Abstract

Data from a New Zealand longitudinal study (Christchurch Study of Health and Development) show that the risk of welfare receipt in early adulthood rises sharply as the extent of exposure to childhood "adversity" or "risk factors" increases. Cross-sectoral evidence supports the proposition that the formation and maintenance of human capital (defined to include the sum total of an individual’s physical, psychological and economic capabilities) is highly sensitive to the extent of adversity experienced over the life course, and particularly that experienced in early childhood. The implications for social policy which seeks to emphasise prevention, early intervention and investment in human capital formation are explored.

The philosophical and rational underpinnings of social policy are under challenge in Australia, as in many other OECD countries where welfare and other social spending continues to consume an increasing proportion of GDP. Australia’s spending on social security increased from 2.92% of GDP in 1971 (a half century low) to just over 7.35% in 1999.

Uncertainty about the sustainability of the current arrangements is a major impetus for the current re-examination of income support policy, but just as important is concern that the regime may be having unintended and damaging effects. Evidence of "inter-generational transmission" of welfare dependency (Pech & McCoull 1999)¹ is frequently cited as evidence of collateral damage (Murray 1994) and attributed to a combination of:

• moral hazard — the temptation to take hand-outs in preference to work;
• poor role modeling by the parents and their family members;
• the combined effects of high replacement rates² (for a few families), high effective marginal tax rates (for a high proportion of welfare recipients), and fringe benefits (such as health care cards, rent subsidies and transport concessions) which can reduce work incentives, particularly for those only able to obtain low paid work in unskilled or low skilled jobs, or limited part-time work in higher skilled jobs; and
• an institutional environment which fosters a culture of dependency rather than of initiative and self-reliance.

A necessary but not a sufficient condition for the success of a policy in achieving its goals is that it be formulated according to a map of the world which is accurate in its essentials. Applied social policy must be responsive to community values and norms — they are usually reflected in the specific destinations which are chosen and the routes selected for reaching them. However, if the map is wrong, then it is unlikely that an intended policy destination can be reached. If the map is right, but a poor route is chosen, then the destination may be much more costly, hazardous and time-consuming to reach. Moreover, in a world of constant change, a desired destination may effectively be "wiped off the map" before it can be reached.

So an important part of the craft of applied social policy-making is continuous review of the map, the intended destination(s) and the route(s) being traversed so that they can be modified to accommodate new evidence about the world. Desirably, there will be regular and systematic observation of relevant aspects of the world to feed back into the policy -maker’s map.

Over the period which has seen a significant growth in welfare expenditure, there have been many changes which need to be incorporated in the social policy-maker’s map. For example: there have been significant shifts³ in:

• the distribution of work across households (although the overall proportion of the population in work has remained relatively stable since 1945);
• the composition of households, with many more single person households and single-parent families;
• the pattern of employment and asset accumulation over the life-course for both men and women;
• the relative demand and wage level for skilled versus unskilled labour;
• the earnings gap between the top and bottom income groups in Australian society;
• the geographical concentration of disadvantage;
• the proximity of parents to their adult children and grandchildren⁴; and
• the proportion and average age of children in childcare.

In recent years the social policy destination which has been most actively pursued in conjunction with reforms to the Welfare State in English speaking countries, is one in which individuals are resourceful, self-reliant and economically active, families are strong and resilient and both communities and businesses recognise that they need to take an active role in seeking solutions to social problems both on their own initiative and in partnership with government.

There are ample signs that the road to this policy destination is likely to be challenging. The real questions are: do we have the right map and what is the best way to get there?

If the welfare state has failed to prevent poverty, family breakdown, mental illness, child abuse and other domestic violence, child behavioural problems and learning difficulties, is it possible that it has inadvertently contributed to them by, for example, facilitating family break-up, voluntary unemployment and the decline of community? If this be so, then consideration would need to be given to addressing perverse incentives in the welfare system. If, on the other hand, the evidence is that perverse incentives are not the only or even the most important factors, then the best route to the desired policy destination may lie in complementing income support measures with strategies to prevent or ameliorate the conditions which most contribute to these problems.

This paper argues that long term social welfare receipt amongst people of working age should be addressed through cross-sectoral cooperation in strategies which promote human capital formation. Such a strategy would be expected to produce other social and economic benefits.

The fact that not all individuals from disadvantaged, or even abusive, families and depressed neighborhoods grow up to lead lives at odds with social policy ideals, may be taken as evidence that their less successful peers were simply not trying hard enough and should be induced to do so. However, when the pathways leading to good and poor outcomes are traced with the aid of data from longitudinal studies, patterns emerge which make sense of the apparent contradictions.

Four important publications appeared last year which among them synthesize research across sectors on the circumstances or factors which individually or in combination increase the likelihood of poor outcomes (risk factors), or good outcomes (protection or resilience factors) in health, behaviour and economic activity (Marshall & Watt 1999; Homel et al. 1999; Marmot & Williamson 1999; Keating & Herzman 1999). The research described in these four publications suggests a possible answer to the conundrums posed by "successful" slum dwellers on the one hand and middle class "failures" on the other. They also explain why both these outcomes are relatively uncommon.

The socio-economic gradient across all learning, health and behavioural outcomes appears substantially to reflect differences in the degree to which families are able to meet the developmental needs of young children (McCain & Mustard 1999, Wadsworth 1999; Sherman 1999; Power & Herzman 1999) Growing up in an impoverished family is an important risk factor for a variety of developmental problems (Marsh & Watt1999, pp107-114) and is often associated with economic, social and health difficulties in later life⁵. However, a strong parent-child bond⁶ can help overcome some of the ill-effects of economic disadvantage (Ibid. pp120-122). Conversely, a child growing up in a middle class family, who may have access to a high standard of living, may suffer serious developmental problems if the parent-child relationship is characterised by abuse or indifference (Wisdom & White, 1997).

Some risk factors which seem to be common to poor outcomes across social policy sectors include low birth weights, prenatal difficulties, teen pregnancy, large family size and short spacing between children, parental conflict, substance abuse, mental illness, poverty or unemployment, single parent⁷, step or blended family, family isolation, abuse, neglect, or parental indifference⁸. These risk factors are exacerbated by difficult temperament, disabilities or chronic health problems in the child. (Homel et al, 1999, p.136) However, resilience factors, particularly strong attachment between the child and a parent or care-giver, but also including: authoritative parenting; high intelligence; easy temperament; problem solving skills; a sense of self-efficacy; and emotional regulation, together with strong attachment to school or community in older children, appear to lessen the adverse impact of risk factors which are present.(Homel et al, 1999, p.138, Marsh & Watt 1999 p120 et seq.).

As children develop, negative outcomes arising from earlier risk factors can themselves function as additional risk factors for subsequent problems. Sometimes this is represented as a pathways paradigm, with critical transition points throughout the life-cycle presenting both risks and intervention opportunities.

Although some authors distinguish between "psychological capital", "human capital and "social capital" (Zubrick et al. 2000), it can be useful to think of "human capital" as the sum of attributes which an individual brings to interactions with others and to the critical task of regulating him/her self. The dimensions of human capital thus comprise physical characteristics, health status, mental health and mood control, self-image and sense of self-efficacy, resilience to stress (which can be expressed physiologically and psychologically), labour market skills, knowledge, cultural awareness, and capacity for responsive personal relationships. It may also be useful to consider networks of personal support as a factor in human capital rather than just as social capital.

Estimates of the contribution of genetic factors to human capital vary (a range of twin studies cited in Marshall and Watts 1999, pp 40-44 reported heritability ranging from 21% to 100% on selected measures of attention and behavioural problems). However, the data on risk factors and the effects of accumulated lifetime adversity suggest that it is unlikely that molecular biologists will one day locate genes for resilience or welfare dependency. The resilient, resourceful, responsible and interdependent individual idealised in social policy goals can only develop in a real world which enables the development of human capital.

Influences on the early formation of human capital can be pictured as being located within a series of nested spheres which centre on the developing child. During the months from conception to birth all outside influences are mediated through the mother; at birth the delivery room, medical staff, medical technologies, and other family members have their first direct impact; during the early years before childcare or school, the family and its network of intimates are the proximate influences, followed by successively greater exposure to the local community and the broad social environment.

Eisenberg (1999) advocates a paradigm in which nature, niche⁹ and nurture are considered to affect child development. "Niche" refers to the location of a family within the community, its economic and its social status relative to others. This approach seems to be well supported by the research on risk and protection factors. (cf Marmot & Wilkinson, 1999; Keating & Herzman, 1999).

Special risks to the formation of human capital arise when developmental needs in early childhood are not met Possible causal mechanisms are now considered to lie in the biological "embedding" of deprivation or environmental "insults". Deprivation might occur in the form of inadequate nutrition or stimulation or lack of the warm, responsive care that all children need; "insults" could include exposure to conflict or trauma.

Recent neurophysiological research gives support for the concept of critical periods not only for sensory development but also for the foundations of major psycho-social capabilities (Cynadu & Frost, in Keating & Herzman 1999). Children whose needs are not met during these periods may suffer absolute and irreversible deficits (McCain & Mustard 1999).

"Relative" deficits in human capital arise where knowledge, skills or cultural background are not matched with the demands of the environment. In principle such deficits could be addressed.

The few birth cohort studies now underway in Australia are too recent to permit the study of effects of childhood risk factors on outcomes in adulthood. A longitudinal study announced by the Commonwealth Government as part of the National Family Strategy will provide a major source of data in the future. The birth cohort of the Christchurch study, which commenced in 1978 with a sample size of over 1000, has now matured. Data from the study were extracted and analysed to answer the question "Is there any association between accumulated risk factors in childhood and welfare receipt in early adulthood?"

Table 1 shows a list of characteristics significantly correlated with increased likelihood of unemployment between 16 and 21 years. The variables are grouped under major headings: social/perinatal, family economic circumstances, family stability/family functioning; child adolescent behaviours; and cognitive educational outcomes. Many of the factors associated with increased risk of unemployment in early adulthood are the "usual suspects". This exercise illustrates the importance of constructing a rich longitudinal data set — it is difficult to obtain accurate data on new variables retrospectively.

Table 2 gives rates of unemployment by childhood adversity score. The latter is constructed simply by summing the number of risk factors present. It should be noted that this is not a sophisticated instrument — some of the variables are not strictly independent, and there has been no weighting of factors. A more sophisticated scale with well-constructed and weighted risk factors would have greater predictive value. (Similar considerations apply in relation to tables 4 and 6). Nevertheless, the table illustrates how steeply the probability of unemployment rises as the adversity score, represented by the number of risk variables present, increases.

This pattern is repeated in tables 3 and 4. Table 3 shows the factors which increase the chance of having dependent children at age 21. Table 4 shows the relationship between adversity score and probability of having a dependent child at 21, having a dependent child and being on welfare at 21, and being a sole mother at 21, respectively. Here again the probabilities rise sharply as adversity score increases. Note that having a teenage parent is a developmental risk factor in this and other studies (Hardy et al.1997). Having a parent aged 21 was not identified as a risk factor in this study, but having lived in a sole parent family between the ages of 0-15 was.

Table 5 and Table 6 show risk factors for being in receipt of welfare at age 21 and the relationship between adversity score and probability of welfare receipt. Once again, likelihood increases steeply with increasing adversity score.

Data from the Christchurch study may be of greater relevance to Australia than data from countries geographically and culturally more remote, particularly as Australia and New Zealand have similar welfare systems. However, caution still needs to be applied in extrapolating findings such as these which may be sensitive to social and institutional differences. The data are consistent with the findings of longitudinal studies conducted in other OECD countries, which show how persistent the ill-effects of early adverse experience can be (for a wide range of such studies cf Marsh & Watt 1999, Ch.2, Marmot & Wilkinson 1999, Ch. 3; Keating & Hertzman 1999, Ch. 3).

Most of the individuals in the sample had low to "moderate" levels of risk exposure — but although the proportional contribution to welfare receipt was lowest for this group, the numerical contribution was the highest. Growth in unemployment levels has been a major source of growth in social security spending over the three decades. Some of the growth in unemployment may be due to an increase in the level of frictional unemployment, as companies have adopted more flexible hiring policies. Some may be deficits of human capital relative to the high-skilled requirements of emerging industries. Some may be associated with the growing geographical concentration of disadvantage¹⁰ in areas where growth of employment opportunity is inhibited by the low concentrations of social and human capital. These areas may spiral downwards as those able to leave in search of opportunity do so.

The developmental risk factors identified in the Christchurch study are similar to those identified in other longitudinal studies as major influences on health and well-being — ie on human capital. One could predict that, in the absence of intervention, the Christchurch sample will also demonstrate over time the previously observed connections between increased adversity exposure and harmful behaviour, relationship problems, long term welfare receipt, chronic health problems in early middle age and lower life expectancy.

Even if welfare systems have contributed indirectly to the problems of growth in welfare expenditure, many of the people who are most expensive in terms of their use of health, criminal justice and welfare services, are individuals on whom adverse circumstances have literally left their mark in the form of human capital deficits including a vulnerability to stress.

It is not clear that poor individuals would fare better in terms of risk and resilience exposure without a strong social welfare system. However, long term benefit receipt is itself a risk factor for all family members. This may result from factors such as financial strain, low relative income, and/or the loss of self-esteem and social status associated with unemployment. Regardless of background, people who are unemployed and on benefits for a long time show a decline in physiological and psychological functioning, and those whose prior experience has been most adverse (in terms of risk exposure) do worst (Wadsworth et al. 1999).

Interventions to give greater access to life-long learning could assist with human capital mismatch problems. Community or social capital development may boost economic activity and employment opportunity in areas in decline. However, those individuals whose life-time exposure to adversity has been highest may only respond to intensive interventions involving teams of professionals from a range of disciplines.

High levels of socio-economic inequality, to which unemployment is a major contributor, seem to have an adverse effect on human capital formation across a whole society¹¹. It is a moot point whether interventions to address human capital deficits associated with inequality could or should be developed in preference to measures for tackling inequality directly. Government action to reduce inequality through higher levels of taxation and spending has not enjoyed great popularity in western democracies for some decades. It remains to be seen whether and at what point a perceived link between lower taxation, lower spending, and social problems might turn public sentiment around.

"Human capital" is a relatively new concept in mainstream public economic policy — the OECD is only now undertaking work on the development of a conceptual framework and measurement tools for assessing human and social capital and their application to the "knowledge economy". In some ways the concept may seem to add little to what is normally understood as the personal capability of an individual. Its advantage lies in trying to provide a better analytical framework for appreciating the dimensions of individual differences and how they reflect or embody the vicissitudes of life. However, there are dangers — similar to those involved with genetic testing (which could form a component of human capital assessment) - that individuals with known high levels of adversity exposure may be unfairly denied opportunity¹². Risk/adversity is, after all, about probability or susceptibility not destiny.

What the longitudinal data does confirm is that social policy which treats individuals as identical in terms of the assistance they should receive or the contribution which should be expected of them is likely to be ineffective as well as unfair. The Australian social security system, like many others in recent times, is attempting to pay more attention to the family circumstances of welfare recipients and to identify individuals who might benefit from complementary forms of assistance.

The data from longitudinal studies suggest the potential benefit of interventions which help prevent developmental problems in childhood: for example: wider promotion of maternal health and well-being; developmental screening of children especially between the ages of 0 and 5; widespread parenting programs, particularly for parents with multiple risk exposure, enriched child-care for children identified as being at risk; teaching of coping skills/problem solving to both young children and adults at risk, particularly where the adults are also parents; and wider spread access to well-validated relationship skill programs (to reduce damaging family conflict).

There are budgetary and other constraints on the large-scale introduction of such interventions. It may be preferable to trial smaller scale interventions and build up the evidence base on best-practice before moving to larger scale programs. Moreover, government intervention in matters concerning family functioning must also balance public interest, community values and respect for individual privacy. The widespread acceptance of a public interest in being informed about and taking action against domestic violence and child abuse is of comparatively recent origin. Attempted inroads into other areas of family functioning may meet with resistance unless a strong public interest case is mounted and accepted¹³.

Further work is required to elucidate the causal mechanisms behind risk and resilience and the formation of human capital. Potentially cost-effective interventions across social policy sectors need to be developed and evaluated. The National Families Strategy announced earlier this year includes a range of initiatives to strengthen families and communities. The evaluation of new initiatives such as these will provide useful feedback for our map and help identify the best route for achieving further progress toward our desired social policy destination.

¹ There is little Australian data on this question. The paper reports a slightly higher level of welfare receipt amongst teenage children of welfare recipients.
² The Australian system is one of safety net social assistance rather than a contributory social insurance model. The latter tend to have much higher replacement rates. However the Australian system can offer quite high replacement rates for some very large families with low earning capacity.
³ Not all of these changes are intrinsically negative. In some cases, such as growth in the number of workless households there are immediate adverse effects. The impact of some others, such as the increased use of childcare, may need to be assessed.
⁴ Personal observation
⁵ Montgomery et al (1997) show that short stature at age 7 is a risk factor for adult unemployment. Bartley et al (1999) conclude that in a society such as great Britain the effect is still present when material and cultural factors are taken into account. They attribute it to psychosocial stress (op. cit. p 93).
⁶ Or with another adult care-giver.
⁷ Kelly (2000) provides a review of research on the risks associated with sole parent families and concludes that sole parent family structure is not in itself a risk factor, but is frequently associated with risk factors such as exposure to conflict or poverty.
⁸ In adolescence association with an antisocial peer group is an important risk factor, but strong attahcment to attachment to the community (eg.school) acts is protective (Homel et al. 1999);
⁹ From "ontogenetic niche" coined by West & King (1984).
¹⁰ Vinson 1999.
¹¹ The impact of inequality is one of the main messages in Keating & Herzman 1999, Marmot & Wilkinson 1999, and Mc Cain & Mustard 1999.
¹² Would this be simply another form of class discrimination?
¹³ The sensitivity of privacy issues around the family in part prompted Gilbert Steiner to title his 1981 Monograph " The Futility of Family Policy Brookings Institute. The other major consideration was the seemingly infinite variation of families.

References and bibliography

Albee, G.W. & Gullotta, T.P. (1997) Primary Prevention Works Sage.

Bartley M, Ferrie J & Montgomery S M (1999) "Living in a high unemployment economy:understanding the health consequences." In Marmot & Wilkinson (1999, pp81 -104)

Bryson, L., Strazzari, S. & Brown,W. (1999) "Shaping Families" in Family Matters Winter 1999 Issue No 53, pp33-38 Australian Institute of Family Studies Melbourne

DeFrain, J. (1999)"Strong families around the world" in Family Matters Winter 1999 Issue No 53, pp6-13. Australian Institute of Family Studies.

Danziger S & Waldfogel J (2000) Securing the Future: Investing in Children from Birth to College New York Russel Sage Foundation.

Eisenberg, H. (1999) Am. J. Pediatrics v.103 No 5, May 1999, pp1031-1035.

Goodfellow, J (1999) Multicare arrangement patchworks: the multiple use of formal and informal care in NSW Paper commissioned by Office of Childcare, NSW Department of Community Services. Paper is still under Ministerial consideration.

Halford, K (2000) ‘Australian Couples in Millenium Three: A research and development agenda for marriage and relationship education’. Paper commissioned by Commonwealth Department of Family and Community Services.

Hardy JP, Shappiro S, Asinne NM, Miller TI, Brooks- Gunn J, Hilton SC. Adolescent child bearing revisited: the age of inner-city mothers at delivery is a determinant of their children’s self-sufficency at age 27-33. Pediatrics 100, 802-809.

Hawkins JD, Catalano RF, Kosterman R, Abbott R & Hill KG (1999) Preventing Adolescent Health-Risk Behaviors by Strengthening Protection During Childhood. Archives of Pediatrics & Adolescent Medicine March 153 13 p 226.

Homel, R.et al (1999) Pathways to Prevention: Developmental and Early Intervention Approaches to Crime in Australia National Crime Prevention, Commonwealth Attorney General’s Department, Canberra.

Keating, D.P. & Hertzman, C Developmental Health & the Wealth of Nations: Social, Biological and Educational Dynamics Guilford: NY.

Kelly J (2000) Legal and educational interventions for families in custody and access disputes. Paper delivered at 10
^th World Conference of the International Society of Family Law, Brisbane 9-13 July 2000.

Marmot, M. & Wilkinson, R. G eds (1999) Social Determinants of Health Oxford.

McCain, M & Mustard, J.F. (1999) Early Years Study Final Report. The Canadian Institute for Advanced Research. Publications, Ontario, Toronto.

Marshall, J & Watt,P (1999) Child Behaviour Problems: A literature review of its size and nature and prevention intervention has in childhood Perth, Western Australia, The Interagency Committee on Children’s futures.

Mental Health Promotion and Prevention National Action Plan under the Second National Mental Health Plan: 1998-2003. Publications — Production unit. Commonwealth Department of Health and Aged Care: Canberra.

Montgomery SM, Bartley MJ & Wilkinson RG (1997) Family conflict and slow growth. Arch. Dis. Child. 77, 326 30.

Mrazek, P.J & Haggerty, R.J. (1994) Reducing risks for mental disorders: frontiers for preventive intervention research, National Academy Press, Washington DC.

Murray C (1994) Underclass: The Crisis Deepens IEA Health and Welfare Unit London

O’Leary, K.D, Malone, J. & Tyree, A. (1994) "Physical depression in early marriage: pre relationship & relationship effects". Consulting and Clinical Psychology 62, 594-602.

Oberklaid, F., Foley, D., Goldberg, S., McLoughlin J., Nagoreka, J & Wake, M. (2000) A review of the Early Childhood literature Paper commissioned by the Commonwealth Department of Family and Community Services.

Pan, H.S., Neidig, Pott. & O’Leary, K.D (1994) "Predicting mild and severe husband-to-wife aggression" Consulting in Clinical Psychology, 62, 985-985.

Power,c & Herzman,C (1999) Health, well-being & coping skills in Keating & Herzman,1999.

Rodgers, B.& Pryor, J. (1998) Divorce and separation: The outcomes for children Joseph Rowntree Foundation York

Russell, G. & Bowman, L (2000) Work and family: current thinking, research and practice. Paper commissioned by the Commonwealth Department of Family and Community Services.

Russel, G. and Edgar, D. (1999) "Organizational Change and Gender Equity: An Australian Case study", in Haas, P. Hwang & G Russel, Eds., Organizational Change

and Gender Equity: International Perspectives on Parents at the Workplace. California: Sage Publications.

Schweinhart,I.J. Barnes, H.V. & Weikart, D.P.(1993) Significant benefits: the High/Scope Perry Preschool Study through age 27 years. High/Scope Press,Ypsilanti.

Sherman, L.W. (1997) Thinking about crime prevention in Department of Criminology & Criminal Justice, University of Maryland Crime Prevention: what works, what doesn’t work and what’s promising. Cited in Homel at al 1999.

Straus, M.A. & Gelles, R. Steinmetz (1986) ‘Societal change and change in family violence from 1975 to 1985 as revealed by two national surveys". Marriage and the Family. 48, 465-479.

Straus, M.A., Gelles, R. & Steinmetz, S.K. (1980) Behind closed doors: Violence in the American family. NY. Doubleday.

Tomison A.M. (1996) Intergenerational tranmission of maltreatment — Issues in Child abuse Prevention National child Protection Clearing House Issues paper No. 6

Vinson T. (1999) Unequal in Life: The distribution of Social disadvantage in Victoria and new South Wales The Ignatius Centre for Social Policy Research.

Wadsworth M. (1999) Early Life in Marmot & Wilkinson, 1999

Wadsworth M.E.J. Bartley M.J. & Montgomery S.M. (1999) Unemployment, health and human resources. Soc. Sci. Med. (forthcoming when cited in Marmot & Wilkinson, 1999)

West & King (1984) cited in Eisenberg 1999.

Wisdom, KS& White, HR (1997) Problem behaviours in abused and neglected children grown up: prevalence and co-occurrence of substance abuse, crime and violence. Crim. Behav. Mental Hlth 7,287-310 cited in Wadsworth, 1999,p 49)

Zubrick, S.R, Williams, A.A., Silburn, S.R. & Vimpani, G (2000) Indicators of Social and Family Functioning. Paper commissioned by Commonwealth Department of Family and Community Services. (to be released)