller 1992, Tomison 1994). However, many Australian and international experimental studies of professional decision making in suspected child maltreatment cases (e.g. Dalgleish and Drew 1989, Tomison 1994), routinely employ marital status or family structure as a 'risk factor', which in combination with other factors, may enable the prediction of families more likely to maltreat their children.

Overall, a large body of research, predominantly from the United States, has been directed towards determining who is most likely to maltreat children (Corby 1993), the objective being, through such identification to facilitate early intervention and an increased probability of protecting the child. Yet it appears that there has been no Australian or international study which has conducted a truly comprehensive investigation of the relationship between family type, particularly non-nuclear families, and child maltreatment.

Much of the research that has been done on parental factors has focused on mothers, predominantly those from low socioeconomic backgrounds (Ammerman 1989, Fantuzzo and Twentyman 1986 as cited in Ammerman and Hersen 1990, Wolfe 1987, as cited in Ammerman and Hersen 1990, National Research Council 1993), thus limiting understanding of the role of other parental characteristics in maltreatment. There has been a relative lack of emphasis on the role of fathers, stepfathers, or other family members, except in the case of sexual abuse (Corby 1993). Thus, attempts to assess relationships between family structure and child maltreatment have been somewhat limited.

In the remainder of this paper, the literature on family structure is discussed as it applies to the two most prevalent forms of non-nuclear families, single parent, and blended families.
[ Note 2].

Single parent families

Creighton and Noyes (1989) found that 25 per cent of children registered as physically abused in England and Wales for the years 1983 to 1987 resided in single parent families. Approximately half of the children registered for neglect or failure to thrive, and a third of those registered for emotional abuse also resided with a single parent. Sack, Mason and Higgins (1985) found the prevalence of physically abusive punishment to be twice as high in single parent families as in two parent households.

Physical Abuse

Gelles (1989) in contrast, using the data from two U.S. national incidence studies of family violence, found that single parent families were not more likely to use physical violence overall, but that single parent households (males and females) were more likely than two-parent households to use severe violence (high probability of injury); this was particularly the case in single father households. Overall however, poor, young, single mothers with young children were most likely to report that they physically abuse their children (Gelles 1992).

It should be noted that the majority of studies assessing the relationship between single parent status and child maltreatment generally assume that the single parent is a mother (Gelles 1989). Because the number of families headed by single fathers is small, (e.g. 1.5 per cent of Australian children in 1992 lived in single father families [ABS 1995]), very little is known about their child maltreatment risk, or the factors that lead them to physical violence (Gelles 1989). Accordingly, the remainder of this paper will refer only to evidence from single mother families.

Neglect

Child neglect is commonly associated with low income, poor housing and living conditions, low educational and employment levels, and larger, multi-problem families who are in receipt of government benefits (Boehm 1964; Daro 1988). The neglecting household is often characterised by 'a shifting constellation of adult and child figures, representing at times desperate efforts by the parent to keep the family together during times of economic and other social crisis' (National Research Council 1993:127). Mothers are implicated as the responsible parent in the majority of neglect cases, and are usually assumed to be solely responsible for physical abuse and neglect in single-parent families (Corby 1993).

Single parent families are frequently implicated in the more severe forms of neglect. 'Chronic' neglect cases often fit a stereotype characterised by single motherhood and a 'chaotic and unpredictable character' (National Research Council 1993), long-term involvement with family support or child protection services (Nelson, Saunders and Landsman 1993; Tomison 1994), and a lack of cognitive stimulation and emotional nurturance for the child (Polansky, Gaudin and Kilpatrick 1992).

Sexual abuse

Significantly, Creighton and Noyes (1989) found that 20 per cent of children registered for sexual abuse came from female single parent households. Finkelhor and Baron (1986) argue that children living with single mothers may be exposed to greater numbers of adult males than those in two-parent households, which in turn may place them statistically at greater risk of being sexually abused. In addition, paedophiles often admit seeking out children who are emotionally deprived (Finkelhor 1984) or lacking a father figure in order to groom them for sexual abuse. An issue to be investigated concerns the actual extent to which there is a proportion of sexual abuse cases in single mother families that involve paedophile offenders who have specifically targeted vulnerable families.

Finally, the literature is presented on the relationship between family structure, single parent families in particular, and the factor with which it is commonly linked, poverty (National Research Council 1993).

Poverty

Pilger (1989) contended that one in five Australian children born in 1988 faced the prospect of long-term poverty, and that a higher proportion of Australian children live in poverty than do their counterparts in Britain, Germany, Canada, Sweden, Norway and Switzerland (Pilger 1989). Discussions of the relationship between poverty and child maltreatment began in the 1960s (Steele and Pollack 1968), gaining impetus with the development of models of child maltreatment based on the interaction of multiple factors (interactionist approach).

Child maltreatment is disproportionately reported among poor families and particularly in the case of neglect, is concentrated among the poorest of the poor (Wolock and Horowitz 1984). The professional debate as to 'whether this association results from poverty-related conditions that precipitate abuse or from greater scrutiny by public agencies that results in over-reporting' (National Research Council 1993:133), continues without resolution.

Daro (1988) contended that empirical evidence for the unique role of poverty as a causal factor for child maltreatment and developmental delays continues to grow. But despite the intuitive strength of poverty as a causal factor in child maltreatment, the reality is that not all impoverished parents maltreat their children. As with all single factor explanations of child maltreatment, it does not seem useful to focus solely on poverty to the exclusion of other environmental or intra-familial factors, such as family structure, class or homelessness.

Linking poverty and family structure

Gil (1970) considered socioeconomic problems to be a major cause of maltreatment. Unemployment in particular, was reported to be a significant influence on family violence in the United States (Krugman 1986), though in the United Kingdom this may be true for the long-term unemployed only (Taitz 1987, as cited in Browne and Saqi 1988).

In a U.S. study, Coulton, Korbin, Su and Chow (1995) noted that poverty, unemployment, racial segregation, abandoned housing, population loss (significant decreases in an area's population over a decade), and female-headed households were all associated with neighborhoods with a higher incidence of maltreatment.

Using U.S. National Center for Health Statistics data for 1976, Bachrach (1983) investigated differences in the socioeconomic characteristics of biological, step- and adopted children. She concluded that families headed by single mothers (those who were never married and those with absent spouses) were the most economically disadvantaged group. She reported that two-thirds of mothers who were never married had incomes under the poverty line. In contrast, there were no significant differences between stepfamilies (stepfather present) and traditional nuclear families with respect to family income.

In an investigation of the effects of community influences on child maltreatment, Coulton, Korbin, Su and Chow reported that '[family] structure did not emerge as a separate dimension . . . largely because of the extremely strong relation between an area's poverty status and its proportion of female-headed families' (Coulton, Korbin, Su and Chow 1995:1273).

Thus, the inextricable links between single parent families, especially those headed by a single mother, and poverty, make it difficult to determine the separate contributions of sole parenthood and poverty to child maltreatment (National Research Council 1993, Tomison 1994, Coulton, Korbin, Su and Chow 1995).

Gelles' (1989) investigation of violence and family structure suggested that the high rate of physical abuse committed by single mothers appeared to be a function of the 'poverty that characterizes mother-only families' (Gelles 1989:499). Gelles contended that poverty and the stresses associated with it clearly placed children at greater risk of physical violence by their single mothers. Regretfully, his cross-sectional methodology did not allow for a determination of whether low income preceded or followed a person becoming a single mother. Thus, it was not possible to conclude whether or not income mediates or moderates the relationship between family structure and violence.

Unfortunately, therefore, the analysis of social factors has not revealed the processes through which poverty, family instability and high concentrations of impoverished families combine to produce high rates of maltreatment. Further research is needed to explore the inter-relationship between these variables (Cicchetti and Lynch 1993 as cited in Coulton, Korbin, Su and Chow 1995).

Stepfamilies

Though it is widely assumed that children are at greater risk of maltreatment in stepfamilies, there are few studies that have actually tested this premise, or that investigate the relationship between stepfamilies and maltreatment. The U.S. National Research Council (1993) noted that children who have had a stepfather are at greater risk of abuse (National Research Council 1993). In contrast, one of the few studies to have attempted to explore the issue in any depth, Giles-Sims and Finkelhor (1984) concluded that the available data was inadequate to determine the nature of the relationship between stepfamily status and child maltreatment.

Gelles and Harrop (1991), using a representative U.S. sample generated by the Second National Family Violence Survey, reported no significant differences between biological and non-biological parents in the rates of severe (high probability of causing an injury) and very severe (higher probability of an injury) physical violence towards children. As mentioned previously, they contended that the data on the elevated risks of child maltreatment by stepparents may be criticised on methodological grounds, given that the data derives entirely from official reports of child maltreatment or other clinical data. Such data may be confounded by professional labelling biases (Gelles 1975, as cited in Gelles and Harrop 1991).

In addition, Gelles and Harrop contend that the relationship between risk and stepparents may be confounded by a number of factors. First, labelling bias (Gelles 1975 as cited in Gelles and Harrop 1991); clinicians may expect that non-genetic parents are more likely to maltreat a child, and thus injuries to children with a stepparent may be more likely to be diagnosed and reported as abuse.

Second, it is known that both divorce rates and violence are highest among lower socioeconomic groups. Therefore the over-representation of reports involving non-genetic parents may be a function of income, and not family structure. However, as mentioned previously, Bachrach (1983) produced data which disputes this contention, finding no significant differences in income between families with a stepparent and biological families.

Giles-Sims and Finkelhor (1984) suggested that the disproportionate number of children abused in stepfamilies may occur because adults generally predisposed to violence may be more likely to experience a marital breakup, and subsequently participate in a stepfamily. In support of Giles-Sims and Finkelhor, a re-analysis of data from a large-scale tracking of suspected child abuse and neglect cases within a Victorian regional child protection network (Tomison 1994), indicated that physically violent spouses were more likely to be reported by professionals in stepfamilies than in biological families.

However, as mentioned previously, the sample may be an under-representation of the extent to which biological parents are involved as the perpetrators of maltreatment or other violence, given that non-offending family members may be more likely to report maltreaters where no biological relationship exists (Wallis 1992).

Physical abuse, emotional abuse and neglect

There is little information available on the relationship between step-parenting, physical abuse, emotional abuse or neglect. Creighton and Noyes (1989) found that in 32 per cent of physical abuse cases the child resided with one natural parent (usually the mother) and one substitute parent (usually male). Blended families were identified in 15 per cent of neglect cases, 11 per cent of failure to thrive cases, and 36 per cent of emotional abuse cases, and were thus clearly over-represented in the child abuse population. As mentioned above, this over-representation is commonly reported among families suspected of child maltreatment, and/or those with a history of maltreating their children.

Corby (1993) contended that there has been little attempt to elaborate why this over-representation occurs, or the process of how abuse occurs in such families. First, many researchers make the assumption that women are the key figures in cases of physical abuse or neglect (Hiller, Goddard and Diemer 1991, Corby 1993). Where there are male and female carers in the family it is not always determined who has maltreated the child, while in single-parent families it is usually assumed that the mother is solely responsible for any maltreatment, unless there is evidence to the contrary (Corby 1993).

Corby notes '[in] both these situations, there is a likelihood that women might well be over-represented in the abuser category because of beliefs about their roles, responsibilities and natures. In this way researchers replicate the views and practices of social workers and other professionals, who carry out much of their work in such cases almost exclusively with women' (Corby 1993:65).

Corby (1987) found that little work was done with fathers or father substitutes in child protection case management. Child protection workers have been criticised in child death inquiries for failing to engage fathers in casework and thus modifying their abusive behaviors (Fox 1990 as cited in Corby 1993).

The failure of workers and researchers to engage father figures in research or casework has been attributed to sexist, cultural and legal assumptions whereby mothers having any role in the care of their children are virtually automatically assumed to be accountable (Hiller, Goddard and Diemer 1991, Corby 1993). Researchers and practitioners do appear to have a tendency to rely too readily on mothers as informants in families because of assumptions about the mothering role, and their greater availability (Corby 1993).

An exception to this trend, has been the investigation of child homicides, where slightly more is known about those who kill children in their care.

Child homicides

Each year a number of children are killed by members of their family. From 1989-1992 at least 88 per cent of children under 15 years of age (78 of 88 cases) were killed by other family members (Strang 1995). In most western countries it has been found that stepparents kill children in their care at a much greater rate than natural parents and that many more children are killed by stepfathers than by stepmothers (Daly and Wilson 1994, Strang 1995). The over-representation of stepfathers in cases of lethal abuse has been found to be particularly evident in cases of very young children (Daly and Wilson 1994), where children aged from birth to two years are estimated to be 70 to 100 times more likely to die at the hands of a stepfather rather than a natural parent (Daly and Wilson 1994). Daly and Wilson (1994) contended that this may not necessarily be because stepfathers are more dangerous than stepmothers, rather that small children rarely reside with stepmothers.

Using a data set of all cases of fatal child maltreatment reported to police in Australia for the period July 1989 to June 1992, Strang (1995) reported that the typical offender in the 24 cases was a young male living in a defacto relationship with the victim's mother. Strang noted that the majority of the families of child victims were characterised by instability. Often the mothers were in unstable relationships with the child's natural father or another partner. This instability was compounded by financial instability - only one of 25 offenders was known to be employed at the time of the child's death.

Daly and Wilson (1994) contended that the significantly higher risk to children of being killed by a stepparent has not been found to be due to reporting or detection biases, to the incidental traits of people who remarry, differential poverty, the duration of cohabitation, maternal age, or to the size of families. It appears that in associational terms, stepparenting itself is a risk factor for child homicide.

One explanation for this may be that stepfathers, particularly of young children, may incur the greatest social pressure from family and friends to feel and act like natural parents, 'a pressure they often resist and resent, sometimes violently' (Daly and Wilson 1994:208).

The absence of a protective bond between a child and step-parent has been implicated as a risk factor (Daly and Wilson 1994 Strang 1995). Strang (1995) asserted that the absence of a genetic link, in combination with the stress caused by financial insecurity, provide a clear indicator for increased risk of maltreatment. She noted that further support for her argument comes from risk factor lists commonly used in decision making studies and child protection systems - both factors are commonly cited on such lists (e.g. Greenland 1987).

Sexual abuse

As mentioned earlier, much of the investigation of the role of family structure has centred around the involvement of non-biological father figures in cases of child sexual abuse (National Research Council 1993). Stepfathers or defacto fathers are significantly over-represented as perpetrators of sexual abuse (Corby 1993, Angus and Woodward 1995, Angus and Zabar 1995).

However, no consistent profile of child sex abusers has emerged (Oates 1990). To date clinical work with perpetrators as well as research studies have failed to typify offenders by class, profession, wealth or family structure (Willis 1993). Despite the over-representation of non biological father figures as sexual abusers, the majority of such men do not commit acts of sexual abuse. Even when the highest estimate of stepfathers as abusers is used, only 20 per cent of stepfathers are known to be sexually abusive (Russell 1984).

Prevention

Strategies for the prevention of child maltreatment in general terms, are the same regardless of family structure. Previous Issues papers on Neglect (Tomison 1995a) and Sexual Abuse (Tomison 1995b) have highlighted effective primary, secondary and tertiary prevention strategies for maltreatment. To avoid repetition, this section will be confined to describing prevention initiatives that apply specifically to single families or stepfamilies.

Given that single mothers in low socioeconomic groups have been the main target for the investigation of family structure (Ammerman and Hersen 1990), and the subsequent identification of an association between single mother families and poverty, there is a need for the development of policy programs that explicitly support single parents and protect children.

Such programs should be targeted at reducing the negative consequences of impoverishment among single parents of both sexes at the societal level (Gelles 1989). Improving the economic wellbeing of single parents may then lead to a significant reduction in the incidence of violence and maltreatment in such households (Gelles 1989).

Family Support Programs

It was previously suggested that the current lack of knowledge concerning abusive and non-abusive stepparents and the apparent failure of the research community to focus on this population mirrors the lack of attention paid to stepparents or defacto parents by professionals during child protection investigations and interventions (Corby 1987, 1993).

In order to prevent maltreatment, and where it is in the best interests of the child, prevention efforts must attempt to engage not only mothers, but natural fathers and other father figures. It appears that the failure to address the needs of male family members may be slowly changing. Smith and Pugh (1996) asserted that there is currently a greater recognition of the role of fathers in childrearing. Powell (1988) noted a general trend towards matching program content and methods to the need and characteristics of parents (such as family structure), while taking account of the influences of the local community within which the family resides. He stated that '[increasingly] programs aimed at low-income and high risk populations attempt to tailor the services and methods to the perceived and expressed needs of participants' (Powell 1988:6).

While it has become widely recognised that parenting/family support programs need to be matched to local contexts and family needs, the field has been hampered in operationalising this approach (Powell 1988). Previous Clearing House publications have indicated the current dearth of valid evaluation studies in the field of parenting programs (Tomison 1996, James 1994). As a consequence there is a limited knowledge base from which to determine the 'goodness of fit' between parental characteristics and program design (Powell 1988). Powell (1988) argues that, to date, discussions of program responsiveness to local communities have been 'vague or lacking in the identification of factors at the community level that impinge on program design and implementation' (Powell 1988:10).

Further research and evaluation of programs is required to assess which variables affect the adaptation or design of community-based interventions.

In addition, it has been contended that current secondary prevention programs give scant attention to interactions among multiple variables in the determination of risk status for subsequent child maltreatment. Efforts to target a single risk factor are not likely to be as effective in preventing maltreatment as are programs based on a multivariate, interactionist model, particularly one focused directly on the family (National Research Council 1993).

Clearly, just as single factor approaches to the etiology of maltreatment have been flawed, so have attempts to prevent maltreatment which focus primarily on one factor. An interactive approach is therefore advocated as a more effective means of preventing maltreatment, where the influence of constellations of factors in interaction are targeted in prevention programs. It should be noted however, that such an approach is likely to cause greater difficulties in producing clear support for particular frameworks through evaluation, given the complexity of analysis associated with multiple factor approaches (National Research Council 1993).

An expanded role for schools in prevention

Another approach currently advocated by a number of professional and community groups across the nation is the enhancement of current educational programs running in primary and secondary schools. This reflects a growing perception that education should not be limited to purely academic subjects (Cohn 1990, as cited in Oates 1990) and that the education system should take more responsibility for the production of capable, functioning members of society.

In preventative terms, such a responsibility would involve running compulsory protective behaviours programs in schools (as occurs in South Australia) and the uniform teaching of life skills. According to the Victorian Parliamentary Crime Prevention Committee (1995), the latter should address the following issues: criminal law, victim empathy, gender and socialisation, sexual education, sexual assault, child abuse, violence, domestic violence, and alcohol/drug use. Under this approach, education is strongly involved in preparing young people to function in society, rather than working to educate on a purely academic agenda.

Life skills programs should aim to increase young people's awareness of the societal and personal factors (e.g. poverty, unemployment, stress, social isolation and family structure), that underlie the perpetration of child maltreatment, equipping them to resist maltreatment and educating them about sources of community and social support which may ameliorate the potential for abuse or neglect.

Conclusion

While large numbers of factors and combinations of factors have been shown to be associated with various types of child maltreatment, child maltreatment is a multi-determined phenomenon that cannot be explained by any one factor (Ammerman 1990). The U.S. National Research Council (1993) noted that little is currently known about the specific contributions of individual factors (such as poverty, unemployment, and violent neighborhoods) in families characterised by multiple problems, or about the conditions under which these factors interact with other social and personal factors to produce maltreatment. Within this context, the evidence for family structure as a risk factor associated with child maltreatment was discussed.

Clearly there has been a failure to date, to extensively investigate the role of parental characteristics and family structure. There is a need for further investigation, in Australia and overseas, into the impact of family structure on child maltreatment in reconstituted and single parent families. Such an investigation should incorporate an assessment of the positive aspects of such families in combination with the more negative consequences (Corby 1993). It is apparent that little is known about the degree of risk, what factors exacerbate the risk, or in what conditions reconstituted or blended families do adequately care for children (Corby 1993). It may be that the reason for this weakness lies in a mislaid emphasis on predicting and targeting the problem. As Corby states there 'is a need to consider the hows and whys of child abuse as well (1993:83).

A number of recommendations for further investigation arise from the current literature. First, national and state statistics on child maltreatment should provide better information about the identity of maltreaters, and family types. This would enable a better assessment of the proportion of non-nuclear families where children are maltreated, and the proportion of non biological maltreaters.

Second, the roles fathers and other male figures play both as offenders and protectors in the family needs further investigation and clarification. More information is needed on the role that fathers, siblings, grandparents, stepparents, and other household members play in moderating the likelihood of child maltreatment (National Research Council 1993). Such research should form the basis for the development of more child abuse prevention programs which engage all caregivers in a family, in conjunction with a greater acknowledgement of the role fathers and other father figures play in childrearing.

Third, though it is widely accepted that children in stepfamilies are at greater risk of maltreatment, studies are required to test this premise, and to fully investigate the relationship between stepfamilies and the different types of maltreatment.

Finally, despite the difficulties associated with multiple factor investigations of maltreatment, any research on the topic needs to take account of the interaction of factors which may affect child maltreatment, and thus, should investigate family structure in combination with other social, parental and child variables (National Research Council 1993).

Note 1. Throughout this paper the terms 'blended family', 'reconstituted family' and 'stepfamily' will be used interchangeably to refer to families where children reside with one remarried natural parent and her/his new spouse.

Note 2. Henceforth, unless otherwise stated, definitions of 'stepparent' and 'stepfather' will incorporate both unmarried and married partners of a child's natural parent who reside in the family household.

References

• Ammerman, R.T. and Hersen, M. (1990), 'Research in child abuse and neglect: current status and an agenda for the future', in R. T. Ammerman and M. Hersen (eds) Children at Risk: An Evaluation of Factors Contributing to Child Abuse and Neglect, pp.3-19, Plenum Press, New York.
• Angus, G. and Woodward, S. (1995), 'Child abuse and neglect Australia 1993-94', Child Welfare Series No. 13, Australian Institute of Health and Welfare, AGPS, Canberra.
• Angus, G. and Zabar, P. (1995), Child abuse and neglect Australia 1992-93, Child Welfare Series No.9, Australian Institute of Health and Welfare, AGPS, Canberra.
• Australian Bureau of Statistics (1995), Australian Social Trends 1995, Cat. No. 4102.0, Australian Bureau of Statistics, Canberra.
• Bachrach, C.A. (1983), 'Children in families: characteristics of biological, step-, and adopted children', Journal of Marriage and the Family, vol.45, pp.171-179.
• Boehm, B. (1964), 'The community and the social agency define neglect', Child Welfare, November, pp.453-464.
• Browne, K. (1988), 'The Nature of Child Abuse and Neglect: An Overview', in K. Browne, C. Davies and P. Stratton, (eds.), Early Prediction and Prevention of Child Abuse, John Wiley and Sons, Chichester.
• Browne, K. and Saqi, S. (1988), 'Approaches to Screening for Child Abuse and Neglect', in K. Browne, C. Davies and P. Stratton, (eds.), Early Prediction and Prevention of Child Abuse, John Wiley and Sons, Chichester.
• Clulow, C. (1993), 'New families? Changes in societies and family relationships', Sexual and Marital Therapy, vol.8, no.3, pp.269-273.
• Corby, B. (1987), Working with Child Abuse, Open University Press, Milton Keynes.
• Corby, B. (1993), Child Abuse: Towards a Knowledge Base, Open University Press, Buckingham.
• Coulton, C.J., Korbin, J.E., Su, M. And Chow, J. (1995), 'Community level factors and child maltreatment rates', Child Development, vol.66, pp.1262-1276.
• Creighton, S. and Noyes, P. (1989), Child Abuse Trends in England and Wales 1983-1987, NSPCC, London.
• Crime Prevention Committee, Parliament of Victoria (1995), Combating Child Sexual Assault: An Integrated Model, First Report , LV North, Government Printer, Melbourne.
• Dalgleish, L.I. and Drew, E.C. (1989), 'The relationship of child abuse indicators to the assessment of perceived risk and to the court's decision to separate', Child Abuse and Neglect, vol.13, pp.491-506.
• Daly, M. and Wilson, M.I. (1994), 'Some differential attributes of lethal assaults on small children by stepfathers versus genetic fathers', Ethology and Sociobiology, vol.15, pp.207-217.
• Daro, D. (1988), Confronting Child Abuse: Research for Effective Program Design, Free Press, New York.
• Edgar, D. (1988), 'Family Change', in J.M. Najman ands J.S. Western (eds), A Sociology of Australian Society: Introductory Readings, Macmillan, Melbourne.
• Edgar, D. (1991), 'Families and the social reconstruction of marriage and parenthood in Australia', in R.Batten, W.Weeks, and J.Wilson (eds), Issues Facing Australian Families: Human Services Respond, Longman Cheshire, Melbourne.
• Finkelhor, D. (1984), Child Sexual Abuse: New Theory and Research, The Free Press, New York.
• Finkelhor, D. and Baron, L. (1986), 'High-Risk Children', in D. Finkelhor, A Sourcebook on Child Sexual Abuse, Sage Publications, Beverly Hills.
• Gelles, R.J. (1989), 'Child abuse and violence in single-parent families: parent absence and economic deprivation', American Journal of Orthopsychiatry, vol.59, no.4, pp.492-501.
• Gelles, R.J. (1992), 'Poverty and violence towards children', American Behavioral Scientist, vol.35, no.3, pp258-274.
• Gelles, R.J. and Harrop, J.W. (1991), 'The risk of abusive violence among children with nongenetic caretakers', Family Relations, vol.40, pp78-83.
• Gil, D. (1970), Violence Against Children, Harvard University Press, Mass.
• Giles-Sims, J. and Finkelhor, D. (1984), 'Child abuse in stepfamilies', Family Relations, vol.33, no.3, pp407-413.
• Greenland, C. (1987), Preventing CAN Deaths: An International Study of Deaths Due to Child Abuse and Neglect, Tavistock, London.
• Goddard, C. R. and Hiller, P. C. (1992), Tracking Physical and Sexual Abuse Cases from a Hospital Setting into Victoria's Criminal Justice and Child Protection Systems: A Report for the Victorian Law Foundation, Volumes 1-3, Department of Social Work and Department of Anthropology and Sociology, Monash University, Melbourne.
• Hiller, P.C., Goddard, C.R. and Diemer, K.M. (1991), 98 Cases of Child Abuse Presenting at a Major Melbourne Hospital: Sociodemographic Factors and Institutional Reactions, vol. 1: The research, Department of Social Work, Monash University, Melbourne.
• James, M. (1994), 'Analysis of the Programs database', in National Child Protection Clearing House Newsletter, vol.2, no.1, National Child Protection Clearing House, Australian Institute of Criminology, Canberra.
• Krugman, R.D. (1986), 'The relationship between unemployment and physical abuse of children', Child Abuse and Neglect, vol.10, no.3, pp415-18.
• National Research Council (1993), Understanding Child Abuse and Neglect, National Academy Press, Washington DC.
• Nelson, K. E., Saunders, E. J. and Landsman, M. J. (1993), 'Chronic child neglect in perspective, Social Work, Vol.38, No.6, pp.661-671.
• Oates, R.K. (1990), 'Understanding the problem', in R.K. Oates (ed.) Understanding and Managing Child Sexual Abuse, Harcourt Brace Jovanovich, Sydney.
• Pilger, J. (1989), A Secret Country, Jonathan Cape, London.
• Polansky, N.A., Gaudin, J.M. and Kilpatrick, A.C. (1992), 'Family radicals', Children and Youth Services Review, vol.14, pp19-26.
• Powell, D.R. (1988), 'Emerging directions in parent-child early intervention', in D.R. Powell (ed), Parent Education as Early Childhood Intervention: Emerging Directions in Theory, Research, and Practice, Advances in Applied Developmental Psychology, vol.3, Ablex Publishing Corp., Norwood, NJ.
• Russell, D. (1984), Sexual Exploitation: Rape, Child Sexual Abuse and Workplace Harassment, Sage Publications, Beverley Hills, CA.
• Sack, W., Mason, R. and Higgins, J. (1985), 'The single parent family and abusive child punishment', American Journal of Orthopsychiatry, vol.55, pp.253-259.
• Smith, C. and Pugh, G. (1996), Learning to be a Parent: A survey of group-based parenting programmes, Family Policy Studies Centre, London.
• Steele, B. F. and Pollack, C.B. (1968), 'A psychiatric study of parents who abuse infants and small children', in R.E. Helfer and C.H. Kempe (eds.), The Battered Child, University of Chicago Press, Chicago.
• Strang, H. (1995), 'Child abuse homicides in Australia: Incidence, circumstances, prevention and control', in D. Chappell and S. Egger (eds) Australian Violence: Contemporary Perspectives II, Australian Institute of Criminology, Canberra.
• Tomison, A.M. (1994), An Evaluation of Child Abuse Decision Making in the Barwon Region: A Report for the Victorian Health Promotion Foundation, Volumes I and II, Department of Social Work and Human Services, Monash University, Melbourne.
• Tomison, A.M. (1995a), Spotlight on Child Neglect, Issues Paper No.4, National Child Protection Clearing House, Australian Institute of Family Studies, Melbourne.
• Tomison, A.M. (1995b), Update on Child Sexual Abuse, Issues Paper No.5, National Child Protection Clearing House, Australian Institute of Family Studies, Melbourne.
• Tomison, A.M. (1996), 'Child abuse prevention programs in Australia: An audit of the National Child Protection Clearing House database', in National Child Protection Clearing House Newsletter, vol.4, no.1, National Child Protection Clearing House, Australian Institute of Family Studies, Melbourne.
• Tower, C.C. (1989), Understanding Child Abuse and Neglect, Allyn and Bacon, Boston.
• Wallis, Y. (1992), The Victorian Community's Attitudes to Child Sexual Abuse, Community Services Victoria, Melbourne.
• Willis, G.C. (1993), Unspeakable Crimes: Prevention Work with Perpetrators of Child Sexual Abuse, The Childrens Society, London.
• Wolock, I. and Horowitz, B. (1984), Child Maltreatment as a Social Problem: The Neglect of Neglect, National Clearinghouse on Family Violence, Ottawa.