Family Resilience and Good Child Outcomes: An Overview of the Research Literature

Ross Mackay
Principal Advisor
Ministry of Social Development

Abstract

A review of the international research literature on family resilience shows that processes that operate at the family level - including strong emotional bonds, effective patterns of communication, the use of coping strategies and family belief systems, especially those based on spiritual or religious values - are important means by which families manage to cope with adversity.  Positive parenting is a key influence on children's development, especially in adverse financial circumstances.  Wider family involvement can also assist families to cope with stress.  In particular, non-resident fathers and other father figures have an important role to play in promoting the development of children in lone-mother families, while the burden of teenage parenthood can be eased by multi-generational co-residence.  On the question of whether it is possible to inculcate resilience in families, evidence from a range of recent evaluations of selected intervention programmes shows that approaches that work best are those that involve early intervention, that are sensitive to families' cultures and values and that assist in relieving families' ecological stresses.

Introduction

One of the enduring mysteries that confronts those who work with families and children - and those who are concerned with child and family policy - is why some families respond positively to serious threats and challenges to their wellbeing, while others in similar circumstances do not manage to do so.  The concept of resilience has been developed by researchers to denote positive adaptation under adverse circumstances.  This paper presents an overview of the literature on resilience, with primary emphasis on how the concept has been applied at the level of the family.  The focus of the paper is on the relation between family resilience and child outcomes: how is that in some families facing adversity the children emerge unscathed from the experience, while in other families facing similar circumstances the children's development is seriously impaired?

Concepts and Definitions

The Concept of Family Resilience

Family resilience is an emerging concept.  Its origins lie in the study of individual resilience.  Historically, researchers interested in resilience have focused on attributes of children that are associated with positive adaptation under adverse circumstances (e.g. academic competence or a sense of self-efficacy).  More recently, research scholars have extended the idea to the level of the family.  While the theory has been elaborated by a number of scholars - although further development is needed in a number of respects - empirical evidence on the phenomenon of family resilience is still rather sparse.

A key element in the concept is that of successful engagement with risk.  A family can be considered resilient where it has encountered adversity and coped successfully with the challenge.  This has led to a focus on family strengths - those qualities that allow families to cope successfully with challenges to their wellbeing.  While early studies of resilience among children tended to take a static view - some of the earliest work used the terms "invulnerable" and "invincible" children - emphasis is now placed on its dynamic nature.  Thus resilience is viewed as a process of adaptation under challenges to wellbeing.  Contemporary theorists also emphasise the fact that resilience is not a categorical state, but a continuum (families can be more or less resilient) and that it is contingent (families may be resilient in some circumstances but not others).

A Definition of Family Resilience

Luthar et al. (2000) provided a useful definition of family resilience as:

a dynamic process encompassing positive adaptation within the context of significant adversity.

This incorporates a number of key features of the concept, including resilience as a dynamic process of adaptation, and encapsulates the idea of successful engagement with risk.

Related Concepts

The concept of family resilience is most usefully understood in relation to a number of other key concepts, especially the concepts of risk factors and protective factors.  Risk factors increase the probability of negative outcomes.  Protective factors interact with risk to change the predictive relationship between risk factors and negative outcomes, reducing the probability of negative outcomes.  Another concept used by some researchers is vulnerability.  Vulnerability factors increase the probability of negative outcomes in the presence of risk.

Protective factors and vulnerability factors are both posited on the idea of an interaction with risk: the former reduce the probability of negative outcomes in the presence of risk, while the latter increase the probability of negative outcomes in the presence of risk.  A number of studies examine how protective factors work to reduce the effect of high-risk ecological circumstances, but there is little research evidence on the interaction between vulnerability and risk.  Much of the research on family resilience is concerned with a search for protective factors, which reflects the emphasis on family strengths.

Another useful conceptual distinction is that between proximal and distal variables.  Proximal variables have effects that are experienced directly.  Distal variables have effects that are experienced indirectly through other mediating variables.  In the context of an analysis of sources of influence on children's outcomes, an example of a distal variable is the socio-economic position of the family, while an example of a proximal variable is the parenting behaviours of the child's parents.  Whereas parenting behaviours impact directly on children, the socio-economic position of the family impacts only indirectly through other mediating factors (including parenting behaviours).

These concepts provide a useful way of framing a resilience hypothesis: families may create a low-risk proximal environment for their children's development despite living in a high-risk distal environment.  In examining this hypothesis empirically, the challenge for researchers is to identify proximal factors in the child's family environment that allow children to thrive despite the challenges of an adverse distal environment.

Critiques of the Concept of Family Resilience

The field of family resilience research has been subject to a number of critiques.  Much of the criticism focuses around concerns about definitional confusion.  The term "resilience" has itself been defined in a range of ways and has been viewed by different researchers variously as a trait, a process and an outcome.

The related concepts of risk, protective and vulnerability factors have been subject to similar criticisms.  In particular, the distinction between risk factors and vulnerability factors has not been clearly articulated, since both focus on the idea of things that elevate the risk of poor outcomes.  A distinction made by some researchers is to use the concept of risk to refer to environmental circumstances, while vulnerability is used to refer to individual (and, by extension, family) dispositions.  Some researchers have used the term "vulnerability" to refer specifically to genetic predispositions to disorder.  However, the term has not been used exclusively for such conditions.

In addition to definitional confusions, there has also been a lack of consensus about measurement, use, findings and interpretation of findings.  Concerns have also been raised about the way the label of resilience can lead to a "blame the victim" mentality by attaching negative labels to those who lack resilience.

Some commentators have raised more profound questions about the concept itself.  Tarter and Vanyukov (1999) noted that the label "resilience" is often applied on a post-hoc basis, with researchers "discovering" resilience whenever a positive outcome occurs under conditions that might be considered adverse.  Others have questioned whether the concept is really one that can be applied at the level of the family and whether work on family resilience has added anything to the work on individual resilience.

Underlying some of these latter critiques is a fundamental question about whether family resilience really exists as a phenomenon in its own right.  Certainly no studies have attempted to operationalise and measure the concept of resilience directly.  Rather, most work on resilience has focused on mechanisms of resilience - such as parenting practices or patterns of family communication.  Thus the notion of family resilience must be considered, if it exists, to be a type of latent phenomenon whose effects can only be observed indirectly through a range of different aspects of family functioning.

The present paper will not delve into these issues.  Instead it is assumed that family resilience offers a useful way of approaching fundamental questions about the ways in which families function, in particular by examining different aspects of family functioning that are associated with adaptive outcomes under stress.

The focus of the discussion will now turn to the empirical research evidence on family resilience.  The discussion falls into two parts.  First, an examination will be made of selected aspects of family resilience - general factors that operate to make families resilient under stressful circumstances.  Second, the discussion will examine families in certain situations of ecological disadvantage - specifically poverty, lone parenthood and teenage childbearing - and will identify protective factors that can help families in these circumstances to avoid the adverse consequences of their disadvantaged position.

Aspects of family resilience

While some progress has been made in theory-building, the empirical evidence on family resilience is still sparse.  Many of the most relevant studies have not been specifically cast as studies of family resilience, but have focused on particular aspects of family functioning associated with positive adaptation in circumstances of disadvantage.  The following discussion will provide an overview of relevant research findings in a range of domains of family functioning which are regarded as aspects of family resilience, including family cohesion, family belief systems, coping strategies and communication.

Family Cohesion

The emotional connections between family members are crucial to the functioning of a family.  Families that have good emotional bonds are better able to rise to challenges to their wellbeing and cope well under stress.  This has long been recognised by clinicians who work with families, and a range of instruments has been developed to measure the strength of the emotional bonds within families.  While different researchers have given different names to the underlying construct - family cohesion (Olson 1993), connectedness (Walsh 1998), affective involvement (Epstein et al. 1978), Steinhauer et al. 1984) - the key idea is that of emotional connection between family members.  In all these models, very low levels of cohesion are indicative of family dysfunction.

On the other hand, very high levels of family cohesion can also be dysfunctional, because emotional connections between family members also need to be balanced against family members' needs for individual autonomy.  Thus families with very high levels of cohesion have been characterised by Steinhauer et al. (1984) as "enmeshed", involving "intense, stifling relationships", while Epstein et al. (1978) used the term "symbiotic", which signals "extreme or pathological interest or investment in each other".  The middle position on the measure in both of these models, described as "empathic involvement", is regarded as the healthiest level of family functioning.  Similarly, the middle position in Olson's (1986, 1993) Circumplex Model of Marital and Family Systems, described as the balanced region of the scale, is considered optimal for family functioning.

There is a body of research results showing that a relationship does exist between family cohesion and family functioning.  Olson et al. (1988) showed that families with a higher level of cohesion within the balanced range of the Circumplex Model experienced fewer intra-family strains and higher wellbeing than families with lower levels of cohesion.  Other work has shown that families at the extreme ends of the model are more likely to display symptoms of dysfunction than families within the balanced region of the model.  Similar results have also been obtained for other models.

A number of studies have explored the relationship between family cohesion and family functioning in the context of changes in family structure.  Although step-families tend to have lower levels of cohesion than intact first families, those with higher levels of cohesion tend to have higher levels of family satisfaction and lower levels of stress.  Thus step-families can benefit from higher levels of cohesion.  There is also evidence that a family's level of cohesion can change over time.  Families with adolescents tend to have lower levels of cohesion than those with younger children.  This is not linked with family dysfunction, however.  Rather it seems to be part of a process of natural evolution within the family as older children undergo a process of individuation and autonomy seeking.

Family Belief Systems

In the view of Walsh (1998), family belief systems are at the core of family functioning.  Family belief systems encompass values, attitudes, convictions, biases, assumptions - "a set of basic premises that trigger emotional responses, inform decisions and guide actions" (Walsh 1998).  The dominant beliefs of a family shape how the family as a unit copes with crisis and adversity.

Walsh distinguished three important dimensions of family belief systems: capacity to make meaning out of adversity, a positive outlook and spirituality or transcendence.  Well-functioning families have the capacity to understand what has happened to them in the past and to visualise a different future.  In Walsh's words, resilient families have "a global orientation to life as comprehensible, manageable and meaningful".  Resilient families are also characterised by active persistence, perseverance, maintenance of hope and optimism, and confidence that they can overcome the odds.

Transcendent beliefs are those that supply meaning in people's lives and are often based on spiritual or cultural foundations.  Walsh notes that such beliefs "offer clarity in our lives and solace in distress; they render unexpected events less threatening and enable acceptance of situations that cannot be changed".  Walsh describes spirituality as a key process in family resilience, as it "involves an active investment in internal values that bring a sense of meaning, inner wholeness, and connection with others".  Because of the importance of religion as a key source of transcendent beliefs, the research evidence on its connection with family functioning is discussed as a separate issue in its own right.

The Role of Religion

Most studies of the relationship between religion and family functioning have focused on the marital relationship or on the relationships between parents and children.  A considerable body of evidence documents a link between religion and marital stability, adjustment and happiness.  For example, Scanzoni and Arnett (1987) showed that religious devoutness was positively related to marital commitment and the use of positive conflict-resolution tactics.  More recently, Ellison et al. (1999) showed that regular church attendance was negatively associated with spousal domestic violence.

Research also supports the hypothesis that religion has a positive impact on parent-child relationships and on children's outcomes.  Mahoney et al.'s recent (2001) meta-analysis suggested that religion facilitates positive family interactions and that it also lowers the risk of child maladjustment and adolescent drug and alcohol use.  Pearce and Axinn (1998) showed that there was a positive relationship between emphasis on religion and affective mother-child relationships.  Brody et al. (1996) showed that, among African-American families, parental emphasis on religion is positively associated with family cohesion and negatively associated with inter-parental conflict.  In turn, these variables were found to be associated with children's outcomes.  Children whose parents placed a strong emphasis on religion were less likely to display externalising or internalising problems.  This finding provides support for the view that, among the African-American population in the United States, religious involvement promotes supportive and responsive family relationships which assist them to cope with stress.

A range of mechanisms has been postulated to explain the link between religion and positive family functioning.  Mahoney et al. (2001) distinguish between the functional aspects of religion and its substantive aspects.  The latter refers to the content of religious beliefs.  Such beliefs carry a freight of positive messages about pro-social values.  They may also offer a framework for coping with difficult situations and enable families to make meaning out of adversity.  Messages from religious leaders may also shape parental beliefs that could shape attitudes to parenting.  The functional aspects of religion may be even more important.  One important mechanism through which religious participation affects family functioning may be by augmenting a family's social capital.  Membership of a church congregation provides links to other people with similar values, which may provide increased access to social support, child care and instrumental or financial assistance in dealing with problems.  Attendance at church services may also help the family to achieve cohesion by engaging in joint activities or by promoting shared value systems.  Participation in prayer could also afford families a form of therapeutic strength.

While most of these research findings are based on American studies and hence may not be highly relevant for New Zealand, which is perhaps a more secularised society, the results for African-American families could be suggestive.  In particular, given the relatively high rates of participation of Pacific families in religious activities, it is tempting to speculate that religion may afford Pacific families some degree of protection from adversity in an analogous way to its protective role for African-American families in the United States.  However, this would need to be confirmed by local research.

Coping Strategies

Coping is a conscious intentional response to stress.  Coping is often invoked to represent competence and resilience.  However, these three terms have distinct meanings.  Where coping refers to adaptive responses to stress, competence refers to the characteristics that are needed for successful adaptation and resilience is reflected in outcomes where competence and coping have been displayed (Compas et al. 2001).

A range of different types of coping behaviours has been identified.  Lazarus and Folkman (1984) distinguish between problem-focused coping and emotion-focused coping.  Problem-focused coping involves confronting a problem to reduce the effect of a stressor or set of stressors, while emotion-focused coping involves dealing with the emotional distress occasioned by the stressor or stressors.  Another distinction has been made between approach coping, which involves confronting the problem, and avoidant coping, which involves disengagement or denial.

There is a body of empirical research that focuses on the association between coping behaviour in children and adolescents and psychological adjustment, social and academic proficiency and symptoms of pathology.  Coping has been shown to be significantly associated with the psychological adjustment of children (Compas et al. 2001).  Different coping strategies can be functional or dysfunctional.  Problem-focused coping and engagement coping have been found to be associated with greater adjustment in children and adolescents, while emotion-focused and disengagement coping have been found to be associated with lesser adjustment (Compas et al. 2001).  Seiffge-Krenke (1995) also found two types of coping styles to be functional: active coping (which involves active support-seeking) and internal coping (which involves internal reflection on possible solutions), while withdrawal and denial were found to be dysfunctional coping styles.

Communication

Communication is a key aspect of family functioning.  In Patterson's (2002a, 2002b) view, communication skills function as a vital family-level protective resource.  Communication is central to the process of meaning-making in families - how family members perceive themselves and their relation with the outside world and how they make sense of the challenges they are facing - as well as to the process of coping.  Effective communication is important in the development of a sense of shared decision-making, which is achieved through negotiation, compromise and reciprocity.  The development of effective communication processes within the family also transfers to relationships with the professional world; for example, in dealings with service providers.

Communication has both a content function and a relationship function.  Instrumental communication corresponds to the ways in which family members inform each other about things that need to be done.  Affective communication corresponds to the ways in which family members show each other love and support and is crucial for fulfilling the nurturant family function.  Walsh (1998) suggests that in every communication, family members either support or contest the nature of their relationship.

Walsh identifies three important components of effective communication: clarity of expression, open emotional expression and collaborative problem solving.  Clarity refers to the sending of clear and consistent messages, in both words and actions, as well as awareness of the need to clarify ambiguous signals.  Open emotional expression refers to the sharing of feelings and emotions, in relationships characterised by mutual empathy and a toleration for differences.  Collaborative problem solving involves identifying problems and options to deal with these and working jointly to surmount them.

Effective communication is especially critical at times of sudden crisis or prolonged stress, as these are the times when communication is most likely to fail.  Resilient families are distinguishable from non-resilient families in that they have the ability to manage conflict well, and management of conflict depends crucially on communication and problem-resolution skills.  According to Patterson (2002a, 2002b), different methods of communication can be effective for different families, but poor communication increases vulnerability to risk.

Resilience in disadvantaged ecological circumstances

Because the notion of resilience can only be understood in relation to risk, it is useful to give some consideration to the way resilience might be manifested in common situations of ecological risk.  Three such circumstances will be considered here: poverty, lone parenthood and teenage childbearing.  In each case, consideration will be given to a specific factor that can provide potential protection from adverse effects.

Poverty

Poverty has been linked to a range of adverse outcomes for children, including reduced cognitive ability, poor academic achievement, poor mental health and conduct disorders.  One of the fundamental building blocks of normal cognitive development is proper nutrition in the antenatal period, infancy and early childhood.  Malnutrition interferes with children's cognitive development and delays their readiness for learning when they enter school, setting them on a trajectory of poorer academic achievement.  This is compounded by adverse effects on mental health, social adjustment and behaviour.

A range of theories has been put forward to explain the linkage between poverty and child outcomes.  One of these emphasises the role of parental investments in children.  Poor families may have a deficit in both material resources (including adequate nutritional care and cognitively stimulating materials, such as books) and non-material resources (which include such things as education, information and skills).  Children from such families are likely to do less well in life because of these deficits.   A second theory emphasises the role of parental behaviours, attitudes and expectations.  The stress of constrained economic circumstances diminishes parental mental health, which reduces parents' ability to nurture, monitor and discipline their children effectively.  In addition, parents in such families may have reduced expectations about their children's life chances, which may be transmitted to their children.

A third theory emphasises the role of social capital.  The concept of social capital has its origins in the economic literature.  Just as the original notion of economic capital has been extended to the notion of human capital, representing the value that attaches to education and skills acquired by individuals, the concept of social capital represents a further extension of the idea, denoting the value that resides in social relationships.  According to the theory, people make "investments" in relationships with others, which acquire a certain value that can be drawn on.  For example, friends and neighbours with whom one has established an abiding relationship can provide support and assistance in time of need, provide access to information and employment opportunities, and so on.  Well-off parents are more likely to have networks of relationships with people and institutions of influence than poor parents, which has consequences for their children's relative chances of success.

Parenting as a Protective Factor

Parenting behaviour is one mechanism that links poverty and children's wellbeing.  Poor parents, on average, provide fewer opportunities for their children, are less involved in their children's lives, are more erratic in their use of discipline and monitor their children less.  Parenting quality has been found to be significantly related to conduct in childhood and is predictive of academic, conduct and social competence in adulthood.

A key mediating mechanism is parents' experience of economic strain - the degree to which they have difficulty in managing to make ends meet with limited income.  In several studies this construct emerged as a more significant predictor of parents' psychological distress and parenting difficulties than income per se.  In other words, how a family fares is not simply a matter of how much income the family has, but how they cope with the struggle of managing with limited resources.  This suggests that, as well as interventions that aim to boost the incomes of poor families, there could be benefits from programmes that aim to improve the coping skills and mental health of poor families.

One way of understanding parenting behaviour was developed by Baumrind (1991), which she referred to as "parenting style".  Parenting behaviours are classified on two dimensions - warmth and control.  Parents who rated high on both warmth and control of their children are characterised as authoritative parents.  Authoritative parenting is regarded as optimal for children's development.  This is contrasted with other less-effective parenting styles - authoritarian (low on warmth and high on control), permissive (high on warmth and low on control) and neglectful (low on both warmth and control).  Authoritative parenting has been shown to be associated with positive child outcomes, while other parenting styles are associated with a range of varying problems in child development.

However, context is also important and some types of higher-risk environments may require different approaches to parenting.  For example, Brody et al. (1998) identified a particular type of parenting, which they characterised as "no-nonsense parenting", among low-income rural African-American lone mothers.  This involved high levels of warmth and affection and high levels of control, including physical discipline.  This varies from the standard accounts of optimal parenting behaviour by including physical discipline.  However, it is possible that in some higher-risk environments, such as those facing African-American children in the United States, a different approach to parenting is more adaptive. In particular, given the higher levels of environmental risks, there may be benefits in higher levels of monitoring and discipline of children to ensure their safety.

Lone Parenthood

Single parenthood poses risks for parental wellbeing and children's adjustment.  Children of sole parents, on average, have poorer records of academic achievement, display higher rates of psychological distress and have an increased likelihood of non-marital childbearing than their peers from two-parent families.  In addition, sole mothers have poorer mental health than do their partnered peers, which affects their capacity to parent their children effectively and thus has a knock-on effect on their children's development.

Three primary explanations have been advanced to explain these differences.  The first theory emphasises the role of economic deprivation.  Substantial economic differences have been demonstrated between single-parent and two-parent families.  One study has shown that these differences account for approximately one half of the variation in child development outcomes between single-mother and two-parent families (McLanahan and Sandefur 1994).  A second theory emphasises the importance of two parents for socialisation of children.  There are clear advantages in having two parents to carry out important parenting functions such as supervision and monitoring of children.  Children are also hypothesised to benefit from the presence of a male role model in two-parent families.  A third theory emphasises the role of stress.  Parenting children without assistance from a partner is more stressful than when another adult is there to share the work.  In addition, changes to family arrangements which often precede lone parenthood (including separation, divorce and bereavement) involve considerable stress, especially where separation involves a degree of conflict between the separating parents.  This stress reduces the capacity of the lone parent to parent effectively.  Changes in family structure also often involve changes in the children's place of residence and school, which are postulated to have a cumulative negative impact on children.

Father Involvement as a Protective Factor

Increasing attention has been paid in recent years to the role non-resident fathers play in their children's lives.  Non-resident fathers who engage in authoritative parenting and provide economic support for their children can make important contributions to their children's lives.  Thus if separating parents can successfully negotiate an arrangement whereby the father remains an active participant in the lives of the children, the father's involvement may serve as a protective factor for the children.

If this protective function is to be realised, it is important that a good relationship is maintained between the parents.  Conflict between separated parents has adverse effects on children, which can undermine the positive contribution made by non-resident fathers.  In cases of serious or prolonged conflict, the adverse effects on the children can be such that the children would be better off not seeing the father at all than continuing to see him in a situation of ongoing serious conflict.  Thus both parents need to work to minimise conflict and co-operate effectively on decisions about the children.

The protective influence of non-resident father involvement is likely to arise from a range of sources.  Financial support provided by the father for the child will have an obvious direct benefit.  If the father engages in authoritative parenting, this too will have a direct impact on the child.  The father may also make a more indirect contribution to the wellbeing of the child, by assisting the mother in sharing the burden of caring for the child and providing other forms of support.  This is likely to have positive effects on the mother's mental health, with flow-on effects for the child.  Finally, children also benefit from having contact with their biological father, in terms of their development of self.

Although the latter role cannot be filled by anyone other than the biological father, other functions can be performed by other male figures.  Thus the attention of researchers has been increasingly directed towards the role of "social fathers" - male figures who act like a father to the child.  There is empirical evidence that social fathers can have a positive influence on children's development, although it appears that this may vary depending on who the father figure is.  Jayakody and Kalil (2002) reported that male relatives (e.g. uncles or grandfathers) can have a positive influence on children's schooling.  This appeared to come about through an indirect impact on the mother's mental health.  On the other hand, where the father figure was the mother's romantic partner, children had lower levels of emotional adjustment.  Other evidence also supports the view that where the new partner of a mother attempts to play a strong role in parenting step-children, this can lead to problems of adjustment in the children.

Teenage Childbearing

Teenage childbearing has been associated with an array of negative outcomes for the teenage mother and her child.  Teenage pregnancy is likely to lead to curtailment of schooling and hence to lower rates of school achievement.  It is also associated with depression and psychological distress.  Depression among teenage mothers is associated with lack of employment and rapid-repeat childbearing.  Teenage mothers also have fewer parenting skills.  For example, they are less sensitive and responsive to their children and are more likely to use punitive discipline and display more physically intrusive behaviour.  This leads to higher rates of adjustment problems among children.

It is unclear how much these effects derive from the fact of teenage childbirth itself and how much they are a result of selection effects.  Some studies have shown that when pre-pregnancy measures of academic ability, behaviour and family circumstances are taken into account, the academic differences between teenage parents and their non-parenting peers disappear.  In other words, young girls from disadvantaged backgrounds tend to have limited prospects irrespective of whether they have a child.

Multi-generational Co-residence as a Protective Factor

Multi-generational co-residence has long been thought to help young single mothers to navigate the challenges of early parenting.  Generally, most teenage mothers live with their own mothers, at least for the first few years after giving birth.  Multi-generational living arrangements are more likely when the mother is younger, is unmarried, has fewer children and has few economic resources of her own, and when the grandmother has provided support through the young mother's pregnancy.

Multi-generational co-residence has been found to have both positive and negative effects on teenage mothers.  On the credit side of the ledger, such arrangements can benefit adolescent mothers' socio-economic outcomes, especially where grandmothers assist the young mothers to acquire more education.  Furstenberg et al. (1987) found that mothers who were unmarried and remained living in their parents' homes for the first five years of their child's life were more likely to have returned to and finished high school, worked and not received welfare than their unmarried counterparts who had moved out of their parents' home.  The greater availability of child care and support for the young mother in multi-generational households may explain much of this positive effect.

On the other hand, there is also evidence that multi-generational living arrangements can have a negative effect on young mothers' parenting competence.  Negative effects have also been reported on young mothers' mental health.  Kalil et al. (1998) showed that grandmother co-residence is associated with increased psychological distress among adolescent mothers, particularly where there is a high degree of conflict between the adolescent mother and the grandmother.  Some of this association is likely to arise out of the negative effects of intergenerational conflict, shared responsibilities for child rearing and other family processes salient to young mothers' transition to adulthood, and grandmothers' mid-life identity as parents and grandparents.

Another study adds a further layer of complexity to this picture.  Schweingruber and Kalil (2000) found that higher levels of grandmother participation in decision making about the young mother and her child were associated with more depressive symptoms only among white (and not black) teenage mothers in multi-generational households.  This suggests there may be some variation across ethnic groups in the extent to which there is a downside to multi-generational co-residence.  It has been argued by some commentators that black mothers might adapt more successfully than white mothers to single-parent status, perhaps because single parenthood is more normative in the black community (Fine and Schwebel 1988).  In addition, among many ethnic communities, extended family living arrangements are more common and regarded as affording protection for parental wellbeing and children's development.  In contrast, Hawkins and Eggebeen (1991) characterise the three-generation white family as aberrant: neither expected nor preferred by its members due to its violation of white family norms of independence.  While the New Zealand cultural context is different from that in the United States, there are some echoes in the New Zealand situation that suggest potential avenues for future research.

Exemplar programmes

The lessons from family resilience research can only be useful in improving the lives of families and children if they can be successfully operationalised in programmes that are delivered to families who are experiencing difficulties or who are at risk of poor outcomes.  As yet, there has been little work that has specifically tried to implement the ideas that have emerged from the research on family resilience.  However, a broader review of the evaluation research shows a number of programmes that have effected improvements in the lives of families and children.  The following discussion will examine a short list of exemplar programmes that cover a range of different approaches to boosting family and child outcomes, and that have met with varying degrees of success.

Nurse Home Visitation Program

One approach aims to boost child outcomes by improving parenting behaviours.  One of the most successful parenting programmes is the Nurse Home Visitation Program (Olds et al. 1999).  The goal of the programme is to improve pregnancy outcomes, early child health and development and the personal development of mothers.  The programme involves weekly visits for the first month following enrolment and fortnightly visits for the remainder of the pregnancy.  After the birth of the child, the mother is visited weekly for the first six weeks, fortnightly up to the age of 21 months and then monthly until age two.  Visits last between 60 and 90 minutes and cover a range of issues, including personal health, environmental health and maternal role and life course.  Various assessments are made of maternal, child and family functioning.

Evaluations of the programme have demonstrated a range of benefits for both mothers and their children.  Among children, the positive outcomes include reduced rates of child maltreatment and reduced rates of ingestions and injuries.  Among mothers, the positive outcomes include reduced rates of smoking during pregnancy, reduced rates of rapid-repeat pregnancy, increased rates of employment and reduced rates of welfare receipt.  Longer-term effects were also found for the children.  At age 15 they were found to have had fewer arrests and convictions, smoked and drank less, and had fewer sexual partners.  However, the intensive levels of visitation - mothers were visited an average of nine times during their pregnancy and 23 times during the child's first year of life - mean that the programme was very expensive.  Even in the United States, where it has been implemented in a restricted number of sites on a demonstration basis, it has been considered too costly to be implemented nation-wide (Duncan and Magnuson 2001).

Comprehensive Child Development Program

The Comprehensive Child Development Program stands alongside the Nurse Home Visitation Program as a cautionary example of how a programme of somewhat different design can produce quite disparate results.  The programme aimed to enhance the physical, social, emotional and cognitive development of children in low-income families and to assist families to become economically self-sufficient.  It also involved visits to the home, not by a trained nurse but by a case manager, who undertook some work with families but also provided referrals to other sources of assistance.  A strong emphasis of the programme was on early childhood education, which was provided in the home - in some cases by the case manager and in other cases by a professional child educator.

In contrast to the Nurse Home Visitation Program, evaluations of the Comprehensive Child Development Program have not produced evidence of any impact in any of the target domains - child learning, child and family health, parenting, family economic self-sufficiency or maternal life course (St Pierre and Layzer 1999).  It is unclear why these two programmes have achieved such disparate results.   Evaluators are still unable to determine which components of a programme are crucial to its success and why one programme succeeds where another fails.  It is important to learn more about this through detailed study of such programmes, since public investments in early intervention programmes are at risk if these questions remain unanswered.

Carolina Abecedarian Project

The Carolina Abecedarian Project is another example of a high-intensity early childhood intervention which has achieved significant success, making an enduring difference to the life trajectories of its subjects.  The programme was targeted at children who were identified as being at risk of mental retardation because of the conditions in their home environments.  The programme provided enriched centre-based child care services for eight hours a day, five days a week, for 50 weeks of the year from birth to age five.  Children were randomly assigned between the programme and a control group, whose members were able to take part in other child care services that are generally available in the community.

Evaluations of the programme have shown a range of impressive effects that have endured over the years as the children have grown to early adulthood (Campbell et al. 2001, 2002).  At the most recent follow-up at age 21, the children had higher scores on tests of cognitive ability, higher rates of participation in tertiary study and were more likely to have delayed parenthood.  The children were also subject to a further experimental trial at school age.  Half of each group were randomly assigned to receive support from a Home-School Resource Teacher who provided additional instruction and acted as a liaison between the parents and the school for a further period of three years.  In contrast to the early intervention, the school-age intervention made no significant difference to the children's learning or other outcomes.

Minnesota Family Investment Project

Another approach that has shown promise in improving outcomes for children involves initiatives to boost the incomes of low-income families, primarily by encouraging participation in the workforce and ensuring that this results in a lift in incomes.  An example of a successful initiative of this type is the Minnesota Family Investment Program (Knox et al. 2000).  The programme includes two components: mandatory participation in employment for long-term welfare recipients, and financial incentives to encourage work efforts.  Employment is mandatory immediately for two-parent families and within six months for sole parents.  Under the programme, participants keep part of their benefit until their earnings reach 120% of the poverty rate.

The programme has resulted in increased rates of employment and increased earnings among participants.  Of more interest in the current context is the fact that it has also produced improvements in school performance among the children of participants, as well as reduced rates of child behaviour problems.  This shows that there may be a range of different mechanisms for improving child outcomes.  While the above programmes have shown the potential for effecting improvements by changing parental behaviour or by enhancing early childhood education, this programme demonstrates the importance of income increments for low-income families.

Families and Schools Together (FAST)

The final exemplar programme is one aimed at improving links between families and the community.  The Families and Schools Together programme (FAST) is targeted at families experiencing multiple difficulties, including child behaviour problems.  The programme is based in primary schools.  It is run over eight weeks and comprises weekly meetings involving groups of families.  The programme is distinctive in the way that it aims to build protective factors around children as part of the programme.

A range of positive outcomes has been demonstrated, including reduced rates of conduct disorders, decreased rates of family conflict and decreased rates of social isolation (Sayger and McDonald 1999).  This demonstrates another potential avenue for improving child outcomes, by improving the social capital of families and embedding them more securely in a network of relationships based around the child's primary school.

General lessons about principles for policy interventions

Despite what has been demonstrated by these evaluations, it is clear that much remains to be learned about how best to assist families to enhance their resilience to stress.  There is much to be gained from a careful study of the evaluation literature - especially from studying the detailed working of successful programmes, as well as those that have failed - in order to isolate the particular elements of programmes that are associated with success.

Some commentators have already commenced this task.  Hogue and Liddle (1999) offered the following recommendations regarding standard techniques for implementing family-based prevention models: intensive recruitment and retention; attention to the psychosocial issues of parents; identification of family protective factors; use of normative developmental guidelines; and a focus on parenting practices, using skills-oriented rather than educational approaches.

Ramey and Ramey (1998) also presented a set of principles for effective interventions.  According to their analysis, interventions are more likely to succeed if they begin earlier in the child's development, continue longer, are more intense, are more comprehensive, and are sensitive to families' cultural beliefs, practices and traditions.  However, Ramey and Ramey cautioned that early effects may not be sustained, since the child's normal environment may not be able to maintain the positive progress achieved in the intense phase of the intervention.  In addition, effects may not be realised for all children.  Children from high-risk families have derived more benefits than others.  This suggests that intervention efforts should be targeted to higher-risk groups.

Finally, Durlak (1998) identified a range of risk factors from an analysis of nearly 1,200 outcome studies.  Twelve factors emerged as common and prominent risks to child wellbeing.  These factors existed at different levels - individual, family and the wider environment - and included poor neighbourhoods, ineffectual social policies, poor-quality schools, negative peer pressure, peer rejection, low socio-economic status, parental psychopathology, marital discord, punitive child-rearing, early onset of the target problem, and stress.  Durlak also identified a range of common protective factors present in successful intervention programmes: positive social norms, effective social policies, high-quality schools, positive peer modelling, good parent-child relations, personal and social skills, self-efficacy and social support.  The pervasive nature of these risk and protective factors across such a large number of studies indicates their salience as influences on child development.  Any intervention aimed at promoting child and family outcomes needs to pay full attention to dealing with the issues they raise.

Concluding remarks

This paper commenced with a reference to the "enduring mystery" of family resilience.  The research findings reported here show that there is really no great mystery, since the things that serve to protect families in adversity are basic processes of family functioning, such as communication, togetherness, secure belief systems and good parenting practice.

If there is any magic here, it is what Ann Masten (2001) has called "ordinary magic", a telling phrase which sums up both the commonplace nature of the family processes that help to protect families from the consequences of adversity and the magical way that most families seem to know these things intuitively and to enact them, at least some of the time.  Still, a minority of families struggle with these aspects of family functioning and it is these families to whom attention needs to be directed.

A range of different approaches have been tried to assist at-risk families to improve their functioning and to boost their children's chances of success.  Taken as a whole, the evidence from evaluations of a range of intervention programmes shows that it may be possible to boost the resilience of families, although much remains to be learnt about how best to do this.  Some of the most effective programmes have involved very extensive interventions and have therefore been very costly.  However, this is no guarantee of success, as some programmes that have been both intensive and costly have failed to demonstrate any benefits for participating families or their children.

As is generally true in the area of social policy, there is not likely to be any giant leap forward in terms of intervention design that will make a huge advance over current understanding.  Instead, it will be necessary to undertake careful study of programmes that have demonstrated success - and those that have not - in order to understand which components of a successful programme are crucial to its success.  Knowledge will need to be built incrementally by careful design of interventions based on this understanding and subjecting these to rigorous evaluation.  Although this may involve a protracted effort before results are demonstrated, it is important that the effort is made.  Unless a better understanding is obtained of what works to boost the outcomes of children, not only are public funds at risk by expenditure on ineffective interventions, but (more importantly) future cohorts of children will continue to fail to achieve to their potential.

References

Baumrind, D. (1991) "The influence of parenting style on adolescent competence and substance use" Journal of Early Adolescence, 11(1):56-95.

Brody, G.H., D. Flor and E. Neubaum (1998) "Coparenting processes and child competence among rural African-American families" in M. Lewis and C. Feiring (eds.) Families, Risk, and Competence, Lawrence Erlbaum, Mahwah, New Jersey.

Brody, G.H., Z. Stoneman and D. Flor (1996) "Parental religiosity, family processes, and youth competence in rural two-parent African-American families" Developmental Psychology, 32:696-706.

Campbell, F.A., E.P. Pungello, S. Miller-Johnson, M. Burchinall and C.T. Ramey (2001) "The development of cognitive and academic abilities: growth curves from an early childhood educational experiment" Developmental Psychology, 37:231-242.

Campbell, F.A., C.T. Ramey, E.P. Pungello, J. Sparling and S. Miller-Johnson (2002) "Early childhood education: young adult outcomes from the Abecedarian Project" Applied Developmental Science, 6(1):42-57.

Compas, B.E., J.K. Connor-Smith, H. Saltzman, A.H. Thomsen and M.E. Wadsworth (2001) "Coping with stress during childhood and adolescence: problems, progress, and potential in theory and research" Psychological Bulletin, 127(1):87-127.

Duncan, G.J. and K. Magnuson (2001) Individual and Parent-Based Intervention Strategies for Promoting Human Capital and Positive Behavior, manuscript, Northwestern University.

Durlak, J.A. (1998) "Common Risk and Protective Factors in Successful Prevention Programs" American Journal of Orthopsychiatry, 68:512-20.

Ellison, C.G., J.P. Bartowski and K.L. Anderson (1999) "Are there religious variations in domestic violence?" Journal of Family Issues, 20:87-113.

Epstein, N.B., D.S. Bishop and S. Levin (1978) "The McMaster model of family functioning" Journal of Marriage and Family Counseling, 4: 19-31.

Fine, M.A. and A.I. Schwebel (1988) "An emergent explanation of differing racial reactions to single parenthood" Journal of Divorce, 11:1-23.

Furstenberg, F.F., J. Brooks-Gunn and P. Morgan (1987) Adolescent Mothers in Later Life, Cambridge University Press, New York.

Hawkins, A. and D. Eggebeen (1991) "Are fathers fungible? Patterns of coresident adult men in maritally disrupted families and young children's well-being" Journal of Marriage and the Family, 53:958-72.

Hogue, A. and H.A. Liddle (1999) "Family-based preventive intervention: an approach to preventing substance use and antisocial behavior" American Journal of Orthopsychiatry, 69:287-93.

Jayakody, R. and A. Kalil (2002) "Social fathering in low-income African-American families with preschool children" Journal of Marriage and Family, 64(2):504-16.

Kalil, A., M. Spencer, S. Spieker and L. Gilchrist (1998) "Effects of Family Living Arrangements and Quality of Relationships on the Mental Health of Low-Income Adolescent Mothers" Family Relations, 47(4): 433-41.

Kalil, A. (2003) Family Resilience and Good Child Outcomes: A Review of the Literature, Ministry of Social Development, Wellington.

Knox, V., C. Miller and L.A. Gennetian (2000) Reforming Welfare and Rewarding Work: A Summary of the Final Report on the Minnesota Family Investment Program, Manpower Demonstration Research Corporation, New York (available online at http://www.mdrc.org).

Lazarus, R.S. and S. Folkman (1984) Stress, Appraisal and Coping, Springer, New York.

Luthar, S.S., D. Cicchetti and B. Becker (2000) "The construct of resilience: a critical evaluation and guidelines for future work" Child Development, 71:543-62.

Masten, A.S. (2001) "Ordinary magic: resilience processes in development" American Psychologist, 56:227-38.

McLanahan, S.S. and G. Sandefur (1994) Growing Up With a Single Parent: What Hurts, What Helps, Harvard University Press, Cambridge, Massachusetts.

Mahoney, A., K. Pargament, N. Tarakeshwar and A. Swank (2001) "Religion in the home in the 1980s and 1990s: a meta-analytic review and conceptual analysis of links between religion, marriage, and parenting" Journal of Family Psychology, 15:559-96.

Olds, D.L., C.R. Henderson, Jr., H.J. Kitzman, J.J. Eckenrode, R.E. Cole and R.C. Tatelbaum (1999) "Prenatal and infancy home visitation by nurses: recent findings" The Future of Children, 9(1):44-65.

Olson, D.H. (1986) "Circumplex Model VII: validation studies and FACES III" Family Process, 25:337-51.

Olson, D.H. (1993) "Circumplex Model of Marital and Family Systems: assessing family functioning" in F. Walsh (ed.) Normal Family Processes (2nd ed.) The Guilford Press, New York.

Olson, D.H., Y. Lavee and H.I. McCubbin (1988) "Types of families and family response to stress across the life cycle" in D.M. Klein and J. Aldous (eds.) Social Stress and Family Development, The Guildford Press, New York.

Patterson, J.M. (2002a) "Integrating family resilience and family stress theory" Journal of Marriage and Family, 64:349-60.

Patterson, J.M. (2002b) "Understanding family resilience" Journal of Clinical Psychology, 58(3):233-46.

Pearce, L.D. and W.G. Axinn (1998) "The impact of family religious life on the quality of mother-child relations" American Sociological Review, 63:810-28.

Ramey, C. and S. Ramey (1998) "Early intervention and early experience" American Psychologist, 53:109-20.

Sayger, T.V. and L. McDonald (1999) Evaluation Report for Australian FAST Schools: Report #4, FAST International, Melbourne (available online at http://www.wcer.wisc.edu/fast/research/research3b.pdf).

Scanzoni, J. and C. Arnett (1987) "Enlarging the understanding of marital commitment via religious devoutness, gender role preferences, and locus of marital control" Journal of Family Issues, 8:136-56.

Schweingruber, H. and A. Kalil (2000) "Decision-making and depressive symptoms in multigenerational teen parent households" Journal of Family Psychology, 14:556-69.

Seiffge-Krenke, I. (1995) Stress, Coping, and Relationships in Adolescence, Lawrence Erlbaum, Mahwah, New Jersey.

Steinhauer, P.D., J. Santa-Barbara and H. Skinner (1984) "The process model of family functioning" Canadian Journal of Psychiatry, 29:77-88.

St Pierre, R.G. and J.I. Layzer (1999) "Using home visits for multiple purposes: the comprehensive child development program" The Future of Children, 9(1):134-151.

Tarter, R.E. and M. Vanyukov (1999) "Re-visiting the validity of the construct of resilience" in M.D. Glantz and J.L. Johnson (eds.) Resilience and Development: Positive Life Adaptations, Kluwer Academic/Plenum Publishers, New York.

Walsh, F. (1998) Strengthening Family Resilience, The Guilford Press, New York.