The United Nations International Year of the Family 1994 (IFY94), gives us an opportunity to bring together threads of social life that have been treated separately and disjointedly.

The rights and protection of children cannot be seen as separate from the context of family life. Care of the disabled, the aged, the mentally handicapped must link public and private responsibility. Youth homelessness, alienation, training and income support cannot be treated in isolation from family forms and changing resources. That most basic shift of modern societies to an assertion of the equal rights of girls and women arises from the repressive structures of family life and demands a rethinking of the nature of family life.

So IYF94 offers us the chance to examine both the benefits and the costs of different family forms, to particular groups in society and to the individuals who together constitute any one family.

THE PROBLEM OF DEFINITION

There is a danger in talking about the concept of 'the family' of elevating an abstraction, a Weberian 'ideal type', to a status of reality it does not have. My preference would be to speak of 'families' and of 'family policies' in the plural rather than the singular.

Because, too, 'the family' carries in each individual's mind so much emotional baggage, and because the term 'family' carries religious and political ideological overtones, IYF94 may well degenerate into the sort of definitional debate that destroyed the US White House Conference on Families in 1980.

Three ways around this definitional quicksand can be suggested.

GENERIC AND SOCIAL PARENTS

First, we must remind people that 'every individual has a family' (of origin at least).

An infant is born of woman and has a genetic father. These two adults may or may not live together but the child's family is defined by that biological union. Both the mother and the father have kin who form part of the child's extended family. Again, living together or not makes no difference to that concept in relation to the child as an individual.

Where a child is adopted by non-genetic parents, it has both a genetic family and a social family. The distinction is an important one, for most societies have made provision for social parenthood even where the blood father could not be identified. In societies disrupted by war or other social chaos, many children are separated from parents and their extended families. In a day to day sense they then have no family and societies try to provide substitutes in the form of orphanages, shelters, adoption, foster care and so on. This is one of the major tasks for the implementation of the World Declaration on the Survival, Protection and Development of Children and Plan of Action for Implementing the World Declaration on the Survival, Protection and Development of Children in the 1990s.

As each individual grows, their social relationship with 'the family' changes. A child may become homeless, its parents may separate, divorce, repartner or remarry. Young adults leave the family of origin to lead an independent life. If they live alone, they still have a family and so family policies need to consider their family obligations, responsibilities, reciprocal caring relationships.

Once an adult partners in a sexual relationship they form a sociologically distinct category - the potential 'family of procreation'. This is sociologically prior to the concept of 'marriage', though legal marriage may be the only socially legitimated way in which sexual partnerships can be formed. Sexual regulators such as incest taboos, courtship rituals, religious and civil marriage laws are in fact social mechanisms for the control of sexual behaviour, and human history is rich in the varieties of mechanisms for controlling sex and the care of children who come from that union. It is ironic that so much so-called theory of the family ignores the sociology of sex and intimate behaviour. Yet such behaviour underpins the emergence of family group culture and the structures of care in society as a whole. Sexual life is obviously central to the gender differentials of social structure.

Even when an adult does not partner and have children, he or she usually still has a 'family', of parents and/or kin, and family policies cannot ignore the 'single' adult. Indeed, especially for single women, the burdens (and satisfactions) of family care - of the aged, the disabled, of other people's children - have been central to many 'single' lives.

Finally, the couple who do have offspring and live through the child-rearing phase will usually find themselves alone again once those offspring have matured, left home and formed their own families of origin. Are such couples not 'families' because they have no children living with them? Is the aged widow or widower to be ignored when we speak of family policies because they no longer live within a unit defined as a family unit through the presence of children? Is the single or divorced mother with children not a family? I suggest not.

LIFECOURSE CHANGES

This brings me to the second point which may keep the definitional problem at bay. That is, 'the family' changes throughout the life course and changes through history as social circumstances change.

So no one definition of family will suffice if we are to understand either the nature of family life or the parameters of family policy.

Both facts are obvious but often forgotten when ideological agenda are being pursued. Most families do take the form of the so-called 'nuclear family' - mother, father and dependent children - at one stage of the life course. During infancy, they often also take the form of the 'breadwinner/housewife' family model. That is, the mother focuses on the caring tasks and the father on earning an income. But this too is very ethnocentric and historically shortlived.

The biological differences between male and female clearly have predisposed most societies to have men as the aggressor/protector of women and children. But the physical weakness and vulnerability of women before and immediately after childbirth do not explain the social extension of that physical power differential into the complex gender systems of patriarchy and female subordination that are now being challenged.

Indeed, it must be remembered that most societies have required both men and women to share the tasks of providing food, earning a living, protecting infants, and there have always been non-parental carers who have substituted for parents in caring for the young. In many UN countries, women are not only the main carers but also the main providers and income-generators. The place of men in these social arrangements is under challenge, as it is in advanced societies where greater equality and income independence enable women to reject a position of subordination.

Separation, divorce and remarriage are part of the life experience for increasing numbers of family members. That restructuring makes them different, but not necessarily less important, families whose needs must be considered in family policies and family support programs.

My point is simply that 'the family' cannot be treated as a static entity defined according to traditional structures or moral preferences.

BEYOND THE FRONT DOOR

A third way around definitional narrowness is to remind people that the family 'does not stop at the front door'.

This is more than just a matter of the unit of analysis one is using in family research or family policy. Most support services for families and most tax and income security transfers are targeted at the household unit. This may be convenient, but it is conceptually only part of the picture. Income transfers between the generations flow in both directions; a household may have no children present but is supporting offspring living, studying or working elsewhere; a single-person or couple-only household may appear to have few family obligations, yet is supporting aged parents or other family members living separately.

But as well, family life is continually centred around the emotional life of family culture, often built up across the generations and sustained by contact between and across several related family households. 'Intimacy at a distance' still has an impact on the way apparently isolated household units construct family life. Policies and programs which ignore such links will very likely miss and/or misconstrue real family needs.

Given the above provisos, how can we conceptualise family life so that its complexities do not prevent us from identifying clear directions for family policies?

I propose the accompanying diagram (Figure 1) as a broad starting point for thinking about the tasks of 'the family'. Later, I will set this within a broader social context (Figure 2) and outline briefly the main parameters for evaluating family policies (Figure 3).

Proposition 1: The family unit is a system of cooperation based upon combining human and other resources and a structured distribution of costs and benefits

1.1 Because human and other resources are unequally held, there will be competition, conflict, negotiation and inequity in any family system.

1.2 However, a family unit is formed on the basis of joint action for mutual benefit, so the necessary condition for family formation and family maintenance is cooperation and interdependence between its members.

1.3 The resources combined in a family unit are personal, social and material/economic.

1.4 Personal resources include the various intelligences, competences, skills acquired through childhood development into adulthood of each partner. They also include emotional resources, character, sexuality and the gender-structured differences in skills produced by society. The underlying capacity is intra-personal intelligence, self-understanding, the sense of control over one's own environment that comes with growing competence.

1.5 Obviously that sense of the competent self is in large part socially constructed. Gender and class differences constrain the capacity for self-control and for being effective in one's environment.

1.6 Social resources are partly personal, in the capacity to understand and empathise with others - what is called inter- personal intelligence - and the capacity to manage and negotiate with others in social situations.

1.7 Social resources also flow from the status and class of one's family of origin, kin, ethnic, religious and occupational groups. Family life chances are dramatically affected by the back-up resources (either transferred in real terms or potentially to be drawn upon if needed) available to each family unit. Family units isolated from such wider family support have to draw more heavily upon their inner resources, or look for public 'welfare' supports. Social resources include advice, know-how, wisdom, moral support, group solidarity, sense of security.

1.8 Material resources are obviously vital to family maintenance and living standards, the sine qua non of survival. Shelter, housing, quality and security are essential, as is adequate food and clothing. Access to such resources usually depends upon income as the basis of exchange for goods and services the family cannot provide for itself. What we often forget is that the all-embracing need of both parents and child is for food and/or the economic means to provide food, clothing and shelter. We forget that from time immemorial, children have had to accompany their parents to 'work' in hunting, gathering, in tending the animals or tilling the fields, or else they have to be left in the care of others, older children or other adults.

1.9 Family structure is closely linked to family resources and living standards. A two-parent family has two adults to share (as they and the society choose or dictate) the material and economic tasks. Many children can be both a resource and a drain on resources. Separation and divorce have major impacts on family living standards, the most prevalent ones being the poverty of women and children. Having access to extended kin networks is also obviously important here and developed societies with high mobility do not necessarily mean a diminution of intergenerational material exchange and other forms of support. While smaller family size in the so- called nuclear family may mean fewer people resources to draw on, greater longevity often means access to three generations of potential support.

Proposition 2: The family system of cooperation focuses around two main sets of tasks: income-generation and care

These are not 'functions' in a Parsonian goal-oriented sense, but they incorporate and put in context the differing life experiences and living standards of all families. Family unit organisation and 'management' of these two tasks depends upon the type of cooperation, negotiation, exchange and conflict that is developed within each family unit and within each social group's normative family culture.

2.1 Income production or generation depends crucially upon the family members' capacity to work and produce goods that can be exchanged or to earn wages as the means of exchange. Obviously this is not merely a matter of personal skills, but also of the economic system in which families have to operate (see Figure 2 and the propositions which follow). If the labour market no longer requires unskilled labourers or is over-supplied with professional lawyers or teachers, some families' capacity to earn an income is reduced. If unemployment is high, or social chaos disrupts the production system, family survival is at risk. The value of domestic labour is often forgotten in discussions of both these tasks; domestic labour is, of course, both productive and caring but not income producing.

2.2 The other side of income generation is its distribution within the family. A major change in Western family life has been the shift from families as productive units to being, in the main, units of consumption. The status and treatment of children also reflects that shift as children consume more than they produce in return and the cost-benefits of investment in child-rearing are delayed and the value of that investment is obscured.

2.3 Families combine resources and earn income in both socially structured and idiosyncratic ways. Both parents and children may 'work', both inside and outside the home. Male income-generation may be separated off from family work as a more complex division of labour develops. The unpaid work of women in the home may or may not be valued by the society and other family members. As women also enter the paid labour force, household work becomes a double burden, is reduced, is done by non-parental carers and home help, separates women as well as men from family life.

2.4 Income-generation is a key factor in the power differentials of family life. Where paid work is the exclusive preserve of men, women are locked into dependency, have restricted autonomy and decision-making power, cannot escape from a violent or unsatisfactory relationship, are constrained in the capacities they can develop. The expectation that women will be 'provided for' by a male builds into a whole set of normative expectations about education, role preparation, deference and self-esteem which may be sustained by the social structures of education, work, mass media.

2.5 Opening up the options for women either through market work or social security payments inevitably challenges the power balance within family life. The nature of joint action within the family system of cooperation becomes open to renegotiation of formerly accepted statuses, and the distribution of family resources, costs, benefits and rewards alters. Such conflict may be both damaging and productive of positive change. Many divorced women report improved quality of life despite reduced family income because they now have at least a reliable income and one over which they have control.

2.6 The dynamic nature of how family units negotiate joint action to produce economies of scale and mutual benefits suggests that the family is not a passive victim of social change. Rather the family is the maker of change (familia faber). It is in the dynamics of family negotiation that labour market demands or legislation about equal opportunity are forged into a behavioural reality that has lasting impact. The 'fallout' may be family conflict, violence, disruption, separation, but couples faced with the task of jointly earning an income and caring for children actively forge a new agreement which embeds an acceptance of new values and ways of coping.

2.7 The other main task within the family system of cooperation is that of care. Care involves both 'caring about' and 'caring for' and both must be kept in mind when thinking about the structures of care in any society.

2.8 'Caring about' is the more important but often ignored side of the family's care task. Because the family is a primary group based on emotional affiliation rather than a formally organised secondary group, members care for one another essentially because they care about one another. In fact it is because we care about, are emotionally involved with, our family that the main burden of caring for those in need falls on the family unit. It is too easy to forget that the family is the major 'welfare' or 'support service' network within society. This too becomes socially structured via gender differentiation to mean that female family members end up doing more of the caring-for tasks, with males viewing their income-earning role as their side of the caring- for equation. This does not however mean that men care less 'about' their wives, children and other kin.

2.9 Care must also be differentiated in terms of the 'object' of care. This may be the self, one's spouse, one's children or other dependants.

2.10 Caring about and for the self is an oft-neglected aspect of family care. Ideologies of self-sacrifice, service, self- effacement can produce self-respect but can also engender self-denigration and inequality in family relationships. The child who is brought up in what has been called a 'vicious' cycle of socialisation rather than a 'benign' cycle, will approach new tasks and other people with an expectation or fear of failure rather than success; they will not have that sense of efficacy or control that helps assert the needs and wishes of the self in negotiations with other people. That is why the sociology of competence and the social psychology of self-esteem are so important in the study of family life. It is also why the experiences of early childhood development need close attention in family policies because the foundations of self-respect are essential for adult negotiations within the family system.

2.11 It is in the family that this basic self-understanding and self- respect are developed. Interaction systems which set arbitrary limits and authoritarian disciplinary patterns for children narrow the concept of self and its capacity to negotiate with others. Permissive child-rearing encourages excessive self-centredness, sets few limits and thus distorts the social nature of selfhood and the need for reciprocity in social relationships. A more balanced setting of limits with encouragement of self-control teaches the child about the nature of social cooperation. That is why the family unit is so central to civil society, for if a child is not socialised to understand the rights of others in the family it will have little respect for the rights of social 'strangers'. If a child is abused, treated as an inferior, of lesser (usually female) status, its prospects for later self- assertion and equal reciprocity will be seriously damaged.

2.12 The family is, in that sense, a key educator in the politics of social relationships. Through family structures, cultural values are reproduced and the culture's taken-for- granted 'recipe knowledge' about social structures is conveyed. The family power structure reflects the power structure of society and its varied interest groups. The social 'place' of self in relation to 'Others' is embedded in early childhood family experiences. The presence or absence of a father and the way he behaves in relation to his spouse and children will be a major influence in this process. So caring for others has its origin in family life.

2.13 As well as self-regard, the caring tasks of family life can be divided into caring about and for one's partner/spouse and caring about and for one's related dependants (children, the disabled, the aged).

2.14 Caring about a partner grows out of intimacy and the gradual accumulation of a shared group (couple) culture. This is commonly called 'love' but it has wider connotations than romantic love. Even in an arranged marriage not based on free choice of partner or on 'love', the very nature of the sexual bond usually results in a sense of mutual caring, shared understandings and 'secrets' that set the couple apart in a special socio-emotional category from other kin and associates.

2.15 This caring about also usually involves caring for, but this varies with religious, ethnic and, above all, gender differences. The mother-child nurturing bond colours expectations of who will do most of the caring-for others in the family. Patriarchal values define male caring-for tasks in terms of earning an income, the 'family breadwinner' role. This is not to be trivialised, for if that role is not fulfilled adequately in a society where employment options are gender-based we suffer 'the feminisation of poverty' or 'the pauperisation of motherhood'. Even in industrialised societies with a high proportion of married women in the labour force, men remain the principal 'breadwinners' because of salary differentials and inadequate provision of non-parental child care.

The imbalance of affective caring reflects the differential socialisation of males towards dominance through instrumental tasks and of females towards socio-emotional group maintenance tasks. Attachment to infants is likely to be stronger for females because of time and task allocation, but it is important to be clear that 'bonding' is not merely a function of biological motherhood; it is socially constructed and thus children can bond strongly to others (including males) if given appropriate child-rearing conditions.

2.16 Caring about and for dependants however is closely gender- based. Time allocation of males to the income- producing task has its reciprocal in the burden of care in the day-to-day caring-for sense being placed on women. Child care, both in the home and in paid or unpaid non- parental care situations, is typically done by women. Physically or mentally handicapped family members are cared for by women. Aged care, as parents become weak and dependent, is usually done by one woman, often one of several daughters (or even daughters-in-law), adding to her other tasks of caring for a male spouse and for offspring now adult but still calling on her for support.

2.17 So when we assert the family is a system of cooperation, pooling resources and distributing the costs of performing the two main tasks of income-generation and care, it must be clear that the negotiation about task allocation is often unequal and based upon the gender- based resources of unequal partners and other family members. Despite that, the family unit is a joint system of action with mutual benefits and economies of scale.

Proposition 3: The family system of cooperation results in a duality of joint benefits and unequal rewards.

This is clear in the relative living standards of different family units and of different members within each unit.

3.1 At one level, we can speak of 'family living standards' by measuring combined family resources - income, housing, other assets including time for leisure - and a shared quality of life. But clearly there are 'trade-offs'. The male trades off earning an income against leisure time. The female may have more leisure time but lacks access to and control over income. If she earns an income as well as running a household and rearing children, her time for other pursuits is severely curtailed. In the distribution or consumption of family-produced benefits and rewards, parents (especially mothers) may sacrifice their share for the sake of the children. The goods, services and leisure produced by each family system are unequally distributed and consumed.

3.2 Separation and divorce affect family members' living standards very unequally. The male usually leaves with a share of property and well-developed income-earning career capacities. Because of home-making tasks, women leave with some property share but far less ability to earn an independent income. Children, usually in the custody of the mother, thus enter with her an impoverished period, perhaps ended on repartnering but with often long-term damage done to their own life chances. The degree of State support for lone parents and proper enforcement of child maintenance will lessen or exacerbate that change in living standards.

3.3 So when measuring family living standards, we must differentiate between individual shares and differences between the quality of life or wellbeing of separate family members.

3.4 The shift towards a more gender-equal family partnership model and attempts via the 'welfare state' to redistribute income and access to such 'social wage' services as education, health and other support services alters the nature of internal family negotiations about the social distribution of costs and benefits.

This leads us to the second and more macro-level depiction of the family's place within society.

Proposition 4: The nature of family life and the nature of the civil society in which it is embedded depend upon but also contribute to the balance between state and free market coordination of social structures

4.1 Put very simply, modern developed societies are coordinated by the actions of both the state and the market economy. Their relative autonomy affects the benefits and costs that accrue to individuals and families within civil society as well as our identity as a nation.

4.2 As we have seen in eastern Europe and elsewhere, excessive coordination of society by the state can degrade civil society, violating the adaptive capacities of ordinary social life and threatening the social reproduction of both culture and individual identity. Coordination becomes an end in itself, unmediated by publicly affirmed and renewed norms of morality.

4.3 On the other hand excessive laissez-faire also degrades civil society because it is then 'colonised by market forces' (Pusey 1991). Culture and identity dissolve into arbitrary and individual choices and individual calculations of utility, with the result that public and private elites overwhelm the independence and autonomy of the masses. In trying to restructure unwieldy and unproductive economic systems, the economic rationalists who seem to dominate economic theory in free market economies, may have made the fatal slip of treating society itself as the enemy. They have put a system above the people it is meant to serve.

4.4 Just as state control and an overblown welfare state degrade the integrity of the individual and the capacity of civic groups to negotiate and cooperate within a reasonably democratic and equitable balance, so would free market nihilism endanger the reproduction of society and the very identity of a civilised nation. It is only by ensuring the integrity of civil society that we can keep a reasonable rein on the relative evils of state or market coordination.

4.5 Social democratic systems have attempted to preserve some balance between 'free' market forces and the coordinating mechanisms of the State. As Durkheim long ago pointed out, a complex division of labour is based upon, and requires for its successful operation, greater coordination because the parts of the system are so inter-dependent. Cooperation, negotiation, compromise, reciprocity are the basis of exchange and social coordination. Civil society, including families, reflects that more open, inter-dependent relationship, is indeed the very basis of a non- repressive, restitutive system of social justice and economic exchange. Such inter-dependence means the inevitability of reciprocal rights and obligations.

4.6 Unfortunately, the dynamics of the marketplace have been sustained by the language of 'familism', by the 'ideology of separate spheres' which accepts profit as the key operating principle and exonerates the public sphere from obligations of social responsibility. As Hooyman argues, when care-giving values became isolated in the home they lost their centrality in the wider public sphere.

I shall return to these issues when discussing the core elements of family policy.

Proposition 5: Social policies should have as their central objective the maintenance and improvement of family wellbeing. By definition, then, 'family policy' involves a focus on the family impacts of other social policies.

5.1 If we are to formulate adequate family policies we must look at all social policies in terms of their impacts on family wellbeing and not permit a false dichotomy to be drawn between 'economic' policies and 'social' policies. Figure 3 may serve as a way of guiding discussion about this general proposition, bearing in mind that the absence of deliberate family-oriented policies is in itself a policy choice which has measurable impacts on the quality of family life.

5.2 As the discussion in Section 4 demonstrates, a nation's approach to economic policy is a crucial factor in how family units manage. Dumon's 'enabling' policies are essentially those that support the economic and material wellbeing of families. If family income depends upon wage labour, access to jobs and income levels will be very different in a free market economy from in a centralised welfare state. Regulation of pay awards, working conditions, hours of work, access to education and training, the provision of parental and other forms of leave, all enable or disable families in their survival and caring tasks. Countries short of male labour (as in times of war) will employ women and offer other forms of family support. Policies about housing interest rates, private ownership versus public rental, taxation on private assets and income tax rates will all affect family production, consumption and living standards. Other forms of income maintenance and subsidy such as family allowances (for children), disability and other benefits, pensions, means-testing of eligibility for publicly-subsidised services are another vast area of economic 'enabling' policies which need to be assessed for their relative impacts on different family types at different stages of the life course.

5.3 Each society will have its own degree of integration or coordination of these economic and material wellbeing enabling policies. There is no a priori benefit in greater or lesser coordination, but an overall evaluation of how one aspect interacts with others is likely to be useful as, for example, in avoiding poverty traps or denying access to employment for specific groups of people such as women or youth.

5.4 The second set of family policies crucial to family wellbeing can be called 'family functioning and support services policies'. Here the central philosophical dilemma is the extent to which family 'autonomy' can be and should be modified by public 'interference'. The very concepts used here reveal the problem. Few family units have ever managed to survive 'alone', having normally to combine with others for protection, production and exchange of expertise. Most societies have provided non-parental care for children to enable both parents time for survival and production tasks. So private family 'autonomy' has always been circumscribed and is even more so in a complex division of labour. Yet the ideology of privacy, freedom of choice, family autonomy and rights is very strong and can prove dysfunctional in several ways.

5.5 The language of discussion thus needs to move towards a more explicit public-private 'partnership' model of family support, so that responsibility for survival and wellbeing is seen not as that of the private family alone, but also of the wider community and the government, employers and other institutions whose actions impinge on family life. Too often the public side is conceptualised as the State welfare bureacracy. Not often enough is attention paid to the damaging structures of work, urban design and schools which could and should be more 'family-friendly'.

5.6 Preventive support services need stronger emphasis and the longer term cost-benefits of prevention need to be documented. Family support services tend to be designed as remedial or crisis- intervention stopgaps which are inevitably more intrusive on family autonomy and more costly than education and preventive universal family support.

5.7 Services such as parent education, nutrition and health support during pregnancy, family planning, family mediation, marriage counselling or family therapy; school courses in human relationships, conflict management, sexuality, home economics and child development all have a broadly preventative role.

5.8 Child day care should not, in my view, be seen as 'family replacement' but as support for parenting and family functioning. More importantly, child care should be seen as child-development services where the 'public parent' (including the mass media) supports private parents in ensuring the full development of every child's potential. Service support for parents in the home in their parenting tasks needs as much attention as formal child care services away from the home.

5.9 The issue of family isolation needs closer attention in the context of mobile urban societies. Kin networks of support are harder to maintain and the active family-as- maker requires new skills of information-collection and organisational family management that could be better facilitated. Targeted, means-tested, access- restricted, crisis- oriented services are stigmatising and counter- productive to the public-private partnership approach. Family resource centres in easy local reach and open to all would be more effective in linking separate families to information and advisory support services. The role of family isolation and privacy in sustaining domestic violence and child neglect/abuse calls attention to the need for new urban centres to sustain a caring civil society.

5.10 Clearly the legal system plays a part in regulating the public- private balance. Yet much of the literature on family policy ignores the role of the law. Here, I refer not only to marriage and divorce laws, but also to those laws relating to adoption, child welfare, foster care, neglect and abuse that circumscribe the limits of family 'autonomy'.

5.11 The law not only incorporates what have become accepted social principles and practices; it is in turn a vital legitimator of social change.

5.12 The new divorce laws promulgated in most Western societies in the 1970s legitimated and codified the shift that had already been taking place towards a less institutionalised form of 'companionate' marriage based on the quality of relationships. They also asserted two new principles that contradicted the older view of marriage and which must be kept in mind in debates about the family and society.

5.13 The first was the principle of equality: between men and women in terms of their joint and equally valuable (though different) 'contributions' to the two core tasks of marriage; in terms of their joint (and continuing) responsibility for the support of children; and, usually, in terms of their equal right to a share of matrimonial property upon divorce. As well, many countries asserted the equality of children whether they are born of the marriage or outside it.

5.14 A second principle enshrined in the new family laws was that of neutrality, of the State and the law in relation to the organisation of emotional and private life. No-fault divorce put an end to State condemnation of private behaviours as a 'cause' of marriage breakdown, though several nations still wrestle with the concept of 'fault'. In parallel, States have developed new legal approaches to the criminality of family violence, whether spousal or against children.

5.16 Nevertheless, rather than guaranteeing preservation of a particular model of marriage and family order, the new laws merely regulated individual and collective interest and legitimated the diversity of existing couple situations. They recognised the affective basis of marital relationships as opposed to an institutional conformity to roles.

5.17 These principles of equality and neutrality in relation to private approaches to marriage are echoed in other areas of the law. Equal pay provisions in employment, human rights and equal opportunity legislation, the legalisation of abortion and removal of the notion of 'illegitimacy' for children born out of wedlock, even policies of anti- discrimination and multiculturalism, all reveal an institutional legitimation of freedom to choose one's own lifestyle, to be entitled to respect, a removal of the State's role in enforcing institutional conformity to any one defined family 'model'.

5.18 What this means in effect is that both within-marriage relationships and state-family relationships become permanently negotiable. Law and justice no longer guarantee marital and family order the stability of a particular, prescribed kind. Each citizen is free to negotiate his or her family, his or her law. Indeed the new rhetoric of mediation and conciliation presses towards private negotiation and conflict resolution, away from litigation based on formal rules or roles, with the law acting merely as a helpful intermediary. The message is that couples in separation, as in marriage, are themselves responsible for decisions affecting their lives and the lives of their children. But there are some heavy provisos.

5.19 The central problem with all this adult autonomy and negotiation is the place of children and the obligations of parenthood. Such trends conflict with the time and resources that must be invested in 'quality' children. The State must intervene in the economic and child-related consequences of marriage breakdown and in the problem of dependency more generally. Much more attention needs to be given to the unintended consequences of such legal principles for separate family members.

5.20 Another crucial element of family policy discussions which seems to me to be lacking is the overall philosophical purpose which should guide decision-making in this area. Too often, family policy is seen as synonymous with pro- natalist policy, or social security policy, with 'family welfare' or 'family wellbeing' as the goals. Few writers go beyond to ask why is family wellbeing important as a goal. Or discussions focus on equity as a goal and thus look at family income/resources as the key elements of family policy, without asking what are the expected 'outcomes' of equity between families, other than the broad outcome of 'social justice' seen as a goal in itself.

While I have no problems with any of these as crucial elements of any family policy analysis, I believe we need to look further at how 'the family' and the quality of family life relate to wider social outcomes such as the development of a sense of moral obligation which is the underpinning for society itself.

The problem is to explain how any society can ensure the positive transference of a sense of obligation to self and close intimates to a sense of moral obligation to others, to 'strangers' in a sociological sense.

5.21 As indicated in Section 4, much of the debate on relations between the modern liberal State and the family rests on a simplistic dichotomy between public control and private freedom. Claims to independence, autonomy, individual rights, are made strongly along with criticisms of dependency, regulation and public interference in private life. Yet in any complex society people are dependent on others to make that society work and no individual can survive alone.

As Wolfe (1989) puts it, 'a people who are completely free are a people unencumbered by obligations, whereas economic growth, democratic government, and therefore freedom itself are produced through extensive, and quite unencumbered, dependence on others'. Such interdependence means the inevitability of reciprocal rights and obligations.

5.22 The dilemma of our modern liberal democracies is that we defend our freedoms but are confused about the social and moral obligations that make that freedom possible in the first place. So it is necessary at the outset to assert that no society can exist without a moral code, a set of rules that define people's obligations to one another. Moral obligation in a 'traditional' society is 'easy' in that the rules are known to others and individuals are not called on to act as moral agents since the rules of social interaction are predefined. People in modern societies, by definition more affluent and more complex, suffer the agony of choice - society poses multiple decision points, conflicting vested interests, and leaves the individual much wider margins to deviate from or oppose the rules of authority. Sociologists since Durkheim have pointed to the consequences of 'anomie', normlessness or 'estrangement', where individuals float free of others and lack that sense of connectedness that would give them a feeling of common purpose and mutual obligation.

5.23 The family as an institution has become the main vehicle through which such dilemmas are expressed. That 'private' sphere of life has increased its salience via the intensity of emotional commitments and high expectations (a haven in a heartless world) while at the same time losing some of its salience (or changing its nature) as a mechanism of social reproduction.

As the first and most emotionally intense social context of children's lives, the family teaches us to know ourselves, teaches us our obligations and rights in relation to intimate others and teaches us our place, our status in the pecking order of society. The power play of intimate relations is a 'global' obstacle that must be coped with and converted into rules of social behaviour that may apply to more specific, new relationships with less intimate others, with strangers beyond the family itself. Yet the more private, the more idiosyncratic becomes the ecology of family-based moral learning, the less generalisable the lessons become, the fewer are the agreed-upon guidelines for organising moral rules that can help us meet our obligations to the 'strangers' upon whom we depend in a complex modern society.

5.24 In a sense, the social sciences have replaced other institutions such as religion as the theatre of moral debate. When people are insecure about whether they are behaving correctly towards others, social scientists become the 'savants' because they can describe what people actually do and they have a theory about how people should act toward one another. We rely increasingly on 'experts', becoming in Habermas's sense a 'therapeutocracy' in which professional expertise comes increasingly to substitute for family autonomy.

The difficulty is, we face at least three broadly competing sources of social scientific moral guidelines: economics, political science, and sociology. Corresponding to these are three sets of institutions or practices charged with the maintenance of moral responsibility: those of the market, the state, and what was once called civil society. Each of these is instructive when considering family policies and their key outcomes in social behaviour.

5.25 All three are theories of regulation as well as theories of freedom. The economic approach says society works best when people can maximise rationally their self-interest. Rather than central state direction to coordinate economic activity, the technique of the market place (voluntary cooperation) should be used. Of course the market is anything but a voluntary mechanism for organising obligations to others. Indeed, prices and other trading mechanisms constrain individual desires and coordinate the market itself, for the collective good. Political science argues that more is needed to constrain private interest: government must regulate if the good of society as a whole is to be achieved. The sociological approach is more optimistic in arguing that people have the capacity to develop their own rules of cooperation and solidarity. As Randall Collins (1975) puts it, such faith is not 'rational' because it would always be more 'rational' to do what is best for ourselves rather than bother about other people. Social cooperation depends on an irrational act of trust: that if I give up part of my personal desires, you will give up part of yours and some mutual benefit will result. Gouldner's 'norm of reciprocity' rests on this irrational act of trust in others not to be completely 'free riders'. It is not merely a rational calculation of the cost-benefits of exchange, rather it is a consideration of the needs of others and our own positive role in maintaining social interaction. The family as the earliest and most crucial institution affecting our interpersonal, social understanding thus needs close attention insofar as it develops in children, or fails to developm, a sense of moral obligation to others.

5.26 What is of interest here is that both the market and the state regulation approach tend to remove a sense of the individual's personal stake in the fate of others from the process of moral decision-making. The market rests on individualistic moral codes; the state on collective moral codes. Should we help the collective good by paying our fair share of tax or maximise self-interest by tax cheating? Is the best solution to the drugs crisis to legalise them and let market forces prevail or to enforce laws against drug use? If army conscription fails to inspire a sense of obligation to serve one's country, will reliance on voluntary military service do any better, or just leave that obligation to those who need defence jobs as a source of personal income? Concerns for efficiency and cost seem to lead to market solutions; they result in problems of inequality and injustice so people turn to the state for solutions.

Neither the market nor the state approach encourages a strong sense of mutual obligation to others as people. Neither recognises that people are able to make their own moral rules, to rely on self- restraint, ties of solidarity, community norms and voluntary altruism in a way that bases moral regulation on a feeling of obligation to the wellbeing of others as people in a civil society.

5.27 Alan Wolfe's argument as a sociologist is that civil society has 'withered away' because of the destructive moral impacts of both the market and the welfare state approaches. He calls for a restoration of an inter-personal sense of moral obligation, a rebuilding of private family processes that will teach people to fulfil their obligations to remote strangers as well as to their intimate associates. Instead of such obligations being felt as abstract and impersonal (via taxes and state coercion, or via the lucky mechanisms of the free market) we need to revive notions of moral agency associated with civil society. Is this in fact possible, or is it yet again a call for return to some romantic past? Does family policy discussion need to address this ultimate goal of encouraging a sense of moral obligation via programs that stimulate agency rather than indifference or passivity?

5.28 The concept of civil society was associated with the rise of bourgeois society and protection of the individual against the monarchical or feudal state. Modern man was seen, as in the thinking of the Scottish Enlightenment, to be capable of self- regulation, curbing his passions to build a good society based on mutual interdependence. The moral energy of capitalism arose from the individual's responsibility for his own actions, the Protestant ethic. Several writers of the time recognised that selfish market energies would create a system of complete interdependence and so freedom would not be in opposition to society, rather freedom was social, resting on the recognition, not denial, of obligations to others. Hegel, de Tocqueville, Durkheim, Weber, Gramsci and others in the nineteenth century began to view civil society not as in opposition to the coercive state, but as standing between the free market and the state, 'a place of transition from the realm of particularism to that of universalism'. Voluntarism and localism, pluralism, Toennie's notion of Gemeinschaft, Cooley's concept of the 'primary group', the Chicago school's emphasis on local communities, Mead's 'generalised other', all reflect this view of the civil society as 'mediating' between the extremes of market and the state.

What seems to have happened, however, is that proponents and theories of the market and the state have come to hold sway over those of civil society. The modern world of bureaucratic firms rationalises away personal responsibility, denies or obscures rather than recognises mutual interdependence, organises the moral order by the same principles that organise the economy rather than understanding that the economy is made possible only because of a pre-existing moral order.

The answer does not lie in doing away with either the market or the state. The market does promote individual choice; the state provides some basic security, promotes equality and creates a better life for most. The family can be violent, communities can be racist, so an idealised view of civil society is no answer.

5.29 There is another meaning to civil society that needs rejuvenation: that is, the themes of autonomy and responsibility. We learn how to act toward others because civil society brings us into contact with people in such a way that we are forced to recognise our dependence on them. We ourselves have to take responsibility for our moral obligations, and we do so through this gift called society that we make for ourselves. Liberal democrats must recognise that reliance on states and markets does not absolve them of responsibility for their obligations to others, their obligation to protect the social order that makes their freedom possible.

The central task of modern societies is to 'extend inward obligations outward', to ensure that we do not rely too much on impersonal mechanisms of moral obligation and hence become 'out of practice as moral agents capable of finding our own ways to resolve the paradoxes of modernity. We need civil society - families, communities, friendship networks, solidaristic workplace ties, voluntarism, spontaneous groups and movements - not to reject but to complete the project of modernity'.

The underlying goal of family policies (and the criteria by which they should be evaluated) ought therefore be to encourage and extend those solidaristic ties and the sense of mutual moral obligations that make societies 'work' positively for every citizen.

In my view, family policy discussion has failed to discuss this central theme of moral obligation and social interdependence. If we were to see the rejuvenation and sustenance of 'civil society' as its goal, we might cut through some of the dilemmas of modern democracy with its confused approach to independence, family autonomy, the role of the state versus the 'free' market. Such discussion would bring to the centre of family policy topics such as the social ecology of childhood; the importance of agency in developing family supportive programs; the efficacy or otherwise of a 'public parent' versus the emotionally and time-committed parent who is 'mad about the child'; the damaging effects of family autonomy and public control and how to achieve a balance between them; and the moral as opposed to merely political dimensions of social justice.

REFERENCES

For ease of presentation, I have not referenced source material throughout this paper, but acknowledge below my use of the ideas and writings of the following authors in particular. More general sociological references are too numerous to be listed here.

• Baines, C., Evans, P. and Neysmith, S. (eds) (1991), Women's Caring, Feminist Perspectives on Social Welfare, McClelland & Stewart, Ontario.
• Collins, R. (1975), Conflict Sociology: Towards an Explanatory Science, Academy Press, New York.
• Pusey, M. (1991), Economic Rationalism in Canberra, Cambridge University Press, Cambridge.
• Schutz, A. and Luckmann, T. (1974), The Structures of the Life World, Heinemann, London.
• Wolfe, A. (1989), Whose Keeper?: Social Science and Moral Obligation, University of California Press, Berkeley.
• The views expressed in this paper are those of the author and do not necessarily reflect those of the United Nations.